Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Hall of Merit > Discussion
Hall of Merit
— A Look at Baseball's All-Time Best

Wednesday, December 05, 2007

2009 Ballot Discussion

2009 (November 3, 2008)—elect 3
WS W3 Rookie Name-Pos
535 178.1 1979 Rickey Henderson-LF
294 96.5 1988 Mark Grace-1B
245 92.0 1987 Jay Bell-SS
241 91.0 1987 Matt Williams-3B
251 63.5 1986 Andres Galarraga-1B*
189 87.3 1990 Kevin Appier-P*
206 63.4 1988 Ron Gant-LF
199 63.8 1990 Greg Vaughn-LF
200 59.8 1991 Mo Vaughn-1B
151 65.8 1991 Mike Bordick-SS
140 61.9 1982 Jesse Orosco-RP
129 49.8 1990 John Burkett-P
109 53.6 1991 Charles Nagy-P
113 50.7 1986 Dan Plesac-RP
115 46.0 1992 Denny Neagle-P
125 37.4 1991 Orlando Merced-RF/1B
038 15.4 1991 Kazuhiro Sasaki-RP

Joey Numbaz (Scruff) Posted: December 05, 2007 at 08:23 PM | 486 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

Page 4 of 5 pages ‹ First  < 2 3 4 5 > 
   301. Paul Wendt Posted: October 30, 2008 at 06:42 AM (#2999356)
Brock #281
Well, the first yearly ballot of a HoM voter is like that. There are maybe 200 plausible players who I should really sort through to get to the 15 top ones. [Agreed -pgw] I mean, look at my Babe Adams defense. I have a group of five comps, from the same team during the same 20-year period. And they're all about equal to Appier, Key, and Newcombe, plus maybe 40 more guys I didn't get to.

maybe, maybe not. Those five comps do not make a group that should embarrass you. Visit Jesse Tannehill at Baseball-Reference and see "Similar Pitchers". Nor is your hypothesis (in the LF rankings) about the role of manager-teammate Fred Clarke is intriguing, better than anything else I have ever read advancing the role of Clarke or that of Frank Chance in Chitown (player-managers of the two dynasties).

By the way, the bottom five of Jesse Tannehill's ten Similar Pitchers make an intriguing quintet:
Urban Shocker, Art Nehf, Bob Caruthers, Carl Mays, Lon Warneke

/curtail reading Brock Hanke's preliminary ballot but I can bearly wait to get back!
Brock's revenge: he makes reading the ballot daunting! (but I do look forward to returning, you're "doing great" Brock.


285. DL from MN Posted: October 29, 2008 at 09:44 AM (#2998494)
To answer another point, I don't think there is necessarily a problem with "culled" ballots. If you just limited your consideration set to everyone who received a vote in last year's election you would have plenty to choose from and wouldn't be likely to miss anyone.

I agree and suppose aloud that that is a conservative position on the matter. --new voter's daunting maybe responsibility

On the other hand, Brock Hanke is a little more likely than most to suffer some post regret by such limiting, because he is much more likely to find more and brighter gems (to his own eye) than is a random newcomer who explores beyond the bounds of 101 "incumbents" from 2008.


DL from MN #287
Dan R's numbers don't include relievers (at least in the version I have) so I'm not using his numbers entirely but I have adjusted the placement of several pitchers on my ballot.

1) Rickey!
2) Tommy Bridges - looks even better in the standard deviation adjusted numbers. Incredibly strong PWAA - even with Ed Walsh, Ted Lyons, Bret Saberhagen, Jim Bunning. Best pitcher currently eligible (though he's not as good as Kevin Brown).


For the non-pitchers I recall Dan R traversing a path away from his original focus on rates and career bulk (longevity), placing essentially no value on season bulk (durability).
--maybe not terribly far away, still a moderately strong weight on rate and longevity compared to many others
--probably he begins from the moderate position in the first edition on pitchers, no longer at the original point of that path (but the high rate, low season-workload Bridges makes me think of it)
--the polar opposite of that early DanR, I suppose, is counting "MVP-type seasons" and "All-Star type seasons" defined by Bill James using win shares; that is a focus on season bulk measures by a rating system that counts a lot for showing up at work.

38) Vic Willis - glad I didn't PHOM him this year

Is there an interpretation of downgrading Willis, with DanR's help I suppose?
Willis is overrated by some, I know, because his teammates are underrated. What a ridiculous generalization that he long suffered without support. Even in a world where W-L is the only measure, his support might have been a lot worse.


DanR #291
League-average innings munchers earn $12 million a year these days with good reason: I'm sure every single HoM voter has had the dreams of a few teams they rooted for dashed by ghastly fourth and fifth starters, or pitchers who taxed their bullpens too heavily (ahem...Scott Kazmir?).
(clapter)
That is the best short post-mortem I have read since someone's (ahem...Bob Brenly?). Oops, no, that wasn't post mortem, thanks to Mariano Rivera's first big BS.


#296-297
where Joe Dimino's most recent update on Pennants Added is located at BTF.

I have checked many a thread and google searched Joe's PA, but I cannot find the most recent version, only one that posted numbers through 1983.

Joe, if you have an updated PA, myself, along with the other newbies for 2009, would benefit from your analysis.
297. Joe Dimino Posted: October 29, 2008 at 08:46 PM (#2999019)
I do have that freak . . . I will upload when I get on that computer later tonight.


Presuming that it is too big,
Recently I uploaded a data file to yousendit.com, which provides 7 days free storage for files < 100MB.
Newcomer 'stax' recommends and uses rapidshare.

from "Pitchers for the Hall of Merit"
316. stax Posted: October 21, 2008 at 02:30 AM (#2990673)
Just a side thing, but for when the yousendit link goes down here's the file on Rapidshare.

http://rapidshare.com/files/156042846/400pitchers.csv
   302. bjhanke Posted: October 30, 2008 at 08:27 AM (#2999372)
Here is the rest of my prelim, along with comments on the holdover guys that I need to comment on. As I said before, these comments are nothing compared to my usual. I'm just running out of time, and you need at least a day or two to vet this stuff and decide if I get a final ballot.

One thing to notice. Most of the pitchers here fall into two families: the Deacon Phillippe family and the Wilbur Cooper family. Deacon's gang have about 13 years, 2600 IP, and rates around 120. Wilbur's have over 3000 IP, but a rate about 115. One of the reasons I have the few other guys ranked as high as I do is that they're not in these groups. I mean, if you rank one member of a group highly, what about the others? That poses a problem in making up a ballot. As you can see, I prefer the Deacon gang to the Wilbur gang, but an individual who doesn't fit into either group, like Babe Adams, scores well with me.

Another thing. I do refer back to my ballots in the "positionals." What I mean by that term is the discussions and ballots we've been doing to rank the guys who are already in the HoM. Not the ongoing discussions about positions. I put that stuff in to clue people in that I have theories here and am trying to be consistent. That is, I do make a raise in Lou Brock's defense because I ignore Fielding Percentage. But I do it for everyone, especially King Kelly, who is the poster boy for lousy FPs. Same thing with the other references. It's just so you can check that I'm consistent and working form theory, rather than just making theories up to fit individual players.

Again, this takes 2 posts because of the character limit on the site.

Thanks for reading.

- Brock

5. Don Newcombe
Needs Korean War credit and at least a year of pre-majors credit to develop enough bulk for this. But that credit seems the least he is due. Of course, you have to guess how good those years would have been, but the war years, at least, come after he's established and pitching well. I view his weak season in 1954 as being due to having to get his game back after the war, so I pull it up to the surrounding seasons. That gives him a very small boost, too. Then there's the bat. Without the bat, Newk would not rank up here.

6. Bobby Bonds
Fine rate, but a short career, and it was clear that he was indeed finished when he retired. I have him here because the rate is indeed high, and because his fielding is very good. Average in center field, excellent in right. Could have played a lot more center on different teams.

7. Dizzy Dean
I go back and forth on whether to cut Dizzy a break for getting hurt in the All-Star game. Today, I'm going forth, although it's not a big boost. What Dizzy has to sell is that wonderful peak and a hot partial in 1938 at just the right time to help the Cubs.

8. Luis Tiant
Very similar rate and career length to Wilbur Cooper. High length, good rate, but not great. But unlike Cooper, he wasn't in the middle of a bunch of guys who did the same sort of thing. So he stands out more. His W/L is higher than Cooper's, but that's mainly the teams. Of the Pittsburgh Five, Bill James prefers Cooper to any of the others. I don't, for reasons in the Babe Adams comment, but the career length does draw attention. I don't doubt that I have Luis about right. I do wonder if I have Wilbur too low.

9. Deacon Phillippe
When I thought there was a consensus of support for Sam Leever, I had him here, because I'm relying on culling other opinions to some extent. For what I think of Phillippe, and why I replaced Sam with the Deacon, see my Babe Adams comment. The very greatest control pitcher ever, and the staff ace of a very good staff at a time when staff ace was a very meaningful designation.

BTW, OCF (I think) mentioned that low walks guys often give up homers. That's true, but it's misleading. The guys who give up the homers are the Robin Roberts guys, who are forcing control by throwing nice high fastballs whenever they get behind in the count. The Pittsburgh guys aren't like that. They are breaking ball specialists who just have amazing control of their breaking balls. That type of pitcher doesn't give up the homers. They're like Maddux and Glavine among modern pitchers, only even more extreme.

You can see this in Bill James' comment on Bucky Walters in the last Historical Abstract. He's comparing Babe Adams to Robin Roberts, and says that both "believed in strike one," but that Roberts believed in relying on the fastball, whereas Adams believed in saving it for emergencies. Well, if you're trying to insure that the first pitch is a strike, and you're saving your fastball for emergencies, then that first pitch is a breaking ball, right? That means you just have to have amazing control of the curve. I call it being Born With Control. The Pittsburgh Five are all like that, and the Deacon here is the one born with the very most control in history. And, if you have complete control over your curve, you don't have to throw fastballs when behind in the count. So you don't give up the homers. You get the low walks, but don't have to pay the homer price. You give up grounders. I repeat: Deacon Phillippe is the answer to Barry Bonds.

There's also a factor involving the ballpark. The Pittsburgh park at the time was huge, especially in left. Homers off the Five are probably weighted to inside the park jobs down the line in left or into the left center gap. So analyzing the Five's homer rates and trying to translate them isn't very helpful unless you have some way of isolating the inside ones.

10. Hilton Smith
Of the remaining Negro League players, Hilton has the best reputation that I know of. His MLEs and rep suggest that this is about right.

11. Tommy Bond
Like so many of the early pitchers, Tommy got his arm blown out in a short set of high-IP seasons. This was the time period when the game was experimenting, trying to figure things out. One of the most important was to figure out just how much workload a pitcher could stand and not collapse in a couple of years. The only way to do that was to test a decade's worth of pitchers to destruction. Until you've gone over the limit, you don't know what the limit is. That's what happened to Bond.

If you ignore the small number of good years and look at the career totals, he's another one with high innings (about 150 more than Wilbur Cooper, whose strength is career length) and a rate comparable to others in his group, like Cooper and Tiant. A slightly lesser rate than Luis.

12. Tommy Leach
Low rate, but very long career, especially for the time period. Tremendous defense, both in the outfield and at third. Did not do well at short or second, although he never got an extended trial at either spot. I have him here by comparison to other third basemen, not other outfielders. His offense isn't so weak at third.

13. Rabbit Maranville
People dismiss him because of the weak offense. They think of him as just a very long career, where his glove carried a weak bat. But that's not the whole story. If you take defense into account, his peak, from 1914-1919, is very very good, and he probably was the MVP in 1914, when he turned in what is likely the best defensive season in history. His 14-19 offense is close to league average, and his defense is just through the roof. The prime is also good, extending out to about 1924, driven again by defense that is so spectacular that it can carry any offense anywhere near league average. Also, his defensive rates look weaker than they were, because he played so many seasons when he was old and his defense had deteriorated. And still, his defense is amazing.

Essentially, there are four, and only four, people who can contend for the title of best defensive shortstop ever: George Wright. Honus Wagner. Rabbit Maranville. Ozzie Smith. Wright and Wagner could hit, but they also played in a time where the best athletes dominated more than at any later time. Rabbit is the bridge between them and Ozzie. Some offense, and the great, great defense. My own opinion is that Rabbit is actually the best glove ever, which is occluded by the long career, where he played for years after his defensive prime. If Ozzie hadn't hurt his arm, it would be close. Honus benefited from a ground ball staff. George Wright doesn't have enough paper trail to be sure.

14. Lou Brock
I give Lou three credits that most people don't, including one that no one that I know of does, but that I proselytize for.

a. Lou was a poor outfielder with a rag arm, all true. But part of his lousy defensive ranking is a bad fielding percentage. I don't count FP, because my opinion is that errors are included in range factor, where they are just the same as balls missed by guys who lack range. All ranking systems that I know of count both FP and Range. In range, Lou is near average for left fielders. It's the FP that drags him down so far, and I think it's an illusion. I don't mean to say that he was a good, or even average, outfielder, but he was better than anyone rates him. This is the one I proselytize for: Stop counting errors twice! Range Factor does it for you; you don't need FP. For more discussion of this, see my King Kelly comments over in the right field positionals.

b. Lou does not have tremendous base stealing value, if you just use the normal break-even points. But those points are computed for runs, not wins. In Lou's time, where very few runs were scored, a single run bought a higher percentage of a win. That's why people steal more bases in low scoring times, and why the deadball guys stole so many at what appear to be such ruinous rates. If you run Markov chains for wins, I bet you'll see that Lou's SB are worth more games than their runs would suggest. Also, in SB, Lou is an outlier, and as I said about the Pittsburgh pitchers, I give those guys extra credit for being able to force issues. Lou could do that, especially in stealing third base and in stealing when the game was late and tight and everyone knew that he was going to go.

c. I give Lou Brock World Series credit. I seldom do that, but I do for Lou. He was in three series and was a star in all three. He never had a weak series, like Reggie Jackson, and he had three tries to fail. That has value.

15. Kevin Appier
Probably will not survive in the final ballot. There are too many guys who are similar. Comparable career length and rate to Deacon Phillippe and Sam Leever. Kevin's W/L is much lower, but that's his teams. I tend to give preference to early guys when this happens, because they had worse playing conditions, so you should expect their careers to be shorter. He was the staff ace of his teams, but that meant less in his time than it did in Phillippe's. I have him about even with Leever.
   303. bjhanke Posted: October 30, 2008 at 08:28 AM (#2999373)
Walters, Bucky
Another Wilbur Cooper type. Among the group, Walters has a high rate and a low length, but both are within the parameters of the group, as opposed to real high rate / low length guys. Bucky hit well for a pitcher, of course, but there's no real value to be added by considering his play at third. He was moved to pitcher because he was hitting like one. A very good fielder for a pitcher, of course. The added hitting and defense means that I won't criticize anyone who has him higher than I do.

Redding, Dick
My read on his reputation is that he was not considered to be as good as Hilton Smith, and by a reasonable margin. I place a lot of weight on rep when dealing with the Negro Leagues, because hard stats are so hard to find.

Puckett, Kirby
Compare to Indian Bob Johnson:

NAME GAMES OPS+
Kirby 1783 127
Bob 1863 138

Even after dropping Bob's rates some for WWII, Kirby is behind on offense, and has even less playing time. Kirby was the better outfielder, of course, but not by a huge amount. Certainly nothing like the offense gap. Johnson was an excellent left fielder, just short of being a center fielder. Bill James has Johnson's grade too low because his system favors center fielders. Like Puckett.

You can complain that I have Kirby ranked too low, out of the top 15. Well, I have him lower than Bob Johnson, for what I think are good reasons. I also have him below Lou Brock. Kirby was a better hitter and a better fielder, but in neither case is the margin as large as it seems. Because of the lack of a decline phase, Kirby's OPS+ is too high, as are his defensive numbers. The gap between his OPS+ and Lou's should be halved, IMO. As for defense, it's a matter of adjusting Lou, not dropping Kirby. When I factor in Lou's career length and other stuff, he comes out ahead.

Johnson, Bob
I have to drop his rates after 1941 because of war competition. That leaves him with a short career and a lower rate, although it's still good. The problem with the career length was a slow start. Bob didn't get to the majors until he was 27, and the reason was that he didn't hit well in the minors until he was 25. No minor league credit, not that I normally give any to anyone except converted negro leaguers. I am completely sure that Bob outranks Kirby Puckett.
   304. bjhanke Posted: October 30, 2008 at 10:45 AM (#2999405)
Paul says, "By the way, the bottom five of Jesse Tannehill's ten Similar Pitchers make an intriguing quintet:
Urban Shocker, Art Nehf, Bob Caruthers, Carl Mays, Lon Warneke"

That's not just intriguing. It's downright hilarious. I would never have thought that any of the five would be comparable to Tannehill, or even to each other. When the pitcher ballot gets done, I just have to work up the group to see what is there. Tannehill is a very interesting pitcher whom I may have underrated. A weird career with multiple weird teams. You should not treat my positioning of him among the Pittsburgh Five as final; it's mostly a result of his having the lowest ERA+ in the group, which can be misleading. This comp list is interesting in that it supports that idea. Those are five pretty good pitchers there. I just never thought of any of them as being similar to Tannehill. - Brock
   305. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 30, 2008 at 12:26 PM (#2999430)
DL from MN--wins above average is flawed because I don't think it accurately reflects real value on the field for pitchers, not because it's difficult to calculate.

Paul Wendt, yes, you're exactly correct. The group convinced me to value in-season durability, because I realized that by valuing 10 WARP in 500 PA more than 10 WARP in 700 PA, I was effectively raising my replacement level well above the empirically accurate level that I have advocated for so tirelessly.

bjhanke, no one looks at fielding percentage here. Brock was, just as you say, league average at turning balls in play into outs (according to TotalZone, which is obviously a much more sophisticated tool than raw Range Factor), but he was really atrocious at throwing. Opposing runners accumulated more value above average running on Brock than Brock himself accumulated running on opposing catchers and outfielders!

As for the baserunning itself, he was 32 runs above average according to Dan Fox's EqBRR (the greats like Coleman/Wilson/Rickey/Raines are in triple digits). The difference between a 4 RPG environment (Brock's) and a 4.75 RPG environment (the modern game) is only one run-per-win: 9 runs/win in 4 RPG, and 10 in 4.75 RPG. (The equation I use is 3.33*RPG^.71). So the run environment only increases the contextual value of his SB by a whopping 0.3 wins. The bottom line is that he just wasn't a particularly great asset on the basepaths.
   306. Chris Fluit Posted: October 30, 2008 at 03:24 PM (#2999589)
First, bjhanke: Thank you for your comments. Though I don't agree with your placement or preference of every player you mentioned (and none of us agree perfectly), you've certainly exceeded the requirements for comments. I'm not the biggest presence in the HoM, but since you mentioned me by name, I have no objection to your ballot being counted.

One thing I would like to address:
Redding, Dick
My read on his reputation is that he was not considered to be as good as Hilton Smith, and by a reasonable margin. I place a lot of weight on rep when dealing with the Negro Leagues, because hard stats are so hard to find.


I do think it's reasonable to look at reputation when considering Negro League players because the statistical evidence can be scarce. However, there are times when reputation can be mis-leading. Concerning Redding vs. Smith, Redding had all of his best years before the introduction of league play which works against him in terms of reputation. Meanwhile, Smith had a couple of advantages regarding reputation: all of his best seasons at the time when the Negro Leagues had the greatest attention, several noteworthy teammates who brought attention to the team while he was playing and after (namely, Satchel Paige and Buck O'Neil), higher caliber of teammates resulting in more championship appearances. I'm not arguing against Smith as a legitimate candidate. I even voted for him a couple of times myself. But, the nature of the eras goes a long way towards explaining the discrepancy in reputations between the two pitchers. Similarly, Catfish Hunter's reputation far exceeded Bert Blyleven at the time, but in retrospect, Blyleven was the much better pitcher.
   307. Paul Wendt Posted: October 30, 2008 at 03:52 PM (#2999630)
Brock Hank
3. John McGraw
Everyone knows the basic case here, right? Very high rates, enormous OBP, very short career due to injuries.


He missed 1896 with a deadly disease that did not prove fatal (typhus?). When he returned, he was the on-base machine of fame. He missed spring training and some weeks(?) of a later season with another severe illness (smallpox?) and/or under quarantine.

Cliff Blau may be a McGraw expert.

. . .
Lyon's OPS+ is 139. McGraw's is 135. Joyce's is 143. Lyons played 1121 games, McGraw, 1099, and Joyce only 904. But still, they're all pretty much in a group. To distinguish them from each other, you need to look at stuff that isn't in OPS+: defense and baserunning.

In baserunning, McGraw has the clear victory. He stole about twice as many bases as the other two, in an era where that was a much more important stat than it is now (in general, SB are more important in low-scoring eras than in high ones, because SB is a one-run strategy).


very low-scoring when he urged Josh Devore and Freddie Snodgrass to run but how low-scoring when McGraw was on the basepaths himself, for HOM discussion essentially NL 1897-1901?

. . .
So which is a better indicator? Third base range factors or getting in some shortstop? I don't know. Third base is a strong defensive spot at the time. There's a lot more to defense than raw range factors. Bill James has McGraw graded as a B+ third baseman, which is very good. He has Lyons at C-, which means he's looking at a lot more than raw range factors. He has Joyce at F, which is completely consistent with Joyce's contemporary reputation.

If you believe Bill, McGraw is your man. I think he probably is, because he has the SB edge, and because he has two of three indicators on defense (shortstop and Win Shares, opposed to raw range). But there is a possibility that Denny Lyons was, indeed, a bit better. None of the defense indicators is definitive. There is a much smaller probability that people have Joyce's defense ranked all wrong, in which case his offensive rates pull him up.


There is some sabrmetric agreement (n>1) about Joyce in the field. (For SS Ed McKean, too, iirc, there is agreement among contemporary reputation and multiple sabrmetrics that he was a poor fielder at best. Anyone else?)

For Denny Lyons it's a mixed bag and I don't know his contemporary reputation. Arlie Latham was the great 3Bman by reputation, in the field, on the bases, and overall --but overall was largely governed by the field and the bases. In the NL there was Billy Nash in the field, the one who finally bumped Ezra Sutton off the bag in Boston.

In short, I'm pretty convinced about McGraw. What I'm not sure of is that I should not include Denny Lyons on my list somewhere.

That is reasonable.
Or even very good or excellent. (Careerists will qualify that severely, of course. Something like "Excellent, given that you have never heard of career.")
   308. Mike Emeigh Posted: October 30, 2008 at 04:12 PM (#2999654)
Similarly, Catfish Hunter's reputation far exceeded Bert Blyleven at the time, but in retrospect, Blyleven was the much better pitcher.


Definitely not "much better", and maybe not even "better".

I don't have my support numbers handy - I'm going from memory, because it's been a while since I looked at this - but among contemporary HOF candidates, Blyleven was one of the worst pitchers in the group in terms of leveraging his support, where Hunter was above-average in that regard. By that I mean that if you look at the numbers on a game-by-game basis, Hunter was better than Blyleven in converting run support into wins. Give either pitcher three runs in seven batting innings of support, for example, and Hunter's team was more likely both to be ahead in the game when he left, and to actually win the game.

-- MWE
   309. Paul Wendt Posted: October 30, 2008 at 04:23 PM (#2999665)
Brock #303
NAME [ in full ] GAMES OPS+
Kirby [Puckett] 1783 127
Bob [Johnson] 1863 138

Even after dropping Bob's rates some for WWII, Kirby is behind on offense, and has even less playing time.


This may be the occasion for some league-rank or standard deviation analysis. Offhand I think that may be more important than the WWII discount, which should be tiny at first (Greenberg is not in the field but that is only about 1% of league batting weight), although Bob Johnson's very best seasons needs serious discount when side by side with Greenberg's 1945, for example. (heh heh, that is another unwise-crack)


304. bjhanke Posted: October 30, 2008 at 06:45 AM (#2999405)
Paul says, "By the way, the bottom five of Jesse Tannehill's ten Similar Pitchers make an intriguing quintet:
Urban Shocker, Art Nehf, Bob Caruthers, Carl Mays, Lon Warneke"

That's not just intriguing. It's downright hilarious. I would never have thought that any of the five would be comparable to Tannehill, or even to each other.


They have superior batting skill in common, and era in common --or they would with Burleigh in place of Parisian Bob. Beside anything else you may be thinking, it's remarkable that they rank #6-10 here because this is similar pitchers without any consideration of batting skill (see similar batters).

[
For those who haven't yet followed the reference, what's wrong with you?, Jesse Tannehill's first five "Similar Pitchers" all played for Pittsburgh including three as teammates on the 1900-1902 staff --the entire core of the pitching staff, with Waddell in 1900, Doheny in 1901-1902.
Jesse Tannehill, similar pitchers
# Deacon Phillippe (938)
# Jack Chesbro (923) *
# Babe Adams (921)
# Sam Leever (920)
# Ed Morris (917)
Babe Adams 1907-1926 after a 4-inning debut in St Louis (aesthetic downfall here), Ed Morris 1884-1890.
]


DanR #305
Paul Wendt, yes, you're exactly correct. The group convinced me to value in-season durability, because I realized that by valuing 10 WARP in 500 PA more than 10 WARP in 700 PA, I was effectively raising my replacement level well above the empirically accurate level that I have advocated for so tirelessly.

But at the same time, I mean to say, that "lesson" is now also in your pitcher ratings in terms of BFP or IP, right? So if someone reacts, "Tommy Bridges, way too high" the reason cannot be that your pitcher ratings await the same adjustment. It must be something else, akin to someone's reaction today, "John McGraw, still way too high".

(I don't think either one is "way too high", partly because placement on the annual ballot is fraught with optical illusion, given the stage of the project and given that Rickey Henderson and other newcomers provide no useful markers. Rank 2 of 15 looks "way too high" for anyone you might put there except Henderson, if you forget the context. What it means is something like "best of 50 guys whom I see between Henderson and the next newcomer", where the consensus opinion is that those 50 rank between 200 and 400 all-time. In some sense rank 2 may be closer to rank 100 than to rank 1.
   310. Joey Numbaz (Scruff) Posted: October 30, 2008 at 04:28 PM (#2999675)
Mike, isn't 3 runs of support in Oakland a much better offensive support than 3 runs at Three-Rivers? That is, it was easier to make 3-runs stand up that great pitcher's park, no?
   311. Joey Numbaz (Scruff) Posted: October 30, 2008 at 04:29 PM (#2999676)
Three-Rivers or The Met. Both were pretty good hitter's parks.
   312. Paul Wendt Posted: October 30, 2008 at 04:39 PM (#2999690)
Brock #284
I couldn't find a link to the final results in the center field ranking among HoM center fielders, but I remember Doby as being somewhere in the upper middle. There is no way to get from there down to "does not belong in the HoM" using the difference between Doby and Smith.

For a while, Brock and other newcomers or recentcomers, all of the threads you know by experience will be easy to find in the first page of the archives. Rather than post a link to that archives catalogue, here are the directions.
Visit the "Hall of Merit" (home page);
select "Important Links";
select "Hall of Merit Archives".

Unfortunately the archives catalogue is a pain to use for archived material that does not show up on the first page.

In the centerfield ranks, the group mind seemed to expect Larry Doby at rank 11 but one enormous gap behind #10.
Jim O'Rourke (up), Duke Snider (down), and Paul Hines (up) overturned that expectation in spectacular fashion, and halved the gap, but Torriente #9 and Doby #13 fared just about as expected!

On Reggie Smith,
Reggie is a clear HoM guy. My only question to myself is whether I have him too low here. I think I do. If you read the Adams and McGraw comments above, you'll see that I'm taking groups of comparable players, none of whom is in the HoM, and arguing that Adams and McGraw are the top ends of those groups. In Reggie's case, I'm arguing that his comp is Larry Doby, who is not only in the HoM, but is way over the borderline. That's a much stronger argument. If I had to fill out my final ballot right now, Reggie would be #2.

Good show. The time you are able to spend is productive, at least for me as a reader.
It's valuable to have someone come from outside --cold, not as returnee or longtime lurker-- with fifty years experience watching or talking baseball seriously, and time enough to write as much as you do.
(I have read your ballot #2 to #4, which is enough for Wednesday and Thursday.)
   313. Best Dressed Chicken in Town Posted: October 30, 2008 at 05:30 PM (#2999777)
This is the one I proselytize for: Stop counting errors twice! Range Factor does it for you; you don't need FP.

Not necessarily for outfielders. Most of their errors are throwing errors, or letting-the-ball-skip-by-them errors. So eliminating the error would not add another play made. There are very few muffed fly balls, at least in the past 50+ years.
   314. Paul Wendt Posted: October 30, 2008 at 05:37 PM (#2999785)
Brock Hanke #302
9. Deacon Phillippe
Born With Control. The Pittsburgh Five are all like that, and the Deacon here is the one born with the very most control in history. And, if you have complete control over your curve, you don't have to throw fastballs when behind in the count. So you don't give up the homers. You get the low walks, but don't have to pay the homer price. You give up grounders. I repeat: Deacon Phillippe is the answer to Barry Bonds.

Generally I sympathize with the theme. I doubt that bottom line. Historical context is extremely important (as we all say) and in large part that is competitive context (as few of us say). By training, systematic or otherwise --that is by explicit teaching and instruction-following or by competitive adaptation-- pitchers and batters at any time are byproducts of the current and recent pitchers and batters. And umpires and catchers and managers.

In Deacon Phillippe's time, what power hitters threatened a pitcher may have been more likely to hit a high strike over an outfielder's head than to jack a low strike (or even a would-be called ball). That is enough to illustrate the point, and all the time I can spend now.

There's also a factor involving the ballpark. The Pittsburgh park at the time was huge, especially in left. Homers off the Five are probably weighted to inside the park jobs down the line in left or into the left center gap. So analyzing the Five's homer rates and trying to translate them isn't very helpful unless you have some way of isolating the inside ones.

Mike Emeigh or KJOK may be lurking with detail knowledge from Pittsburgh.
The Pittsburgh Five did not work in the same ballpark, although three of them (Tannehill, Phillippe, Leever) did so along with Chesbro and a generation after Ed Morris; and two of them (Phillippe, Leever) did so along with Willis.
Babe Adams was still a babe when they moved to Forbes Field and Wilbur Cooper was a schoolboy.

For the others, what you say about home runs may be true regarding the majority of major league teams home ballparks. --outfielders did play closer to the action, so there were home runs on fly balls hit over their heads, but inside-the-park homeruns were routine manywhere.

Many or most of the classic ballparks were constructed and filled by NL and AL teams 1909 to 1916 where the latter marks Chicago North Side's move from Federal League to NL. If I understand correctly, those ballparks changed the home run possibilities significantly and differently.

Perhaps ballparks that fostered itp or alternatively otf homeruns were always common. Perhaps the difference after 1909-1916 was that every team played in a permanent ballpark. Earlier there had been plenty of mlb parks with frequent otf homeruns somewhere, maybe even behind the entire outfield. But they were temporary or permanently temporary ballparks for mlb teams. (permanently temporary: We won't be here long. We will have a real major league ballpark when we can afford it.)

/stop. having read #3-15 plus notes. That's enough for Wed, Thu, and Fri
   315. karlmagnus Posted: October 30, 2008 at 05:49 PM (#2999799)
BJhanke, I give Leever extra credit for not having started till 27. His early years were in the deep 1893-96 recession when baseball had contracted recently, so he stayed schoolteaching and didn't go in for baseball until times improved. His published career is thus artificially short, which needs adjustment and pushes him decisively ahead of Adams, although not so much ahead of Philippe, who was only 5 months younger (I don't buy your ace argument, which may only apply to 1903 and not other years.)
   316. Paul Wendt Posted: October 30, 2008 at 05:50 PM (#2999800)
(cant' stop)
All ranking systems that I know of count both FP and Range. In range, Lou is near average for left fielders. It's the FP that drags him down so far, and I think it's an illusion.

Many people use fielding percentage and range factor to judge fielding. In the third millenium I suppose they are few but people who look up players in a web encyclopedia such as baseball-reference (maybe some print encyclopedia too?). Check career FP and RF against league. They may look up several players a few times a month, or vice versa, hence hundreds of players in a year, probably with some repetition. I suppose that few who use computers for more than the internet now have rating systems that rely on FP and Range. Those are for quick-and-dirty use by systematic people and for people who only use baseball records by lookup. I suppose.
   317. mulder & scully Posted: October 30, 2008 at 06:44 PM (#2999842)
If you wanted to look at possible indications of "Ace"-hood or -ness, you could look at a pitcher's opponents. Who did Clarke use against Matty or McGinnity? I don't have the time to check - full time job and 2 graduate classes, but Retrosheet could help answer the question.

2 cents
   318. bjhanke Posted: October 31, 2008 at 12:38 AM (#3000131)
Dan says, "As for the baserunning itself, he was 32 runs above average according to Dan Fox's EqBRR (the greats like Coleman/Wilson/Rickey/Raines are in triple digits). The difference between a 4 RPG environment (Brock's) and a 4.75 RPG environment (the modern game) is only one run-per-win: 9 runs/win in 4 RPG, and 10 in 4.75 RPG. (The equation I use is 3.33*RPG^.71). So the run environment only increases the contextual value of his SB by a whopping 0.3 wins. The bottom line is that he just wasn't a particularly great asset on the basepaths."

You may be right. I'm working from memories of Brock, and while I have a lot of them (I live in St. Louis), they are all from my teens and twenties. I'd have to go chase down game logs to work it all out. Obviously, that isn't going to happen by Tuesday.

Chris Fluit says, "I do think it's reasonable to look at reputation when considering Negro League players because the statistical evidence can be scarce. However, there are times when reputation can be mis-leading. Concerning Redding vs. Smith, Redding had all of his best years before the introduction of league play which works against him in terms of reputation."

All true. And I'll get around to looking at all that during the year before the next election. This stuff takes too much time to make deadlines. Oh, and the reason I mentioned you by name is that I didn't remember seeing your handle on the positional posts. I assumed that you did not know that writing just a very little is not my strong point. The guys who have followed those elections know that I'm a wordhack. And yes, I do make my living as a writer, mostly of computer systems documentation, but occasionally something else. The odd thing is that I can, when I need to, write to word count. Did it for years for a newspaper. It just takes effort that I don't usually put in because I'm trying to be thorough.

Paul says, about John McGraw's base stealing era, "very low-scoring when he urged Josh Devore and Freddie Snodgrass to run but how low-scoring when McGraw was on the basepaths himself, for HOM discussion essentially NL 1897-1901?"

Aaargh! You are SO right. I've been doing this to myself for years. My head just can't get rid of the idea that John played in the deadball era. He didn't. He played in the 1890s, which was a high-scoring era. Please nail me whenever I do this to myself. Thanks SO much.

Mulder and Scully say (says?), "If you wanted to look at possible indications of "Ace"-hood or -ness, you could look at a pitcher's opponents. Who did Clarke use against Matty or McGinnity? I don't have the time to check - full time job and 2 graduate classes, but Retrosheet could help answer the question."

Very true. Exactly what I need to do. Ain't gonna happen by Tuesday. Oh, and BTW, I promised to send you a couple of my old BBBAs that you missed, but have forgotten which two years. Between Don Malcolm and myself, we have them all. Just post up, and we'll get them out to you.

Karlmagnus says, "BJhanke, I give Leever extra credit for not having started till 27"

That's fine as long as you do it to everyone, which ain't easy. Check out Sam Thompson and Bob Johnson, to start. The main reason I don't do it is because of people like Johnson. As I said in his comment, there are people who didn't get to the bigs until late simply because they didn't get their games together until late. Separating the two groups is really really hard and unreliable, so I don't do it. Then there's Ken Boyer, who I believe lost 2 years of MINOR league play to the Korean War, delaying his entry into the bigs, and who gets punished in analysis for a short career. Giving Negro League credit, where you have something like Chris Cobb's MLEs based on incomplete stats, is hard enough. Check out Chris Fluit's comment on Dick Redding. He is absolutely right. Redding suffers because he was older before the NEGRO major leagues got started. There is simply no way to really assess how good he was when young. Then there are all the guys who were older when the National Association got started, including some tremendous players, like George Wright. When you're dealing with games for which there are no stats, it's hard to be fair to the whole group.

Oh, and for everyone who posted: Thanks for letting me know that modern analysts have figured out about the errors thing in range factor. I wasn't aware that people had gotten to that one. Last time I was really plugged into a sabermetric community, no one did that, and so I fell into preaching. Thanks again. I'll shut up now. - Brock
   319. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 12:41 AM (#3000138)
Paul Wendt, yes, I apply the same salary estimator to both pitchers and hitters.
   320. Chris Cobb Posted: October 31, 2008 at 03:35 AM (#3000219)
OK. Here’s the first post with actual data following my attempt to lay out a pitcher consideration set that included every pitcher worth giving a serious look. It’s a study looking at pitchers’ career runs above replacement.

Background

Dan R has been arguing lately for the importance of rating pitchers based on value above replacement, if replacement level can be established. The electorate has available to them two readily available systems that give something like value above replacement—WARP and Win Shares—but it is clear that neither system sets its point where value = zero reliably near to any empirically determined replacement level.

The electorate has quasi-access to two systems that strive for a much more accurate calculation of value above replacement: Joe Dimino’s Pennants Added and Dan R’s WAR. Unfortunately, the most recent work of neither is currently available, though Joe’s should be soon. Their systems, when fully implemented, have the added advantage of calculating the marginal value of high seasonal peaks, so that the real value of peak seasons is accounted for as part of career value above replacement.

Partly because their data isn’t around yet, and partly because I like figuring, I have put together a career runs above replacement measure (RAR) for all post-1892 major-league starting pitchers who are part of what could be called our consideration set: unelected, eligible pitchers who either a) received a vote in the 2008 HoM election, b) appeared in the Bill James top 100 pitchers list, c) pitched 4000+ innings in this era, or d) were mentioned by someone as an omission from my original list.

These career RAR totals are meant to offer a rough sorting of our pitching candidates, not a definitive ranking. The sorting is rough because a) it doesn’t attempt to establish the value of peak performance or b) measure seasonal peaks or c) do any fine-tuning such as adjusting for standard deviation or d) offer compensatory credits or penalties for things like military service, or e) make contextual adjustments to innings pitched. Given how closely bunched pitchers in the backlog are, these sorts of fine-tunings could significantly affect the order of one's rankings. These lists should, however, help to identify some pitchers that merit close attention who are not getting it and some other pitchers who don’t merit the attention they are getting.

I have calculated the RAR totals as the sum of career pitching runs above average (PRAA), career batting runs created above position (RCAP), career fielding runs above average (FRAA), and a replacement level modifier.

PRAA is calculated by multiplying WARP1’s DERA (defense-neutral run average) by the pitcher’s career innings pitched. DERA adjusts for quality of defensive support and normalizes pitchers to a 4.50 run-environment. I use this method rather than just grabbing WARP’s PRAA totals because those are calculated using their adjusted innings pitched, which is, as Dan R has plainly put it, a junk stat. My premise here is that an inning pitched is an inning pitched, and a run saved is a run saved.

RCAP is taken from Lee Sinins’ Baseball Encyclopedia. Some RCAPs have been estimated: the old Windows laptop on which I accessed my encyclopedia copy died recently (the Encyclopedia is not Mac-compatible), so I couldn’t get the data for some pitchers recently added to the consideration set. Using WARP’s BRAA as a benchmark, I have been able to estimate trustworthy RCAP totals for these pitchers based on the RCAP/BRAA ratio for pitchers who are their contemporaries.

FRAA is taken straight from WARP1.

The replacement level modifier is calculated as the difference in value between an average pitcher and a replacement level pitcher, per inning, in a 4.50 run environment. Based on a study (not comprehensive) of the ERA+ of starting pitchers who lost their jobs the following season, I believe that an ERA+ of 85 is around replacement level for most of post-1892 pitching, so that is what I have used. The difference between an 85 ERA+ pitcher and a 100 ERA+ pitcher works out to about 20 runs / 250 ip, so I have added 20 runs / 250 career IP to each pitcher’s total to get career RAR.

Note that two elements that go into career RAR—PRAA and FRAA—are normalized stats, but RCAP is not. This may make for some small inaccuracies, especially for pitchers with very high or very low RCAP values whose run environments deviated a great deal from 4.50 r/g. But as the sorting I am doing here is just a first effort to group the candidates and not a basis for final rankings, I won’t sweat this detail.

I have broken up the pitcher listings into several groups, based upon comparisons to Bucky Walters. Walters is the highest-ranking returning pitcher on the ballot. If no newcomers besides Rickey place higher in the rankings, and the order of the backlog ranking does not change, Walters will be elected. Since the electorate is going to be somewhat different this year (how different we don’t know), I expect the backlog will shift quite a bit, but who knows how it will. I have never been a supporter of Walters, but I have been trying to look at his case freshly this year, as I hope everyone will do.

For purposes of comparing the rest of the consideration set to Walters, I have sorted the post-1892 MLB pitchers into four groups:

1) pitchers with more IP and a lower DERA than Walters

Of the 66+, there were 14 pitchers in this group. Listing them in order from highest to lowest IP, they are Willis, Quinn, Koosman, Newsom, Friend, Reuschel, Tiant, Cooper, Harder, Jackson, Cicotte, Luque, Leonard, and Finley.

2) pitchers with more IP but a higher DERA than Walters

There were 14 pitchers in this group, also. In IP order: John, Kaat, Powell, Tanana, Grimes, D. Martinez, Morris, Hough, Hoyt, Lolich, Hunter, Blue, Hershiser, Uhle

3) pitchers with fewer IP but a lower DERA than Walters

There were 26 pitchers in this group: Bender, Adams, Breitenstein, Langston, Shawkey, Cone, Viola, Bridges, Warneke, Vaughn, Trout, Tannehill, W. Wood, Trucks, Shocker, Leever, Appier, Key, Rommel, Gomez, Guidry, Rucker, Joss, Hahn, Dean, Parnell. (I dropped Joe Wood because his major-league IP total is very small, and the multi-position career makes RCAP very difficult to figure out.)

4) pitchers with fewer IP and a higher DERA than Walters

There were 5 pitchers in this group: Mays, Gooden, Phillippe, Newcombe, Sain

Including Walters, that’s 60 pitchers. To my knowledge, no one with anything like a serious HoM case has been omitted. (We can safely let the cases of George Mullin, Earl Whitehill, Paul Derringer, Larry French, Freddie Fitzsimmons, Rube Marquard, and so on, lie unconsidered.)

Over the next couple of days, I’ll post the table for each of these four groups, with Walters included in each, and some discussion. Here is the first.

Group 1: Pitchers with more IP and a lower DERA than Bucky Walters.

Pitcher  IP    DERA PRAA RCAP FRAA RAA Rate*  RAR 08 Rank
Reusche 3548.3 3.79  280   6   17  303 21.34  587  29
Quinn   3920.3 4.06  192 
-10   33  215 13.69  528  68T
Tiant   3486.3 3.83  260   5  
-20  245 17.54  523  12
Willis  3996   4.04  204 
-36   14  182 11.40  502  20
Leonard 3218.3 3.84  236 
-19   20  237 18.41  494  
Cicotte 3226.7 3.90  215  
-2   16  229 17.75  487  56T
Koosman 3839.3 4.10  171 
-16   -2  153  9.94  460  
Cooper  3480   4.10  155  36  
-15  176 12.62  454  56T
WALTERS 3104.3 4.14  124  57   19  200 16.12  449   5
Luque   3220.3 4.14  129  31   4   164 12.72  421  
Jackson 3262.7 4.08  152  
-6   11  157  12.05 418    
Finley  3197.3 3.99  181   0  
-24  157  12.29 413  83
Friend  3611   4.13  148 
-25    0  123  8.55  412    
Newsom  3759.3 4.13  155 
-23  -25  107  7.09  407    
Harder  3426.3 4.14  137 
-28    8  117  8.54  391  
*rate is RAA/250 IP 


From the standpoint of career value above replacement, as measured here, it is not clear why Walters is our top-ranked candidate, as eight pitchers in this cohort alone have more career RAR value.

Now, it’s easy enough to see how the contribution of Walters’ peak value would likely move him ahead of Koosman, how Cicotte’s case would collapse under game-throwing penalties, but moving Walters higher than that in this list is not easy. Leonard, who is next above Cicotte, is 5 wins ahead of Walters, and Reuschel is 15! Wilbur Cooper, immediately ahead of Walters, had an excellent peak himself (his career rates are dropped by a couple of terrible seasons), and Dolf Luque, immediately behind Walters, has a top-notch peak season and, with CWL credit, could gain significantly in career value, depending upon one’s view of that part of Luque’s career. There’s a lot of competition here, most notably from Leonard on up.

The pure peak voter may not care about career, but 5-15 wins is a lot of value to disregard in preferring Walters on the basis of peak. It’s worth noting, I think, that, looking also at RAA and career rate, Reuschel, Tiant, Leonard, and Cicotte all have higher RAA totals and better career rates, meaning that, for their careers, they threw more innings than Walters and were better on a per inning basis than Walters was. To argue that Walters’ seasonal durability and peak rate can outweigh the fundamental “pitched more and pitched better” case that these pitchers have seems to me problematic. Cicotte, of course, has the gambling anchor sinks him in the backlog. Leonard has no peak to speak of, and little in-season durability cachet at all, and Tiant was no workhorse in the context of his time. His IP totals were acceptable for a #1 starter, though, and he had some outstanding seasons. Reuschel is solid on peak, all-around play, and durability. He has a good peak, including one extraordinary year to match Walters’ 1939 and solid though not top-notch durability. He has the best DERA in this group, was a good hitter, and a good fielder. The knock on Reuschel is that he didn’t stand out in relation to his peers as Walters did, because his prime was in the 1970s along with those of Seaver, Carlton, Perry, Niekro, Palmer, Jenkins, Ryan, and Blyleven, all of whom rank ahead of Reuschel. Walters had the good fortune to hit the seam between the great primes of the Grove/Hubbell/Dean early-1930s cluster and before the great primes of Newhouser/Spahn/Roberts after WW2. His only peak competition is Bob Feller, who is conveniently in the other league, and NeL pitchers Satchel Paige, Ray Brown, and Hilton Smith. Once those pitchers are factored in, Walters does not stand out against his peers quite so much, but he was clearly dominant against his National League peers during his peak in a way that Reuschel was not.

It’s also worth noting that Vic Willis, who trails Walters in career RAA, was as much or more of a durability monster than Walters, and, like Cooper, his career rates above average suffer from a couple of very bad seasons, which a peak analysis would factor out. He, like Reuschel, has the problem of a great peer group, but without some serious standard deviation adjustment (which might be proper) it’s hard to see Walters making up the difference in career value with peak.

On the other hand, it’s also clear that in this company, Jackson, Finley, Friend, Newsom, and Harder don’t measure up. All of them being at or near 400 career RAR, they don’t lag terribly far behind, however.

In sum, by this RAR measure, Reuschel’s case to be the best of the lot is strong. Tiant, Quinn, Willis, Leonard, Cooper, and Luque should also be included with Walters in a conversation about the best eligible ML pitcher (with Cicotte thrown in for those who aren’t bothered by the fact that he threw games as well as pitches).
   321. mulder & scully Posted: October 31, 2008 at 04:01 AM (#3000222)
Chris,

Great post. I look forward to the rest. I currently have Walters, Willis, and Cooper on my ballot but have been rethinking much of my ballot recently.

Cicotte is a non-starter for me because I zero out his 1919 season which reduces his value considerably.
Leonard has the lack of peak and lack of durability and is barely in the consideration set.
From what I remember, Quinn has a lack of peak as well though he was effective forever. The one year in the top 10 in innings is also troublesome. I lean to more of the workhorse peak pitcher than the marathoner.
Koosman, again the problem of lack of big years. I just don't like the long, low-prime accumulators.
Tiant and Reuschel (Hey, I spelled it right.) need more consideration than I have given them previously.

Thanks, Chris.

Back to my midterm.
   322. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 12:02 PM (#3000284)
Chris, my pitching numbers most certainly are available! I don't entirely swear by them, for the countless reasons I've mentioned previously, but I'm happy to send them to anyone who wants them. Also, for post-1987 seasons, I would *strongly* recommend adding up the team FWAA on my new WARP v2.1 (whose defensive numbers have about a .85 correlation with PBP metrics) and using those instead of BP for defensive adjustments for pitchers.

I think an 85 ERA+ is too high for starting pitcher replacement level, Chris Cobb. You're using 20 runs (BP runs, I imagine, which are 1/9 of a win, so 2.2 wins) per 250 IP, and I use 2.1 wins per 200 IP, which would be 2.6 per 250. Tangotiger finds that long relievers (who get promoted to start when a starter goes down) are .410 pitchers as starters (2.0 wins per 200), and then you have to figure that the gap between a long reliever promoted to starter and the mopup man who replaces him is at least another 0.1 win (probably 0.2 or 0.3 actually; I should probably drop my replacement level an extra 0.2 points to 2.3 per 200 for that factor). The difference between 1.75 wins per 200 IP and 2.3 is quite substantial, a full 8 wins over a typical HoM-length career. This matters not just in terms of ranking pitchers among themselves, but also in terms of the right % of pitchers to have in the HoM. Where did you get 85 from?
   323. DL from MN Posted: October 31, 2008 at 01:49 PM (#3000345)
Once again Chris comes through and provides exactly what I was looking for. When can I steal your spreadsheet?

From the calculations I did this week using Dan R's WARP - Dutch Leonard suffers a lot with a standard deviation adjustment along with most pitchers of that era (Trucks, Trout, etc).

Just for reference, how low is the BP PRAR replacement level - 3 wins per 200 IP? 4?
   324. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 02:24 PM (#3000376)
It seems to vary wildly. For Nolan Ryan, it's 5.6, while for Cy Young, it's 3.3 (these are seasonally adjusted numbers, not the all time adjusted ones). I imagine this has something to do with how BP handles the pitching/fielding split. Anyways, this only teaches us what we already knew: that PRAR is the busted sister of the already beaten-with-an-ugly-stick PRAA. Fortunately, moving from DERA (which is definitely reliable, as long as you trust BP's quite questionable defensive adjustments) to legitimate pitching wins above replacement is delightfully simple arithmetic.
   325. DL from MN Posted: October 31, 2008 at 02:29 PM (#3000382)
Should "replacement" level be empirically calculated for pitchers using the bottom 3/8 of the barrel or can we use overall runs allowed to do the work for us?
   326. Mike Emeigh Posted: October 31, 2008 at 03:00 PM (#3000407)
Mike, isn't 3 runs of support in Oakland a much better offensive support than 3 runs at Three-Rivers? That is, it was easier to make 3-runs stand up that great pitcher's park, no?


Over his career, Hunter's starts came in in environments where teams averaged 3.89 runs per 9 innings. Over HIS career, Blyleven's starts came in environments where teams averaged 4.33 runs per 9 innings. During the period when both pitchers were active (1970-1979) - and in which Blyleven built up most of his disadvantage in leveraging the support he got compared to Hunter, actually - the averages were 3.94 for Hunter, 4.18 for Blyleven. So no, I wouldn't suggest that Blyleven was pitching in situations where it was THAT much more difficult to make runs stand up.

-- MWE
   327. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 03:08 PM (#3000414)
OK, well, Tangotiger has the average non-ace reliever in a reliever role at a .500 winning percentage, the same pitcher converted to a starter as a .410 winning percentage, and a true replacement pitcher in a reliever role as a .470 winning percentage. I would suspect that relievers promoted to "emergency starters" are long relievers whose leverage index must be about 0.6, so the total "chained" loss in the bullpen is .018, making the composite effect of a replacement pitcher a .392 winning percentage, or 2.4 wins below average per 200 innings (I have been using 2.1, and I now think this is too high, because it doesn't take sufficient account of the gap between an average reliever and a replacement pitcher).

Tracing this over time is very tricky. I certainly think the overall replacement level (i.e., the won-lost record of a team of all replacement playesr) has to be fixed for us to be fair to all eras. So if we're going to be moving the replacement level for pitchers (besides correcting for the slight distortions among starters and relivers caused by changes in bulpen usage), then the replacement level for position players needs to move inversely. At which point we get into the great historical pitching/fielding split question that I just mentioned on my WARP thread. Until I find a way to tackle this, I think we are better off just using one fixed global replacement level for position players as a whole and another for pitchers as a whole. The 3/8 method I use is to track the evolution of the spectrum *among* position players, but as you will see, the sum of the 8 positions always adds up to 12 wins below average in a non-DH league (plus 5 for the pitchers' hitting below average = 17 for position players overall).
   328. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 03:20 PM (#3000425)
For his post-1974 career, Blyleven's WPA was 22.3, and his WPA/LI was 26.0, suggesting that he lost a grand total of 3.7 wins by not "pitching to the score." Go ahead and subtract 3.7 wins from his career total; it should do absolutely nothing to your ranking of him. To put that in context, among starting pitchers (post-74), the leaders in "clutch pitching" are Vida Blue at +9.2, Darryl Kile at +5.8, Al Leiter at +5.6, Jeff Suppan at +5.6, and Ismael Valdés at +5.4, while the trailers are Nolan Ryan at -6.8 ("just a .500 pitcher," as they said), Steve Rogers at -6.5, Ken Forsch at -6.4, Jim Bibby at -6.3, Jeff Fassero at -6.3, Storm Davis at -6.1, and Curt Schilling at -6 (which actually might be significant for his HoM case). Hunter from 1974-79 was just as bad as Blyleven from 1974-1992, posting a -3.6 mark in those six years.
   329. Chris Cobb Posted: October 31, 2008 at 04:33 PM (#3000506)
Thanks for the discussion of my post! I hope the discussion will continue.

Some quick responses:

Mulder & Scully wrote: "I lean to more of the workhorse peak pitcher than the marathoner.
Koosman, again the problem of lack of big years. I just don't like the long, low-prime accumulators."

OK, but why? I have this feeling in our pitcher rankings that comparison of the "low-prime accumulator" to the "peak stud" is universalizing the conclusions of James' classic Drysdale vs. Pappas study. That showed only that for two pitchers with the same career length and the same overall level of quality, the "peak stud" would have more pennant-value for his team. But what about the cases, like those we are facing here, in which the "low-prime accumulator" was, overall, much more effective than the "peak stud" and/or pitched many more innings? I am doubtful that James' conclusion that Drysdale>Pappas means that Walters>Leonard. I might take Walters over Leonard but I think we need to look harder at that conclusion, esp. when it becomes not Leonard but Reuschel or John being compared to Walters.

Dan R wrote: "I think an 85 ERA+ is too high for starting pitcher replacement level, Chris Cobb. . . . Where did you get 85 from?"

I got 85, as I mentioned very briefly in the explanation, by looking at the ERA+ values of starting pitchers who lost their jobs as starters from one season to the next, and it came in pretty consistently around 85, I think. I never had time to do a full study, and I might be misremembering my own conclusions, even, so I don't want to defend that number too strongly, but I will say that it is aiming to model between-season replacement level for starting pitchers as starting pitchers, not to model the in-season replacement level for pitchers as a group, which I would expect to be lower. Anyone who wants to use my numbers, which I'll upload to the yahoogroups space over the weekend, can easily re-set the level where they like. I don't disagree with the numbers you put forward from Tangotiger, though I do think it would be worth discussing whether, given the complexities of pitcher usage, the in-season replacement level for all pitchers is the right replacement level to use.

The only things I claim for the replacement level I have used here are that 1) it gets rid of WARP1's crazily shifting replacement level, 2) it models the fact that starter replacement level has been pretty consistent since the change in pitching distance, and 3) it is a starter replacement level.

Entering a discussion with peak voters about paying attention to career value, I think it is appropriate to show the conclusions about the weight of career value that arise from the application of a high but seriously modeled replacement level to pitchers' careers. We can and should start arguing over what the right replacement level should be, but the claim that matters more to me is that career value above replacement should be getting more credit as a measure of merit from the electorate.
   330. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 04:38 PM (#3000514)
2) it models the fact that starter replacement level has been pretty consistent since the change in pitching distance


Do we know this? I assume it for lack of a better option, but I've never seen it empirically demonstrated.
   331. DL from MN Posted: October 31, 2008 at 04:40 PM (#3000517)
> Tracing this over time is very tricky.

Be careful. Your analysis relied heavily on modern bullpen usage. In previous years the "replacement" pitcher was often another starter on short rest. There weren't always 12 pitchers on a roster. I think you have a good idea for a study to see if you can infer the pitcher-fielder split from the empirical replacement value of pitchers.
   332. DL from MN Posted: October 31, 2008 at 04:46 PM (#3000522)
I agree with separating out replacement value for a starting pitcher and replacement value of a relief pitcher.

Chris - Are you uploading this spreadsheet to the Yahoo group?
   333. Chris Cobb Posted: October 31, 2008 at 04:55 PM (#3000526)
Do we know this?[that starter replacement level has been pretty consistent] I assume it for lack of a better option, but I've never seen it empirically demonstrated.

I am not able to empirically demonstrate this at this time. I should have some data somewhere on it--I started the study last February but then set it aside for the positional ballots--but it is not comprehensive. I saw enough from what I found to decide that it made sense to use it as the working assumption in this study. I also have some data that shows that starter replacement level was higher pre-1893.

I will review the data that I have collected and see what I can put into presentable form.
   334. rawagman Posted: October 31, 2008 at 04:59 PM (#3000530)
2009 Ballot prelim - by the way: when are we holding the actual election?

Use a sort of peak-over career number that measures ink by playing time with a strong preference for players who had good in-season durability. Combined with rate stats and a glove measurement, I feel this gives me both context for what the player actually achieved versus what the league around him was able to do. I think it also may be time to go a little more into my baseball philosophy, which may help in clarifying my rankings. I don't believe in the single stat theory of baseball, meaning I don't use WS or WARP in my rankings. Essentially, I follow this as I think a large percentage of what contributes to baseball is not counted. Well, no one has ever counted them as statistics as far as I've ever heard. This includes things like manager's prerogative, and actions that would require a historical pbp analysis currently unavailable. I search for what I consider "total ballplayers", guys who can do it all. I believe in positional representation and abhor the thought process that says that relievers were all failed starters and 2B are all failed SS, etc... A team cannot win without a 2B, nor without someone in LF. When I look at a player's career, I try to ask myself how I would feel about him as his manager - would his presence require special tactics to protect him, or is he completely reliable. I hope it can be seen by my rankings that the "reliable" players generally rise above the ones with clear holes in their games. I thinnk bjhanke voiced similar preferences vis-a-vis his admiration for "extreme" players. There are always exceptions, but this is what I have. The stats I look at to get here tend to be traditional and rate, both offensive and defensive. Contemporary opinion also helps.

This year's notes. Rickey Henderson is an obvious top spot claimant. Kevin Appier looks to me to be very close to Chuck Finley in (my perception of) value. Appier had slightly better rates and Finely kept it up a wee bit longer. He makes the expanded ballot, but isn't really a threat to get votes from me. I was surprised how well Jay Bell placed. Just missed the expanded ballot. Matt Williams desn't really impress me very much. Close to the expanded ballot, but not quite on it. I might be underestimating him a bit, but I can't see him going higher than mid-70's at best. Mark Grace is a slightly lesser version of Steve Garvey. The Big Cat is maybe a hair ahead of Cecil Cooper in my positional rankings. And Mo Vaughn looks like Kent Hrbek (not literally). None of the other newcomers impressed me very much. PHOM entries are Rickey Henderson, Bret Saberhagen and Dizzy Dean. I am open to arguments that I am underrating Reggie Smith and will keep him in consideration for a bottom-of-ballot vote and/or PHOM consideration ahead of Dean.

1)Rickey Henderson - I have nothing to add to the conversation of what he accomplished or where he stands as a measure of his place among the all-time greats. But a question - realistically, how many more years could he have contributed to a winning MLB team, even in a diminished role? In other words, why Julio Franco and not Rickey? (PHOM)
2)Hugh Duffy (PHOM)
3)Tommy Bridges (PHOM)
4)Ben Taylor (PHOM)
5)Kirby Puckett (PHOM)
6)Lefty Gomez (PHOM)
7)Bus Clarkson (PHOM)
8)Dale Murphy (PHOM)
9)David Cone (PHOM)
10)Vern Stephens (PHOM)
11)Gavvy Cravath (PHOM)
12)Bob Johnson (PHOM)
13)Dick Redding (PHOM)
((13a)Bret Saberhagen)) (PHOM)
14)Tony Oliva (PHOM)
15)Bobby Veach (PHOM)
16)Dizzy Dean (PHOM)
17)Orlando Cepeda (PHOM)
18)Reggie Smith - open to being convinced that he should move up a few spots.
19)Al Oliver
((19a)Andre Dawson))
20)Albert Belle
21)Jack Clark
22)Jim Rice
23)Wally Berger
24)Don Mattingly
25)Dan Quisenberry
26)Lee Smith
27)Bruce Sutter
28)Ernie Lombardi
((28a)Jimmy Wynn))
29)Ron Guidry
30)Al Rosen

Other notables....
39)Kevin Appier - Just ahead of Finley. I prefer the better rate to the longer career, but very, very close.
41)Bucky Walters - Very similar to Pierce in overall picture - but built differently.
45)Tony Perez - I wasn't giving him the extra glove credit he earned through 5 seasons as the Reds' 3B. Still, no peak. As far as 1B go, I have Cepeda up higher because of his very nice peak and his not too short career as a regular. Ben Taylor suffers from a lack of documented stats. The stats there show that he could flat out mash the ball by dead-ball standards. Contemporaries say his glove was the best they had ever seen at 1B. Very apt, in this year of the Hernandez discussion. How much as a scoop worth? I think it's worth alot. I maintain that while a below average defensive 1B can cause little harm, an above average glove at 1B will provide a hefty bonus to the team lucky enough to employ one.
58)Mark Grace - It's always fun when a player's name can fit with his on-field ability/persona. A Graceful first-baseman, with the stick and with the glove. Splitting hairs betwen him and Garvey. I think Garvey stuck out just that much more among his 1B peers.
72)Tommy Leach - With his recent rise in the standings, I took another look at him for our penultimate election. I can see arguments that would have him around or even above someone like Brett Butler, or maybe even a bit more, but that would only mean 20-30 ballot spopts for me, and not significant at this stage. Not being convinced either way, he stays down here. Fine player, but not HOM quality.
87)John McGraw - Hurt alot by my readjustment - no durability. Tsk, tsk.
90)Matt Williams - Definitely hurt by the strike of '94, but hurt more by missing half of the following season. His peak was high, but he was fairly one-dimensional in his offensive game outside of that 1993-96 period.
   335. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 06:09 PM (#3000593)
I maintain that while a below average defensive 1B can cause little harm, an above average glove at 1B will provide a hefty bonus to the team lucky enough to employ one.


Do you have any evidence at all to support this notion?
   336. rawagman Posted: October 31, 2008 at 06:18 PM (#3000608)
Not empirical. That comment needs updating in any case. I base this on the presumption that an "average" first baseman will catch the overwhelming majority of throws that don't cause him to move his feet. With some exceptions, many 1B errors can be attributed to plays that may have been beat out for infield singles anyway. However, I also beleive that an exceptionally sure-handed 1B will have the ability to scoop more off-target throws, saving errors from the rest of the infield.
   337. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 06:35 PM (#3000625)
OK, but by definition, then a less sure-handed 1B will scoop fewer off-target throws, adding errors to the rest of the infield. I don't see why the distribution wouldn't be fairly close to normal.
   338. rawagman Posted: October 31, 2008 at 06:44 PM (#3000632)
The point is that those misplays would be attributed to the thrower, leaving the failure to scoop off his record.
   339. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 06:59 PM (#3000652)
Well, that's a problem with the record-keeping. Clearly, the ideal method would be similar to what the PBP metrics do: you'd have an observer (or even better, a PitchFX-style camera) set up along the first base line, determining the likelihood that each throw will be caught and assigning credit/blame accordingly. But obviously, we can't do that retroactively. Sean Smith has done some work with 1B scooping, but his methodology can't be broken down into seasonal numbers, which is what I'd need for my WARP. I think this is a case where we just have to rely on reputation, as poor a substitute as that is. I just don't see how we can ever really know.
   340. rawagman Posted: October 31, 2008 at 07:14 PM (#3000669)
That's just it, really. I don't think we ever can really, truly know retrospectivly. So contemporary reports suffice for me when they come from reliable sources. Scouting will always have a place in the game.
   341. bjhanke Posted: October 31, 2008 at 11:13 PM (#3000803)
Chris Cobb says, "I got 85, as I mentioned very briefly in the explanation, by looking at the ERA+ values of starting pitchers who lost their jobs as starters from one season to the next, and it came in pretty consistently around 85, I think"

If I understand right, you're trying to compute what I call the "starting line" back in the positionals. That's not the same as replacement rate. It's the point at which the player, although still worth a roster spot, is no longer good enough to be a starter. You and Dan are using the complicated phrase "starting pitcher replacement rate" or some variation. I think that confuses the discussion, and may be why some of the posts about Chris' work seem unsure. I'm going to suggest here that we all adopt the phrase "starting line" to refer to this, or some other phrase that gets the consensus here. But I do think using a phrase that doesn't contain the wording "replacement rate" will help clear the discussions. I, for one, have no problem with thinking that 85 is the starting line, but it's obviously too high for general replacement rate. It also should be lower in early baseball, because the difference between replacement rate and starting line is very small. In the 1870s, it is almost zero: they're the same concept.

I really like rawagman's ballot because it illustrates the problem of a first-time voter. It's not that I have my guys ahead of his. It's that I didn't have the time to look at everyone, and I didn't see a rawagman list early enough to include it in my short list of who to examine. I'll get to people like Hugh Duffy and Vern Stephens, but it won't be by Tuesday. And when I do get to them, it will help to have a ballot that tries to sort them out, even if I disagree with the sort. Thanks! - Brock
   342. Chris Cobb Posted: November 01, 2008 at 01:15 AM (#3000817)
Continuing the post-1892 pitcher survey. Here’s the table for –

Group 2: Pitchers who top Walters in IP, but not in DERA

Pitcher  IP     DERA PRAA RCAP FRAA RAA Rate*  RAR 08 Rank
John    4710.3  4.27  120  
-4   42  158 8.41   535  35
Grimes  4179.3  4.32   84  44   35  163 9.73   497  24
Powell  4389    4.19  151  12  
-40  123 7.02   474    
Kaat    4530.3  4.36   70  29    3  102 5.65   465  71
WALTERS 3104.3  4.14  124  57   19  200 16.12  449   5
Tanana  4188.3  4.30   93  
-1   12  104  6.21  439  60
Hoyt    3762.3  4.21  121 
-17    4  108  7.19  409    
Hershis 3130    4.17  115  24   17  156 12.44  406 101T
Lolich  3638.3  4.19  125 
-10  -19   96  6.62  387    
Uhle    3119.7  4.25   87  61  
-10  138 11.03  387    
D
Mart 3999.7  4.53  -13  -4   31   14  0.85  334    
Hough   3801.3  4.45   21  
-7   14   28  2.00  332  99T
Blue    3343.3  4.33   63 
-17   -1   45  3.38  313 
Hunter  3449.3  4.50    0  23  
-19    4  0.29  280    
Morris  3824    4.56  
-25   0   -4  -29 -1.93  276  94T
*Rate is RAA/250 IP 


Commentary.

This is not as strong a group of candidates, unsurprisingly, as those who top Walters in both IP and DERA, though this group would benefit from a lower replacement level. Without IP or SD adjustments, the only two candidates who really stand out as having a strong case to be better than Walters are John and Grimes. John has the most IP of any unelected post-1892 pitcher, and while his career DERA and RCAP are not extraordinary, his FRAA is among the top on the board (the groundball pitcher effect Mike Emigh noted applying to the pitcher himself?) John has 9.5 career wins above replacement over Walters, which is second behind Reuschel among all pitchers listed so far. Grimes has fewer career RAR than John, but he has much more peak cred, with a number of big seasons and the profile of an innings-eater, which John was not. Without big era-adjustments, it’s hard for me to see how Walters has an edge on Grimes from any angle except possibly consecutive peak.

Kaat and Powell top Walters in career value, but not by much, and I don’t think anyone would say that they have better peaks than Walters, though both have legitimately big years scattered through their careers, and both have better in-season durability than Tommy John.

The pitchers on the lower part of this list are far below Walters, though with a lower replacement level they would be closer. Hough and Morris are two of the weakest players to get a vote in 2008.

And now for the largest table –

Group 3: Pitchers who top Walters in DERA but not in IP

Pitcher  IP     DERA PRAA RCAP FRAA RAA Rate*  RAR 08 Rank
Trout   2725.7  3.78  218  30   17  265 24.31  483  91
Shocker 2681.7  3.73  229  17    1  247 23.07  462  52
Cone    2898.7  3.78  232   3  
-17  218 18.79  450  15
WALTERS 3104.3  4.14  124  57   19  200 16.12  449   5
Bridges 2816.4  3.77  228  
-8   -9  211 18.77  437  21
Breiten 2964.3  3.89  201  
-9   -1  191 16.10  428    
Tannehi 2705.3  4.01  147  50   14  211 19.53  428    
Bender  3017    4.02  161  30   
-6  185 15.32  426  
Adams   2995.3  4.00  166  23  
-10  179 14.97  419    
Rucker  2375.3  3.61  235   0   
-6  229 24.09  419    
Rommel  2556.3  3.86  182 
-10   35  207 20.22  411    
Hahn    2029.3  3.36  257   3  
-18  242 29.82  404    
W
Wood 2684    3.88  185 -15   14  184 17.13  399    
Warneke 2782.3  4.04  142  22    3  167 15.02  390    
Guidry  2392    3.76  197   0    0  197 20.56  388   
Joss    2327    3.77  189 
-13   26  202 21.67  388  67T
Viola   2836.3  3.97  167  
-2   -6  159 14.02  386    
Vaughn  2730    3.90  182  
-3  -12  167 15.29  385    
Appier  2595.3  3.85  187   0  
-12  175 16.90  383  n/e
Dean    1967.3  3.46  227  11  
-16  222 28.25  380  19
Trucks  2682.7  3.87  188 
-20   -7  161 14.98  375  96T
Key     2591.7  3.99  147   0   15  162 15.61  369  
Langsto 2962.3  4.13  122   0    8  130 10.95  367  
Leever  2660.7  4.03  139  
-1    0  138 12.96  351  89T
Shawkey 2937    4.10  131 
-10   -9  112  9.49  346  
Gomez   2503    4.03  131 
-28   -6   97  9.66  297  58
Parnell 1752.7  3.82  132   0    3  135 19.32  276    
*Rate is RAA/250 IP 


This large group has a lot of really good pitchers, but I would call attention to seven as the ones who might be considered serious candidates. At the top are Trout, Shocker, and Cone, who top Walters in RAR despite pitching fewer innings. Using Dan R’s replacement level, Walters would pass Cone but fall just short of passing Shocker. Trout is far in advance, but a good bit of air needs to be let out of his numbers, I think, for two reasons. First, so much of his value is tied up in his gigantic 1944 season, which is surely inflated by wartime competition. He was a good hitter, but most of his RCAP were also accumulated during the war years. He loses more to the war than most because he loses meaningful value on both pitching and hitting (though Walters does, too, it should be remembered). Second, Trout underperformed his expected wins more badly on a per inning basis than any long career pitcher that I know of: WARP1 has him at -18 wins versus expected. By comparison, the much maligned Bert Blyleven is at -23, in 2200 more innings pitched! As Chris J. put it back in the day, looking at wins in relation to run support and expected wins “makes Dizzy Trout look like a great big pile of poo.” The low rank of Trout despite stellar component stats, suggests that the combination of wartime penalties and win-based examinations of his record are already affecting the electorate’s view of him. Shocker’s lack of support is more puzzling, but his candidacy has never gotten much traction. He doesn’t have a blow-them-away peak and he wasn’t a premier innings-eater, but he had six top 10 finishes in both IP and ERA+, and in 1922 he was top 3 in both, which is an outstanding year. He’s a lot like Reuschel, I think, with better rates and less durability. Cone is more of a known quantity, so I’ll say less about him, but these numbers suggest that the electorate is right to have him in the mix of top backloggers, but there’d need to be strong arguments for more difficulty of dominance or for him having lost a lot of value to the 1994-95 strike to put him at the very top of the pitchers’ group. It is worth remembering that during the first half of the 1990s he was a workhorse, placing in the top 10 in IP 1991-95 (though he lacks gray ink for 1992 because he was traded between leagues, but his IP total would have broken the top 10 in either one). My image of him is of the six-inning starter he mostly was after his 1996 arm injury. After the injury, his ERA+ remained top 10 for several years, but the top 10 IP finishes were no longer there.

Tommy Bridges, although he trails Walters in RAR, still does well by this measure. With war credit, he has a good argument to rank ahead of Walters. Many arguments have been made recently about his peak, so I won’t rehearse those here, except to say that it wasn’t shabby, and that he, like Shocker, was a pitcher who never really had a bad year.

In addition to the top 4 on this list who are clustered around Walters, three other pitchers merit notice for their value above average: Noodles Hahn, who is third after Trout and Shocker in RAA with 242, Nap Rucker, who with a solid 229 is well above Walters in this measure, and Dean, whose 222 is next in value after Rucker and who tops the group in rate. These are not pitchers I support, because I weigh peak and career both, but peak voters should certainly give Hahn and Rucker a look. It seems hard to believe that Dean ranks at 19, while they receive no votes at all. Rucker may be overlooked because pitching in front of bad defenses killed his ERA+, so he looks like a seven-year workhorse but not especially lights-out; his DERA tells a different story. Hahn is not quite as impressive relative to his contemporaries as Dean, but he may have been overlooked because he was eligible so early and dropped out of view without being given a fair shot. Surely he is a more deserving candidate than contemporaries Joss or Leever. Joss’s support especially is not borne out by this study. He was a very fine pitcher, but if he is a HoMer, then so is Ron Guidry, to whom he is extraordinarily similar in value.

I should not end without noting that Lefty Gomez does very poorly by this measure. He does have a few great years, but those are offset by many mediocre ones, and his value is significantly diminished by his being one of the worst hitters in this group, and a poor fielder as well. The only pitcher with less career value garnering votes is Jack Morris.

I’ll conclude this long post with the final table, which is the smallest –

Group 4: Pitchers with fewer IP and a higher DERA than Bucky Walters

One might expect that no pitchers with fewer IP and a higher DERA than Walters, whose value is supplemented significantly by hitting and fielding and whose merit lies more in his peak than his career, would actually merit serious consideration, but, surprisingly, that is not the case.

Pitcher  IP     DERA PRAA RCAP FRAA RAA Rate*  RAR 08 Rank
WALTERS 3104.3  4.14  124  57   19  200 16.12  449   5
Mays    3021.3  4.25   84  61   57  202 16.71  444  39
Gooden  2800.7  3.97  165  19    2  186 16.60  410  72
Newcomb 2154.7  4.14   86  74   
-5  155 18.01  328  23
Phillip 2607    4.18   93   1    0   94  8.98  302    
Sain    2125.7  4.39   26  35    1   62  7.29  232
*Rate=RAA/250 IP 


No pitcher here tops Walters in RAA, but Carl Mays is very close, with a remarkably similar profile to Walters, overall, though in context his in-season durability does not compare to Walters’. Mays’ case isn’t a bad one, however; he certainly deserves to be in the mix. The pitcher here who might actually have more career value than Walters is Don Newcombe, who probably deserves four full years of credit, two for minor-league play in 1947 and 1948, and two for military service. If these seasons would add 1000 innings at Newcombe’s career average rate, he would have about 480 RAR.

Gooden does not do badly here, but unless one values his stratospheric 1985 season hugely indeed, he doesn’t push in among the top candidates. Sain, even with war credit, falls well short, as does Phillippe.

For those who want to see all these numbers gathered together and work with them yourselves, I’ll post the spreadsheet to the yahoogroups site tomorrow for retrieval.
A basic integration of the lists produces this list as the top pitchers by career runs above replacement.

1) Reuschel 587
2) John 535
3) Quinn 528
4) Tiant 523
5) Willis 502
6) Grimes 497
7) Leonard 494
8) Cicotte 487
9) Trout 483
10) Powell 474
11) Kaat 465
12) Shocker 462
13) Koosman 460
14) Cooper 454
15) Cone 450
16) Walters 449
17) Mays 444
18) Tanana 439
19) Walters 437
20) Breitenstein 428
20) Tannehill 428
22) Bender 426
23) Luque 421
24) Adams 419
24) Rucker 419

It's not the order I would rank them in for my ballot, but it's a rank-order that will carry weight in how I sort the pitchers.
   343. Howie Menckel Posted: November 01, 2008 at 02:12 AM (#3000827)
Brock,
You'll do fine.

There's a 'perfectionist' element to this exercise that has always bedeviled a crop of potential and occasional voters - 'what if I'm missing someone?' basically.

Thing is, those of us who have voted every year aren't any smarter and often are less smart. It's just a matter of having the serenity to accept the things we cannot know, the effort to find out what we can, and the wisdom to accept the totality of what we do know.

The results have been rewarding and useful, while we acknowledge that they can never be perfect.
   344. Paul Wendt Posted: November 01, 2008 at 05:08 AM (#3000877)
Regarding the replacement level for pitchers (all pitchers by league-season?) implicit in DT cards, presumably utilized in calculation of PRAR or DERA.
324. David Concepcion de la Desviacion Estandar (Dan R) Posted: October 31, 2008 at 10:24 AM (#3000376)
It seems to vary wildly. For Nolan Ryan, it's 5.6, while for Cy Young, it's 3.3 (these are seasonally adjusted numbers, not the all time adjusted ones). I imagine this has something to do with how BP handles the pitching/fielding split.

Anyways, this only teaches us what we already knew: that PRAR is the busted sister of the already beaten-with-an-ugly-stick PRAA. Fortunately, moving from DERA (which is definitely reliable, as long as you trust BP's quite questionable defensive adjustments) to legitimate pitching wins above replacement is delightfully simple arithmetic.


Do we know that Davenport estimates the effect of team fielding on the DERA scale, which is runs per nine innings?

Probably you are right about this calculation of PRAA from DERA. Probably I missed something in the glossary, which I read sometime during my gathering data on 400-odd pitchers. (I find that the glossary and other documentation is sometimes difficult to use; reading it doesn't provide me what I "expected" --not that I still expect it in 2008.)

Reading what I found left me uncertain, and agnostic as a matter of policy. I half-guess that the fielding effect is estimated on the RAA scale, which is denominated in runs (per team season, I suppose, but we call that a counting statistic). Then subtract those runs attributed to team fielding from RAA; the difference is PRAA. Use translated innings pitched XIP to calculate DERA.
   345. Bleed the Freak Posted: November 01, 2008 at 12:08 PM (#3000906)
342. Chris Cobb Posted: October 31, 2008 at 09:15 PM (#3000817)
For those who want to see all these numbers gathered together and work with them yourselves, I’ll post the spreadsheet to the yahoogroups site tomorrow for retrieval.
A basic integration of the lists produces this list as the top pitchers by career runs above replacement.

1) Reuschel 587
2) John 535
3) Quinn 528
4) Tiant 523
5) Willis 502
6) Grimes 497
7) Leonard 494
8) Cicotte 487
9) Trout 483
10) Powell 474
11) Kaat 465
12) Shocker 462
13) Koosman 460
14) Cooper 454
15) Cone 450
16) Walters 449
17) Mays 444
18) Tanana 439
19) Walters 437
20) Breitenstein 428
20) Tannehill 428
22) Bender 426
23) Luque 421
24) Adams 419
24) Rucker 419


Additional Peaksters:
Gooden - 410
Hahn - 404

Pitchers who are not directly comparable by using Chris Cobb's measures mentioned above, however, they merit consideration by the electorate:
Jim McCormick
Dick Redding
Leroy Matlock
Hilton Smith
Bruce Sutter
Lee Smith
   346. Paul Wendt Posted: November 01, 2008 at 03:11 PM (#3000937)
"Weighing" career pitchers using those data I have not seriously worked with a "point of reference" or "zero point" so low as 85 on the ERA+ scale. I have lovingly perused tables using 91, in effect, where the measure incorporates batting and pitching performance. The pitching component is essentially DERA so it is no surprise that the same pitchers generally rank high, broadly.

The chief beneficiaries in my personal picture of the historical ballplayers terrain, so to speak, are Shocker, Leonard, Trout, and Tiant --during the Fall 2008 stage when my use of DERA has been systematic with broad although not comprehensive scope-- plus at least Quinn, Cooper, Reuschel and Cone over a longer period.
(The others high on Chris Cobb's concluding list are of course those with greatest longevity, who benefit most from using 85 rather than 91 as a point of reference: Powell, Willis, Grimes, Kaat, and John.)

I do not have any estimate of pitcher fielding performance and I have my own jury-rig based on OPS+ for batting, so I will be interested to download Chris Cobb's dataset and get what he has compiled or jury-rigged, too.

--
I don't like Virgil Trucks, and I wonder how he seems close to #61 in Bill James eyes. But at the same time I am surprised by the poor showing of Virgil Trucks in Chris Cobb's series. It must be my private setting, with two breaks in his favor: I have him merely weak as a batter rather than woefully weak, and I do not have the debit for his own poor fielding.

Ned Garver? Should be in the consideration set here?
I always think of him and Virgil Trucks together, thanks to Virgil Ned Garvin.
   347. Chris Cobb Posted: November 01, 2008 at 03:24 PM (#3000944)
I half-guess that the fielding effect is estimated on the RAA scale, which is denominated in runs (per team season, I suppose, but we call that a counting statistic). Then subtract those runs attributed to team fielding from RAA; the difference is PRAA. Use translated innings pitched XIP to calculate DERA.

At the team level, XIP = IP in WARP1, so if one is concerned that XIP is distorting DERA, it should not do so at the team level. The difference between RAA and PRAA seems to be shared out among pitchers based on XIP.

Of greater concern in understanding WARP1 and assessing its reliability is that Team RAA, which is the difference between NRA*IP and 4.5*IP, does not equal FRAA+PRAA.

For example, the 1998 New York Yankees have a 3.86 NRA in 1456.7 IP. Their team RAA, therefore, should be 104. (It's listed as 107, but that's probably due to adding up rounded numbers.) Team PRAA is 35, so RAA-PRAA=72. 72, it would seem to me, should equal team FRAA. It does not, however. Team FRAA is 110. So it's not at all clear to me what happens to the 38 run difference between team FRAA and Team RAA-PRAA.

For the 1904 Cincinnati Reds, Team RAA-PRAA = 24. Team FRAA = 73.

I don't understand this accounting. I wouldn't expect runs above replacement to balance, because replacement levels might be different at different positions, but surely runs above average should add up?

For the record, my garden-variety calculation of team fielding value, which compares team defensive efficiency (with rough park adjustment) to league defensive efficiency to determine non-HR hits saved by the team, then multiplies hits saved by the average run-value of a non-HR hit for that season, sees the 1998 Yankees fielders as saving 67 runs above average. This is in the ballpark with the 72 normalized runs that separate team RAA from team PRAA.

For the 1904 Reds, my calculation sees team fielding as 55 runs saved above average. This is not in the ballpark with the 24 normalized runs that separate team RAA from team PRAA.

The upshot of this quick study is that I trust the RAA-PRAA fielding runs saved value (which gives us DERA) more than I trust team FRAA, and I trust DERA less than I did.
   348. Chris Fluit Posted: November 01, 2008 at 06:36 PM (#3001013)
Including Walters, that’s 60 pitchers. To my knowledge, no one with anything like a serious HoM case has been omitted.


Comments like this inspire me to want to take part in a HallofVeryGood project once the positional elections are done. There are currently 63 pitchers in the Hall of Merit. I think it would be interesting to see how many of the pitchers on Chris Cobb's list would make the second level. And it wouldn't be all of them. The HoM 63 includes three groups that Chris C. left out of his evaluation- 6 pre-1892 pitchers, 8 Negro Leaguers and 4 relievers. If the second level has a similar balance as the top level, that's another 18 spots. Which means we'd be looking at something like 45-50 of the pitchers on Chris C.'s list getting into the HoVG.

We've argued about the same guys a lot. I think it would be fun to argue Hunter vs. Blue, or Morris vs. Hough. I think it would be fun to see if Teddy Ballgame's teddy bear Mel Harder even makes the second level. I wonder if any Hall of Famers would have a hard time even making the Hall of Very Good- guys like Dean, Gomez and Happy Jack Chesbro.

Even the categories that Chris Cobb left out could be fun:

pre-1892 HoM: Caruthers, Clarkson, Galvin, Keefe, Radbourn, Spalding
would Mickey Welch even make the HoVG? what about Candy Cummings? Or would they be beaten out by the likes of Tommy Bond, Will White, Tony Mullane, Jim McCormick, Bobby Mathews, Dave Foutz and Silver King?

NeL HoM: Brown, Dihigo, Foster, Foster, Mendez, Paige, Rogan and Williams
would we induct one of the 19th century pitchers in Stovey or Fowler? Once again, would any Hall of Famers miss the Hall of Very Good? How deep do we go into the pool of Redding, H. Smith, Matlock, Day, Donaldson, Byrd, Cooper, Manning and Brewer?

relievers: Eckersley, Fingers, Gossage, Wilhelm
would we grab one of the early firemen like Marshall or Hiller? are we impressed with HoF inductees and candidates Sutter and Smith? What are the chances of a dark horse like Quisenberry or Henke?
   349. Chris Cobb Posted: November 01, 2008 at 07:52 PM (#3001045)
The spreadsheet with the RAR data from which the tables I have posted were drawn is now available in the Hall of Merit yahoogroups space.
   350. Paul Wendt Posted: November 01, 2008 at 08:15 PM (#3001055)
Max Lanier: Not sure why he is on this list. Paul?
Ellis Kinder: If we had better tools for assessing pre-1960 relievers, we could tell for sure whether he ought to be in the consideration set or not.
Sal Maglie: Probably should be included; I'd guess he was better than Johnny Sain. Estimating MxL credit for pitchers is tough, though.


I think of them as a group, 1940s pitchers whose full careers are probably now impossible to assess. Lanier was a veteran star and Maglie a 28-yr-old veteran of 84 mlb innings (and some WWII) when they went to Mexico. Perhaps was the only star, or the closest to a star, who went to Mexico?
Did they apply for reinstatement, supplicants to Happy Chandler? Lanier was readmitted in 1949, Maglie in 1950.

WWII? None of Lanier, Maglie, and Kinder served long in WWII.

Kinder debuted in the majors just shy of 32 years old in 1946.
I recall hearing that some of the Red Sox pitchers Mel Parnell, Tex Hughson, and Kinder drank too much for the good of their pro baseball careers, but I doubt that would be a reason for very late mlb debut. It could be a reason for experimental move to the bullpen (Kinder only), if destructively in phase with the pitching rotation.


ing but Lanier and Maglie played in the Mexican League and Kinder debuted
   351. Paul Wendt Posted: November 01, 2008 at 08:22 PM (#3001059)
Chris F #348
Comments like this inspire me to want to take part in a HallofVeryGood project once the positional elections are done. There are currently 63 pitchers in the Hall of Merit. I think it would be interesting to see how many of the pitchers on Chris Cobb's list would make the second level. And it wouldn't be all of them. The HoM 63 includes three groups that Chris C. left out of his evaluation- 6 pre-1892 pitchers, 8 Negro Leaguers and 4 relievers. If the second level has a similar balance as the top level, that's another 18 spots. Which means we'd be looking at something like 45-50 of the pitchers on Chris C.'s list getting into the HoVG.

There is a lot of crowding in the ranks. Are there only 250 Very Good players without Merit? Literally selecting a "second 250": I understand the appeal of having the list but I suspect it would be tedious to do the selection. Well, tedious to carry it out with HOM focus of attention, so that in practice it would not support so much attention.
   352. bjhanke Posted: November 02, 2008 at 08:33 AM (#3001247)
Paul says, "There is a lot of crowding in the ranks. Are there only 250 Very Good players without Merit? Literally selecting a "second 250": I understand the appeal of having the list but I suspect it would be tedious to do the selection. Well, tedious to carry it out with HOM focus of attention, so that in practice it would not support so much attention."

Paul is right here. I have experience. This is what killed the Baseball Maniacs project. It is MUCH harder to identify, say, first basemen #21-40 than it is to identify #1-20. If you try to do a Hall of Very Good by taking the differences between #1s and #20s and using them to give you approximate numbers to put in the HoVG, you end up with LOTS more players, because you're in the very belly of the achievement distribution, rather than dealing with the little spike on top. If you just put the same number in the HoVG as are in the HoM, then the differences between the tops and bottoms of the HoVG positions are very small.

I should add that the Baseball Maniacs were trying to do this 30 years ago, which means without mathematical ranking methods, and using only encyclopedias for source material. But if you do use the mathematical methods, you're entering the territory where anything that is NOT accounted for by the method can really warp the rankings.

Here's one of my favorite analytical horrors: How much WWII credit do you give Warren Spahn? I say none, and I wouldn't argue with someone who wanted to DEDUCT WWII credit form Spahn. Why? Well, let's look. Spahn entered the bigs in 1942, 21 years old, in a cup of coffee that was just about good enough to get him invited back to spring training. Then he went into the war credit zone for three years, returning to baseball in 1946. He was obviously and immediately ready and proceeded to have a monstrously long career, leading the league in wins 8 times and, I believe, holding the record for most wins ever by a lefty.

What's the most important piece of info in the above paragraph? He was 21 when he left for the war. As you all no doubt know, one of the biggest drains on arm durability is workload when young. Well, Warren got three years off when he was young, although he was certainly a competent big league pitcher by no later than 1943. Did this hurt his career length? How could it? I argue that it HELPED his durability, and is one of the main reasons he holds his record. He took three years off when young, his arm got really settled in, and it became all but bulletproof. That's not war credit; it's war debit, if anything.

No mathematical method that I know of is going to deal with that. If it deals with war credit at all, it's going to give Warren two or three years, rather than the none that I argue he deserves. The point extends beyond Warren Spahn, which is good, because he's above the HoVG anyway. The point is that no math method can account for everything. They can get close, but really can't be perfect. Well, as you descend deeper and deeper into the achievement pile, small differences become worth more ranking places. Any small problems with math methods become more important. And that leaves you with doing much more of the type of analysis that I just did for Spahn, as you try to make sure your method doesn't betray you.

And who are you analyzing? Players who are very very similar to each other, no matter how you cut it. I'm willing to go through 200 players to work up a decent HoM ballot, given enough time. I'm not willing to go through 500 to get a 250-man HoVG, especially when the difference between the best and worst of the 500 is maybe half of the difference between the top and bottom of the 200. Before you start on this, think about it. It's a HUGE amount of work to dig this deep into the pile, for smaller and smaller diminishing returns. I say don't do it unless you really REALLY just have to, in order to think you've done what you need to do. I've seen the effort essentially collapse one group of analysts. I like this one. I'm not saying no, but I am saying really think first. Don't just think it will be as easy as the HoM is. It won't. I tried it.

- Brock
   353. Howie Menckel Posted: November 02, 2008 at 03:27 PM (#3001286)
- I'll pass on a "Hall of Very Good," thanks. I just want to vote on pitcher rankings.

- It's one thing to pretty much not give Spahn war credit. But deducting? That's a bit much, too cute by half, I'd say. Next we'll be deducting from Lefty Grove because he got to 'coast' with his arm by being stuck in the minors at a young age, and so on. And don't even get us started on Dazzy Vance, then..
   354. Chris Cobb Posted: November 03, 2008 at 01:56 AM (#3001419)
Time for a preliminary ballot!

2009 Preliminary Ballot

My 107th ballot, and the second cast in real time.

Of the new eligibles, Henderson is a slam dunk, no doubt HoMer, an all-time great, the best leadoff hitter of all time. So he’s easy to place ?. The best of the rest are Appier, Bell, and Grace. None of them are HoMers, though Appier is close enough that he’s worth discussing: I imagine he’ll get some votes. Bell and Grace were good to very good players who look unlikely to place into my top 100 eligibles.

Following the model of my rightfielders ballot, I have now shifted over to basing my rankings fully on Dan R’s WAR for post-1892 position players. That’s both easier and more reliable than mucking about with lots of patches to WS and WARP1 to straighten out replacement-level issues. I still assess career and peak in the same manner, adding career value above replacement, career value above average, and 5*peak rate. For starting pitchers, I use the same calculation, using career RAR, career RAA, and 5*peak rate/250 IP. For ranking pitchers against position players, I have gone with Dan R’s suggested replacement level of 80 ERA+; that is plausible, and it seems to put the pitching candidates and positional candidates on a reasonably level playing field.

(#) = Last year’s ranking
Total = result of new system

The preliminary ballot simply follows the system's results--I will make subjective adjustments for the final ballot.

1. Rickey Henderson (n/e). Total = 230.6. An inner-circle Hall-of-Meriter; #3 left-fielder after Williams and Musial (will be #4, of course, after Bonds becomes eligible). My in-out line is about 120 points, so Rickey is pretty close to having two Hall-of-Merit careers, just as Bill James has argued.
2. Rick Reuschel (4). Total = 140.7. The best combination of peak and career among eligible pitchers, and the strongest candidate in the backlog.
3. John McGraw (n/r). Total = 131.4. Long examination of Dan R’s WAR has convinced me that his assessment of McGraw is correct. Despite his brief career, the value is there.
4. Bert Campaneris (60). Total = 129.2. First of five shortstops on my preliminary ballot. They may move around a bit before the final ballot. I don’t think we’ve yet honored all the deserving players from the 1960s and 1970s. Shortstops Campaneris and Concepcion are the two of the most significant omissions among position players.
5. Luis Tiant (12). Total = 128.7. A nice combination of prime effectiveness and career durability. Rick Reuschel lite, except with respect to style!
6. Phil Rizzuto (27). Total = 127.8
7. Reggie Smith (6) Total = 126.0. Never had dominating seasons because his batting and fielding peaks are separate and he was out of the lineup a lot, but he never had a bad year, either. We’re short on both 1970s shortstops and 1970s outfielders, even though we are long on both positions overall, so Smith should go in with Concepcion and Campaneris. This ranking does not include Japan credit; if I decide to give that, Smith could move as high as #3 on this ballot.
8. Dave Concepcion (5). Total = 125.8 Deserves a head-to-head comparison with Campaneris and Rizzuto for top shortstop spot on the ballot.
9. Tommy John (7). Total = 124.4. He doesn’t have a great peak, but his 12-14 year prime is about as good as any eligible pitcher’s, and he adds another 4-6 good years on top of that, which is quite extraordinary.
10. Johnny Pesky (25). Total = 123.8 WAR highlights the strength of his peak. The presence of McGraw, Pesky, and John on my ballot should show that I am not favoring either peak or career at the expense of the other!
11. Dave Bancroft (11). Total = 122.1. The best combination of bat and glove among eligible shortstops, but in an easier era. If he could have stayed in the lineup more, we’d have elected him long ago, as he was a slightly better ballplayer than Sewell with a longer career. But having few seasons of 145+ games hurts him.
12. Urban Shocker (9). Steadily excellent in a tragically shortened career.
13. Tommy Leach (19). Total = 121.5. Fine player for a long time. Andrew Siegel’s brief analysis of his case is excellent, and Dan R’s war2 shows that his play in relatively high SD leagues was still very valuable, as does Joe Dimino’s PA analysis using Dan’s numbers.
14. Buddy Bell (3). Total = 120.86. Very similar to the recently and deservedly elected Nettles, though not as good as I thought when I was overadjusting for the DH. Will almost certainly drop behind Cone when I get missed time credit calculated for pitchers. He and Leach are both in danger of being bumped for the ballot as I finish consideration of a few complicated cases, but both would be deserving HoMers in the unlikely event they are elected this year.
15.David Cone (10). Total = 120.82. Slips a little bit in my new pitcher system, but not much. Excellent peak, and just enough career added by several effective seasons after coming back from arm injuries.

Players still under consideration for ballot spots, all of whom need to be assigned credit for non-ML play, for military service, or for ML play prior to 1893: Tommy Bridges, Bus Clarkson, Gavvy Cravath, Hugh Duffy, Mike Griffin, Burleigh Grimes, Herman Long, Leroy Matlock, Don Newcombe, Jimmy Ryan, Dick Redding. It’s also possible that another pitcher could bounce up—I haven’t finished my calculations there, but I’ve done all that I think truly likely. Rucker and Hahn probably have the best shots. It might be that none of these players will make my ballot, but it’s possible that 2 or 3 will. Many will end up in the 16-35 range, I think.

The next 20 among players who have been fully evaluated in the new system are as follows: (this listing excludes players with significant pre-1893 playing time, Negro-Leaguers, and some pitchers – that’s why this is a preliminary ballot!)

16) Johnson, Bob 120.2
17) Grimes, Burleigh 118.9
18) Harrah, Toby 118.73
19) Cey, Ron 118.68
20) Cash, Norm 118.4
21) Stephens, Vern 118.3
22) Fregosi, Jim 118.2
23) Bonds, Bobby 117.65
24) Trout, Dizzy 117.59
25) Klein, Chuck 117.2
26) Cuyler, Kiki 116.8
27) Walters, Bucky 114.5
28) Butler, Brett 114.0
29) Guidry, Ron 113.4
30) Fernandez, Tony 113.3
31) Bridges, Tommy 112.7
32) Willis, Vic 112.5
33) Lemon, Chet 111.4
34) Veach, Bobby 110.8
35) White, Roy 110.0
   355. Howie Menckel Posted: November 03, 2008 at 02:06 AM (#3001423)
So far I'm inclined to replicate my 2008 ballot after Rickey No 1 replacing Raines (I had Saberhagen 5 so others move up and I have a slot open. I didn't vote for Lundy...)

Sell me on your favorite missing player
:)

1. RICKEY

2. DAVID CONE - Extremely similar to HOMer Stieb. 8 major prime seasons to Saberhagen's 5. I suspect people are underrating Cone's remarkable 1994, giving not enough credit on a strike-ruined season. 175-96 from 1988-99. Even won all 5 of his World Series starts, with a 2.12 ERA.
3. CANNONBALL DICK REDDING - A longtime favorite who climbed his way back onto my ballot in recent years and now finally has climbed back to "elect-me." I liked him as an all-around candidate, but the HOF research suggests he's more of a peak guy. Those types don't always fare well with me, but I see no better player on the board.

4. REGGIE SMITH - Completes a 10-year renaissance from off-ballot to my top modern full-time OF. Not certain he's a HOMer, either, but solid in so many respects. Weird patterns of high-skill vs high-PA, but he's just damn good every year, basically. Played a lot of CF, and quietly was one of the better OFs of his era. Closer to Puckett defensively than I had thought, and finally he wins this battle.
5. BOB JOHNSON - I really like this sort of consistency over an endless span, though I'd hardly say he's a 'must-elect.' Sort of the Joe Gordon of OFs in career shape, or a slightly longer and flatter version of Kiner. I am concerned by 1944 being his highest OPS+; seems like he took advantage of the weak competition, so I discount that a bit. But has more than a decade's worth of excellent hitting, for a prime that I like better than Van Haltren's or almost any other holdover's.
6. ALBERT BELLE - Eerily Kiner-esque and Keller-esque, and I like if not love these mashers. Wouldacoudashoulda been such an easy pick if not for the sudden career crash. It is true that in subsequent years even more of these types have proliferated.
7. BOB ELLIOTT - Good to see him mentioned in a discussion thread 6-7 years back, at least. Six seasons of at least 134 OPS+, ALL of them as a 3B. Wish he'd played all 3B and not much OF, but c'est le vie - Sewell seemed to get treated as a full SS by some. Beats out HOMer Boyer (see Boyer thread for details) and compares remarkably well with HOMer Santo as a hitter (see Santo thread for more details). Better than HOMer Hack as well, and better than HOMer DaEvans (see DaEvans thread).
8. BEN TAYLOR - Had meant to reconsider him for years; finally did so 4 yrs ago. Long career, excellent fielder, consistent player. I'm not 100 pct sold on the hitting MLEs, but very good reputation and for sure a quality player. Moves up slightly.
9. VIC WILLIS - Won a Howie SP bakeoff with Grimes and Walters, with slightly more career than Walters and better peak than Grimes. It's close, but I'll stick with Willis for yet another year.
10. DAVE CONCEPCION - 8th time on my ballot. Peak is as good or better than Fox's; not quite as consistent, but a slick fielder and a very useful offensive weapon many times. Not fully buying the "other teams were stupid enough to play ciphers at the position" argument; that helped the Reds win pennants, but Concepcion can't get full credit for that stupidity. But he needs the modest credit in that regard to outlast Rizzuto. Similar case to Bancroft, whose prime I preferred in other years to Concepcion's length. It's close.
11. KEN SINGLETON - Bob Johnson-like, but not quite as good for quite as long. Equally underappreciated in his time.
12. KIRBY PUCKETT - Good prime for a CF, but not amazing. I had said if I wasn't sold on him being an excellent defensive CF in his first 6-8 years, he'd drop a bit. And now I believe his defensive prime didn't last that long, so he has dropped slowly but surely. I'll pass on the intangibles, but he holds his own against BobBonds offensively, for instance, and has just enough at the finish line to pass DaMurphy on long prime.
13. BUCKY WALTERS - 4th pitcher on my ballot. Seemed to get Jim Palmer-like defensive support, without enough super-stats to make that irrelevant. Proved his mettle outside of 'war years.' Lemon-esque, though I wasn't a big fan there.
14. DALE MURPHY - His modest fan club will rejoice that I've seen the light, at least in terms of a 14th-place vote. A different peak-primieness than Belle, and a different fade as well.


AMONG THE TOP 10 RETURNEES, BUT I'M NOT VOTING FOR THEM (YET)
JOHN MCGRAW - Wow, never thought I'd see him on THIS list again. Actually was too underrated collectively for a while. Incredible .466 OBP, but basically in a 7-8 year career of semi-regular play, and relatively mediocre in slugging. Weird career that I think befuddles some of the group's strict systems of voting. But not a terrible use of a ballot slot by any means.
TONY PEREZ - Faked it for 5 years at 3B, mostly a 1B and a lotta stat-padding on the downside. Doesn't have the peak I want in guys like this, nor the sufficient OPS+ toward the end.
GAVVY CRAVATH - Have voted for him before; do give him some minor league credit, absolutely. A reasonable pick; I just think that not only did he get a huge boost from the Baker Bowl, others could have done the same. Anyone else who has THIS much more MLB production in his 30s than his 20s? Yes, but not many.
PHIL RIZZUTO - Wow, the required "2nd look" nearly put him on my ballot. I'll grant a lot of war credit and stipulate to the great fielding. But even 3 war credit years gets him only to 13 main years, and the fielding made him above-average overall but not excellent in most seasons. Yet at closer look, similar case to Concepcion when you cancel out the irrelevant parts.

JUST MISSED
ORLANDO CEPEDA - Suddenly popped up on my ballot 10 years ago with the reevaluation. Had been losing out to Perez with positional consideration, but closer look shows a sterling top-4 and top-10 offensive line. DH opportunity added nothing to his case.
DAVE BANCROFT - Not sure if I ever voted for him before 10 years ago. But look at the prime: fantastic fielder at SS, with OPS+s of 120-19-19-09-09-09-04. Won a fresh 3-way evaluation vs Fox and Concepcion at one point, then fell to Davey. Similar to Randolph, but an SS.
BURLEIGH GRIMES - Compare to Ruffing, Rixey, Wynn and other such HOM pitchers - ok, Sutton, too. I dismissed him as short of Rixey and Ruffing, and he was. But he's just one 130 ERA+ year short of climbing onto this ballot. Better peak than Tommy John, and a lot more durable relative to his era.
LUIS TIANT - Looks like he has the peak at first glance, but notice that the IP just aren't quite there. Plenty good when he did pitch, but with that lack of innings you have to be even more dominant. Maybe he winds up as the era's last P electee, but probably not.
RON CEY - In the past I have had him over Nettles and Bell and nearly on the ballot, but that's because I may like his fielding better than most. Closest of the trio to Bando in hitting. Talk me into voting for him.
LEE SMITH - Very tough one. 10 seasons I really like a lot, only 1 or 2 I love. Sutter has more to love, less to like. A lot of RPs do. Off my ballot, but may get back into consideration.
   356. Juan V Posted: November 03, 2008 at 03:18 AM (#3001442)
You're missing your 15th guy.
   357. bjhanke Posted: November 03, 2008 at 03:51 AM (#3001450)
Well, here's my final ballot for the 2009 (? Is the year right?) HoM list. I gave up on tweaking and obsessing, because it's to no real purpose, given the deadline. Right now, this is the best I can do.

- Brock

1. Rickey Henderson

2. Reggie Smith
I now think that there is a big gap between Reggie and whoever should be #3. Maybe there's some sort of quality of competition issue or something, but if not, I think the HoM missed one here.

3. Babe Adams
It's the combination of career length and rate, mixed in with the wonderful control. If I were just a peak guy or just a career guy, Babe would not rank as high. But I try to look at the balance between the two, and that's where Adams shines. He has both. So does Wilbur Cooper. The difference is the Series. Cooper probably should be higher than #16, but I don't have time to work out where to place him, so he'll have to wait for next year.

4. John McGraw
If the next three guys didn't have short careers, I would have to move John down because of his. By next year, I will have a much better idea of who goes where.

5. Bobby Bonds
Bobby, John McGraw and Babe Adams have such different credentials that I have no idea whether I have them in the right order. I do think they're all ahead of Don Newcombe, but that's the best I can do.

6. Don Newcombe
I moved Don down because the career really is short. He remains high because of the bat.

7. Dizzy Dean
I moved Dizzy down because the career really is short, too.

8. Luis Tiant
Very similar rate and career length to Wilbur Cooper. High length, good rate, but not great. But unlike Cooper, he wasn't in the middle of a bunch of guys who did the same sort of thing. So he stands out more.

9. Deacon Phillippe
I do tend to compare Deacon and Sam Leever to Maddux and Glavine. I think that, if the LP pair were to be somehow transported to current times, they would have longer careers with the same sort of wonderful control and about the same value as MG. The reason I think that their lengths would improve is that they would have more choices of breaking ball. As far as I can figure out, Maddux and Glavine don't use the curve itself a lot. They use cut fastballs, sliders and assorted changes. That reduces wear and tear on the arm compared to throwing curves.

BTW, just in case anyone is wondering, the reason I am so sure of the "born with control" effect is that I have some of that myself. I'm no professional athlete, but I've always had accuracy, even with things like guns and bows that I seldom handle. At my level of competition, which is university intramurals and not one level higher, I have great control over pitches. I can't throw hard enough to do anything with it, but I can throw a softball over the low outside corner of the plate, for called strikes if you don't swing, pitch after pitch after pitch. Phillippe and Leever were professional pitchers with at least my level of accuracy. Very likely much more. And they practiced baseball a lot more than I ever did.

The point is that control is at least part talent, and at the Deacon's level of control, he has to have maxed that talent out. Plus throwing hard if not really hard. Plus practice. He has to have everything. He has the record for control. What he didn't have was a slider or a cut fastball, because they didn't exist. Move him to the 1990s and he has them and the control, too. That's Greg Maddux or maybe even better.

10. Sam Leever
A ridiculous oversight on my second prelim. I meant to move Deacon Phillippe into Sam's spot, but I didn't mean to just discard Leever. There isn't enough difference between the two pitchers to justify anything like that. In fact, I don't see enough room between the two to place anyone else in between them.

11. Hilton Smith
Of the remaining Negro League players, Hilton has the best reputation that I know of. His MLEs and rep suggest that this is about right.

12. Tommy Bond
I ran a sort by IP of all ML pitchers before 1890. Tommy Bond is 7th, behind four guys whose careers start about 1880 instead of 1875 (Bond starts in 1874), Pud Galvin, and someone named Bobby Mathews, whom I have never heard of. Bobby Mathews wasn't bad. His ERA+ is 107, the same as Galvin's, only 4 points behind Bond. His career is quite a bit longer than Tommy's, although not as long as Pud's. He looks like he might even be a HoM candidate, at first look. Does anyone know anything about him? I'll certainly check him out before next year.

13. Tommy Leach
Career length and defense.

14. Rabbit Maranville
Even more career length and defense than Tommy Leach, but an even weaker bat.

15. Lou Brock
Without World Series credit, would not be on this list. He has career length to sell, but not defense.
   358. DL from MN Posted: November 03, 2008 at 03:55 PM (#3001612)
> This ranking does not include Japan credit; if I decide to give that, Smith could move as high
> as #3 on this ballot.

Probably heard this before but I believe Smith deserves at least 1 year of credit. He clearly could have had a job in MLB but chose to go to Japan because they offered the best money.
   359. DL from MN Posted: November 03, 2008 at 04:01 PM (#3001614)
I believe the second "Walters" in Chris' list is actually Tommy Bridges.
   360. Chris Cobb Posted: November 03, 2008 at 04:20 PM (#3001632)
I believe the second "Walters" in Chris' list is actually Tommy Bridges.

That is correct. The list is at the end of the long post, #342. #19 says "Walters," but it should say "Bridges."
   361. Al Peterson Posted: November 03, 2008 at 09:49 PM (#3002082)
Why I can’t vote for Rick Reuschel – one voter’s opinion.

I’ve never had a feeling that he was a big game pitcher. I’m going to go through the times Reuschel’s teams were in a pennant race. Let’s see if there is a pattern. Numbers are from bb-ref.com, I think I've pulled them properly from gamelogs/splits:

1973: on Aug 1st Cubs are in 2nd place, 2 GB

Pitching Line April – July: 11-8 2.54 ERA
Pitching Line Aug: 1-5 3.43 ERA
Pitching Line Sep/Oct: 2-2 5.26 ERA

1977: on Aug 1st Cubs are in 1st place, 2 up

Pitching Line April – July: 15-3 2.28 ERA
Pitching Line Aug: 4-2 3.83 ERA
Pitching Line Sep/Oct: 1-5 3.51 ERA

1978: on Aug 1st Cubs are in 2nd place, 4.5 GB

Pitching Line April – July: 9-9 2.72 ERA
Pitching Line Aug: 4-2 3.97 ERA
Pitching Line Sep/Oct: 1-4 5.05 ERA

1984: on Aug 1st Cubs are in 2nd place, 0.5 GB

Pitching Line April – July: 4-5 5.56 ERA
Pitching Line Aug: 1-0 1.69 ERA (10 IP)
Pitching Line Sep/Oct: 0-0 6.75 ERA (4 IP)

1987: Traded on Aug 21st to Giants. Giants are in 1st place, 1 up

Pitching Line April – Aug 20th: 8-6 2.75 ERA with PIT
Pitching Line Aug 21st – Oct: 5-3 4.32 ERA

1988: on Aug 1st Giants are in 2nd place, 4.5 GB

Pitching Line April – July: 14-5 2.93 ERA
Pitching Line Aug: 3-2 3.47 ERA
Pitching Line Sep/Oct: 2-4 3.33 ERA

1989: on Aug 1st Giants are in 1st place, 1 up

Pitching Line April – July: 13-5 2.45 ERA
Pitching Line Aug: 2-1 5.02 ERA
Pitching Line Sep/Oct: 2-2 4.05 ERA

Combine that with the 1-4 5.85 ERA numbers Reuschel sported in the post-season and I’m not seeing someone saying “jump on my back, I’m bringing the title to town”. April and May games should count the same as August and September but that doesn’t make the late season results any easier to swallow.
   362. Chris Cobb Posted: November 03, 2008 at 11:00 PM (#3002139)
Al,

How do Reuschel's August-October numbers look in his other seasons? I don't dispute that he didn't "carry" his teams in these instances, but the claim that Reuschel was not a "big-game" pitcher would only be borne out if he pitched just as well in the second half as the first half in seasons in which his team was not in contention. If his seasons show the same pattern in every season, then what we are looking at is a pitcher who consistently wore down in the second half of the season. That wouldn't be a surprising result for a pitcher with Reuschel's physique. His conditioning, or lack thereof, surely affects his merit, but saying that Reuschel was not a "big-game" pitcher is another way of saying that he is a "choker," and this evidence is not sufficient to justify that claim.

Certainly the April-July Reuschel should get some credit for helping to get his teams into contention on August 1. If Reuschel had pitched full seasons like he pitched the first halves of these seasons, he wouldn't be a borderline candidate: he'd have been an obvious choice.
   363. Al Peterson Posted: November 03, 2008 at 11:12 PM (#3002148)
Chris,

I didn't do an exhaustive accounting of the Reuschel splits. Here are records and ERAs by month:

Apr/Mar 22-22 3.68
May 41-25 3.25
June 45-21 3.04
July 31-39 3.55
Aug 42-33 3.14
Sep/Oct 33-41 3.66

Maybe slightly higher ERA for the 2nd half of the season. The differential seemed most extreme in the seasons his team was in contention.

1. If Reuschel had pitched full seasons like he pitched the first halves of these seasons, he wouldn't be a borderline candidate: he'd have been an obvious choice.

That I agree on but borderline is where he's at. I'm not saying throw him from the consideration list, just he is lower than some others I like.
   364. sunnyday2 Posted: November 03, 2008 at 11:13 PM (#3002150)
Pitchers who are not directly comparable by using Chris Cobb's measures mentioned above, however, they merit consideration by the electorate:
Jim McCormick
Dick Redding
Leroy Matlock
Hilton Smith
Bruce Sutter
Lee Smith


This wouldn't be my list, specifically, but yes yes yes. Long-time HoM voters know there are lots of excellent pitchers not on Chris Cobb's list. I would add:

Don Newcombe

Creative number crunching is cool. But there's a need for creativity at a higher level as well.
   365. sunnyday2 Posted: November 03, 2008 at 11:15 PM (#3002152)
PS. Considering Chris has John McGraw #3, I am shocked that his pitcher ratings don't have room for any peak candidates.
   366. sunnyday2 Posted: November 03, 2008 at 11:16 PM (#3002156)
Brock,

Tommy Bond has been on my ballot for 100 years. He was more better than his peers than just about anybody not elected yet. Say what you want about pitching in his day. I personally give 50 percent of pitcher value to his fielders. But his value above his peers is very very much.
   367. Chris Cobb Posted: November 04, 2008 at 12:28 AM (#3002215)
sunnyday2,

The pitcher data I compiled are, as I explained, specifically measuring career value above replacement: they are not measuring peak value. They provide data for use in ranking: they are not a ranking system. Pitching peak has not lacked for advocates, and I didn't see that I had any new insight to add to arguments about peak pitching, except that it is being overemphasized, I think, at the expense of career value.

Rather than being shocked at my not including peak candidates, why not make an argument as to why peak value is more important than career value? Being shocked at my not including peak candidates is a bit silly anyway, since I began the study with a pool of over sixty pitchers, including every pitcher who received a vote in the 2008 election, so it's not as if I disregarded anyone. Is it any surprise that Dizzy Dean does not have one of the top 25 career runs above replacement totals among eligible pitchers? That's why he's not on that list, because that's what the list was showing. It was _not_ a ranking, nor did it present itself as being a ranking. I'm sorry to sound cranky, but I went to great trouble to be inclusive in my study, so it is rather frustrating to get responses from a old-timer such as yourself that don't actually engage with what I did.

As to my own ballot, I am still in the process of sorting through pitchers, which basically entails adding peak value to the career measures I have already compiled. I specifically mentioned peak pitchers Rucker and Hahn as having a shot at making my ballot, as well as Don Newcombe, who, with appropriate credit, is a medium-career length candidate. Urban Shocker is on my ballot now. Is he not a peak candidate? With only 2681 IP for his career, he's not exactly a career candidate. I suppose you would call him a prime candidate? I would think that my inclusion of McGraw would show that I am not averse to peak candidates, as long as the peak is good enough. I am not yet persuaded that any of the eligible pitchers who has only a peak case has a good enough one to make my ballot.

My standard for a "peak and nothing else" candidate is essentially this: if the player or pitcher has a good argument to have been the best in the game during their five-year peak run, that five years is enough to elect them. That was the case for Hughie Jennings; that was the case for Sandy Koufax, and I supported both of them. John McGraw is more like Wes Ferrell: he has about seven years worth of greatness. This is not a subjective standard. Rather, it is the point in my system at which a peak only candidate tends to past the electability threshold. It seems, however, to make subjective sense. I don't think any of the "nothing-but-peak" pitching candidates were quite that good, so I doubt they will reach that threshold. But I am going to run the numbers in the new system for Rucker, Hahn, Dean, and see what happens.
   368. Paul Wendt Posted: November 04, 2008 at 12:48 AM (#3002222)
Brock (bjhanke)
10. Sam Leever
A ridiculous oversight on my second prelim. I meant to move Deacon Phillippe into Sam's spot, but I didn't mean to just discard Leever.


The Deacon Phillippe comment was a puzzler.


357. bjhanke Posted: November 02, 2008 at 11:51 PM (#3001450)
Well, here's my final ballot for the 2009 (? Is the year right?) HoM list. I gave up on tweaking and obsessing, because it's to no real purpose, given the deadline. Right now, this is the best I can do.
334. rawagman Posted: October 31, 2008 at 12:59 PM (#3000530)
2009 Ballot prelim - by the way: when are we holding the actual election?

Brock and Ryan,

I understand that the date "November 3, 2008" is the prospective date to open the Ballot Thread, except that we have moved to Tuesdays, and this Tuesday may be special. Deadlines are posted more prominently at the end of the preface.

Anyway there is no deadline to post the preliminary ballot before the Ballot Thread opens. On the other hand there is some recommended lag between prelim ballot and final ballot, so the election deadline is not a practical prelim ballot deadline.

Ryan (rwagman) on several 1Bmen
. . .
45)Tony Perez - I wasn't giving him the extra glove credit he earned through 5 seasons as the Reds' 3B. Still, no peak. As far as 1B go, I have Cepeda up higher because of his very nice peak and his not too short career as a regular. Ben Taylor suffers from a lack of documented stats. The stats there show that he could flat out mash the ball by dead-ball standards. Contemporaries say his glove was the best they had ever seen at 1B. Very apt, in this year of the Hernandez discussion. How much as a scoop worth? I think it's worth alot. I maintain that while a below average defensive 1B can cause little harm, an above average glove at 1B will provide a hefty bonus to the team lucky enough to employ one.

I suppose the point about scoops is a modern one, because other plays at 1B (or those other which occur frequently) are considered routine while scooping is considered uncommon. The general point is the only the good old chestnut about the shape of skill distributions in baseball (some say talent). Fundamentally the point is simply that anyone who would be far below average with the glove at 1B does not play 1B in the majors. Right?


356. Juan V, posting on behalf of Juan V. Posted: November 02, 2008 at 11:18 PM (#3001442)
You're missing your 15th guy.

I understand that is the basis for Howie's remarks such as
> Sell me on your favorite missing player
> :)


Chris Cobb #354, footnote to prelim ranks 1 to 15:
. . . It’s also possible that another pitcher could bounce up—I haven’t finished my calculations there, but I’ve done all that I think truly likely. Rucker and Hahn probably have the best shots. It might be that none of these players will make my ballot, but it’s possible that 2 or 3 will. Many will end up in the 16-35 range, I think.
365. sunnyday2 Posted: November 03, 2008 at 07:15 PM (#3002152)
PS. Considering Chris has John McGraw #3, I am shocked that his pitcher ratings don't have room for any peak candidates.

Marc (sd2),
-- peak only or "pure peakist" I suppose you mean. Not Rucker or Hahn, consistent prime candidates, but Tommy Bond and Dizzy Dean?
   369. Paul Wendt Posted: November 04, 2008 at 01:00 AM (#3002225)
Howie Menckel #355
7. BOB ELLIOTT - Good to see him mentioned in a discussion thread 6-7 years back, at least. Six seasons of at least 134 OPS+, ALL of them as a 3B. Wish he'd played all 3B and not much OF, but c'est le vie - Sewell seemed to get treated as a full SS by some. Beats out HOMer Boyer (see Boyer thread for details) and compares remarkably well with HOMer Santo as a hitter (see Santo thread for more details). Better than HOMer Hack as well, and better than HOMer DaEvans (see DaEvans thread).

Several account books include for Ken Boyer a line for military service in Korea.

10. DAVE CONCEPCION - 8th time on my ballot. Peak is as good or better than Fox's; not quite as consistent, but a slick fielder and a very useful offensive weapon many times. Not fully buying the "other teams were stupid enough to play ciphers at the position" argument; that helped the Reds win pennants, but Concepcion can't get full credit for that stupidity. But he needs the modest credit in that regard to outlast Rizzuto. Similar case to Bancroft, whose prime I preferred in other years to Concepcion's length. It's close.

but Bancroft does not make the 15?
By the way I'm half as confused as Juan V. Are these all recycled comments?

13. BUCKY WALTERS - 4th pitcher on my ballot. Seemed to get Jim Palmer-like defensive support, without enough super-stats to make that irrelevant. Proved his mettle outside of 'war years.' Lemon-esque, though I wasn't a big fan there.
14. DALE MURPHY - [7th outfielder] His modest fan club will rejoice that I've seen the light, at least in terms of a 14th-place vote. A different peak-primieness than Belle, and a different fade as well.


I may be in the fan club. The HOM chapter is really weak.
By the way, three votes for those who play the basepaths: Bob Elliott, Dave Concepcion, Ben Taylor.
"those who play the basepaths" counterclockwise from the catcher

ORLANDO CEPEDA - Suddenly popped up on my ballot 10 years ago with the reevaluation. Had been losing out to Perez with positional consideration, but closer look shows a sterling top-4 and top-10 offensive line. DH opportunity added nothing to his case.

Practically, the DH rule was just too late for Cepeda, Frank Howard, Killebrew, maybe Oliva (depends on injury details, a waste at DH if healthy).

--
I see that I didn't add much here.
The implicit definition of basepathsmen may be the highlight.
   370. Joey Numbaz (Scruff) Posted: November 04, 2008 at 01:56 AM (#3002252)
Regarding Reuschel not being a big-game pitcher - don't all games count the same? I've never understood that.

I could maybe see it if you were talking about how a guy played against other contenders, but I'd want to see it over the whole season, not just September. Those games in April count too.

I always though Reuschel was in excellent shape - wasn't he considered to be a fast runner? I know he has some high baserunning ratings in APBA from the early 80s anyway.
   371. Joey Numbaz (Scruff) Posted: November 04, 2008 at 01:56 AM (#3002253)
And yes, I do realize he was kind of fat.
   372. bjhanke Posted: November 04, 2008 at 12:47 PM (#3002410)
Paul says, "The Deacon Phillippe comment was a puzzler."

Paul, can I ask what was a puzzle, so I can try to make it more clear? I mean, I wrote the thing so it all makes sense to me. That doesn't mean it's clear to everyone. The only thing I can think of that would be puzzling is the personal confessional about control. Now that I think about it, this group probably doesn't need to be sold on the idea of talent. All I can say is that, when I wrote the paragraph, I was recalling several disputes I have had with people about that. There appear to be a lot of baseball fans who think that there is really no such thing as talent, that success is entirely a matter of work ethic and "character." If that's what it was, please believe that I didn't think that anyone in particular here needed any such thing. I'm just used to including it in any discussion that comes down to any effect of talent. If the puzzle was something else, I promise to try to write it more clearly as soon as I know what it is.

Thanks, - Brock
   373. bjhanke Posted: November 04, 2008 at 12:54 PM (#3002412)
Paul also says, "Brock and Ryan,

I understand that the date "November 3, 2008" is the prospective date to open the Ballot Thread, except that we have moved to Tuesdays, and this Tuesday may be special. Deadlines are posted more prominently at the end of the preface."

How embarrassing. I completely misunderstood the deadline. Sorry. I'll post up again on the actual ballot thread, although I've budgeted my time for other things (that I've been putting off to do this), so it may only be a repeat of what is here. - Brock
   374. bjhanke Posted: November 04, 2008 at 01:01 PM (#3002415)
Joe says, "Regarding Reuschel not being a big-game pitcher - don't all games count the same? I've never understood that.

I could maybe see it if you were talking about how a guy played against other contenders, but I'd want to see it over the whole season, not just September. Those games in April count too.

I always though Reuschel was in excellent shape - wasn't he considered to be a fast runner? I know he has some high baserunning ratings in APBA from the early 80s anyway."

There does seem to be some sort of extra pressure involved in pitching (or playing) games late in the season in tight races. I don't normally adjust for that, just because it comes too close to those guys I was talking about who want to attribute everything to "character." But I will admit that part of the reason I didn't really look at David Cone is that I got to see him implode in some late games, early in his career. He just seemed to have trouble handling the pressure. So I can see what the argument would look like.

My memory of the early Reuschel, in the 1970s, is that he was tall and lanky. He put on weight later on. Besides baserunning, he could hit some early, too. Look at 1974-77. That, too, went away as he gained the weight. Terry Forster is the guy who could hit no matter how much he weighed. - Brock
   375. bjhanke Posted: November 04, 2008 at 01:08 PM (#3002417)
sunnyday2 says, "Brock,

Tommy Bond has been on my ballot for 100 years. He was more better than his peers than just about anybody not elected yet. Say what you want about pitching in his day. I personally give 50 percent of pitcher value to his fielders. But his value above his peers is very very much."

I agree. That's why he's on my ballot. I ran the sort just because you can't deal with 19th century pitchers by looking at their seasons played, you have to count the IP. I wanted to make sure I wasn't voting for someone who was just swamped by other IP totals. Bond wasn't. He's only swamped by Galvin, who is a real outlier. Then I saw this name Mathews, and had to look him up because I didn't recognize him. - Brock
   376. Joey Numbaz (Scruff) Posted: November 04, 2008 at 01:11 PM (#3002419)
Wow Brock, I have a completely different view of Cone.

IMO he was one of the 'clutchiest' pitchers of his generation. I haven't looked at it though, but anecdotally, as a Yankee fan, he was the guy I wanted on the mound in a big game.
   377. Chris Fluit Posted: November 04, 2008 at 04:01 PM (#3002596)
As a Blue Jays fan, I was surprised by that comment as well. Could it have been that Cone was not clutch early in his career as a Met before becoming clutch later in his career in the American League?
   378. bjhanke Posted: November 04, 2008 at 11:07 PM (#3003133)
Chris (and Joe as well) says, "As a Blue Jays fan, I was surprised by that comment as well. Could it have been that Cone was not clutch early in his career as a Met before becoming clutch later in his career in the American League?"

Very possibly. I don't know as much about his late tour with the Yanks, because I follow the AL much less than the NL. My memories are mostly from his early tour with the Mets. That's why I distinguished which part of the career I was talking about. I don't remember him melting down with the Yanks, but that doesn't mean much. I certainly can't include it in my comment, because I simply wasn't following him after he moved to the AL. If I really spent some time scouring newspapers from the late 1980s, I could probably document at least some of it. At least a couple of his meltdowns occurred on national TV, and so made the papers. If I knew what term the paper guys use for what I call imploding, I might even be able to Google it. I'll try to see if I can find the right term tonight. Always better to document rather than just remember. - Brock
   379. bjhanke Posted: November 04, 2008 at 11:17 PM (#3003138)
Well,whaddaya know? I found some. Goggling David Cone and meltdown produced a lot of results, most of which weren't about the subject at hand. But ultimateMets.com has this exchange, which is certainly part of what I remember, especially the second one, where he held the ball and imploded at an umpire. - Brock

"I remember one of his first starts with the Mets and he was getting shallacked. He was on the mound and he was either crying or hyperventalating or something. He lost it. He came a long way and turned into a real stud of a pitcher. Trading him away was a huge mistake by the terrible GM Al Harazin. I still love this guy although I can't stomach having him on the Yankees. I root for most ex-Mets and I rooted for him on the Royals and Blue Jays but not on the Yankers. I have to hope he doesn't make the Hall of Fame because if he did, he'd probably totally devastate me and go in as a Yankee. He's border line. I didn't like the rumors about his bullpen antics. They're not fit for print. He's a weird guy.

Me
January 5, 2001
Say what you want about his personality, but he's got one of the nastiest sliders in all of baseball. That's why he's lasted this long.

SHORTY
January 18, 2001
Remember when Dave held onto the ball and argued with the first base umpire about a play at the bag while like 2 other base runners were rounding the bags and scoring on him? Was it 2 guys who scored? I think it was maybe against the Expos. I don't know I just remember how funny it was that his team mates were trying to get his attention and he still just belligerently argued with the ump. That was classic. Still in all a great pitcher. They really screwed up letting him and Doc go. Then they go and pitch a perfect game and a no hitter for the Jankees. Sometimes I think we're cursed."
   380. bjhanke Posted: November 04, 2008 at 11:26 PM (#3003141)
BTW, the bits from my quote above about Cone's "bullpen antics... not fit for print" and "his personality" have to do with things like sexually exposing himself to female fans while in the bullpen (yes, seriously, I just didn't inlcude that as part of the quote, because it's not germane to his play), stuff like that. It's NOT what I'm talking about in terms of analysis. I'm talking about implosions on the field. He seems to have been just really volatile when young. Couldn't handle pressure. It's perfectly plausible that he might have grown out of that by the time he reached the Yanks. - Brock
   381. Paul Wendt Posted: November 05, 2008 at 05:59 PM (#3003766)
Brock,
Narrowly, the Phillippe comment was puzzling because it suggested to me that Leever would be somewhere in the vicinity --a little closer than same URL!
http://www.baseballthinkfactory.org/files/hall_of_merit/discussion/2009_ballot_discussion/P300/
Broadly, I haven't yet read the essay on Babe Adams & Co., with the preceding URL, so I must say no more.

--
ah, Google. Without that quotation by brock I would never in a million years searched the web for cone "shallacked" or "hyperventalating" but, lo,

Nhits
11700 hyperventalating
6070 cone hyperventalating
953 david cone hyperventalating
57 'david cone' hyperventalating

Here google helpfully suggests an alternative with "hyperventilating", which yields more hits, although many of the unaltered hits do highlight "hyperventilating".

1910 shallacked [the very first is pitcher "Dice-K still not right"]
23 cone shallacked
9 david cone shallacked
9 'david cone' shallacked

Here google makes no helpful suggestion and the very first hit for "shallacked" is pitcher "Dice-K still not right". I suppose that shellac is now obsolete, with its only surviving application to the old contest between batsman and twirler.
   382. bjhanke Posted: November 06, 2008 at 12:03 PM (#3004208)
Paul -
You do better if you ignore the guy's lousy spelling and use "hyperventilating" (as you noted) and "shellacked." "David Cone shellacked" produces pages of hits, most of which are not germane, but some of which are. Just as an afterthought, do you know if the current term "shelled" comes from "shellacked" or from military shelling? Could be either, as far as I know.

Oh, and thanks for the response on the Phillippe post. Yes, that was just a brain fry. And BTW, since you seem to have an interest in those early Pittsburgh staffs, too, have you looked at Vic Willis? He's the poster boy for Fred Clarke's influence. He was a veteran pitcher when he moved to the Bucs, and then he stopped walking people. Well, that's an exaggeration, but his walk rates do go way down the day he hit Pitt. I don't include him as one of the Five (they would be Six with him) because his move to the Bucs is so late. But there is more evidence in his move for a Clarke influence than anywhere else.

Thanks! - Brock
   383. Howie Menckel Posted: November 06, 2008 at 01:58 PM (#3004242)
My comments were recycle at this point. Rickey and 13 returnees, I suppose. So short one for Could be Bancroft.
   384. DL from MN Posted: November 06, 2008 at 03:07 PM (#3004299)
Just a thought Brock - how much of the Pittsburgh pitching success could actually be Honus Wagner?
   385. Paul Wendt Posted: November 06, 2008 at 07:54 PM (#3004546)
From the perspective of, say, Frank Chance in the rh batter's box,
Leach, Clarke, Wagner for a half dozen years, then
Clarke, Wagner, Leach for a half dozen years.

Vote Leach!
   386. bjhanke Posted: November 07, 2008 at 06:40 PM (#3005009)
DL (disabled list?) asks, "Just a thought Brock - how much of the Pittsburgh pitching success could actually be Honus Wagner?"

On their overall performance? Very high. He was without question one of the best defensive shortstops ever, and he hit like a demon. That will help you win some games and keep your ERA low. But in terms of control, I'd say none. I don't see how Honus could improve the control of a bunch of pitchers. He was neither the manager nor the catcher. I don't know exactly what sort of algorithms people use to try to separate the defense from the pitching, but the Pirates of that time are a hard test case. They had an extreme ground ball staff, if I read the sources right. The effect of that may have been to actually make Honus look a bit better on defense than he actually was. The approach I'm used to seeing, to determine the effect of a particular defender on a particular pitcher, amounts to comparing the defender's stats with one pitcher against a control group of the other guys on the staff. In this case, the whole staff is ground ball pitchers who never walk anyone. With that as a control group, Honus ends up looking worse, because he appears to be helped by the ground ball effect all the time. I don't know if this is accurate currently, because I don't know who is adjusting for that and how. But in simple terms of helping the cause, of course Honus was a help. When you're a ground ball pitcher in a high-errors environment, it helps to know that the infield behind you is going to get a lot of those grounders you're giving up by throwing all those low strikes. It also helps to know that Honus Wagner is coming up in your batting order to support you with runs.

Was that the question you wanted answered? It's possible I went off on a tangent there.

Thanks, - Brock
   387. David Concepcion de la Desviacion Estandar (Dan R) Posted: November 07, 2008 at 10:32 PM (#3005178)
I'd note that my outdated DRA numbers, which account for the GB tendency of pitching staffs, show Wagner as a good, not great, SS, with 8.1 fielding wins above average from 1903-1916. His ranks among the 16 starting MLB shortstops for those years were 5, 11, 3, 4, 5, 6, 6, 9, 9, 6, 10, 4, 7, and 9.
   388. Joey Numbaz (Scruff) Posted: November 11, 2008 at 05:47 PM (#3006854)
I guess it's time to get this puppy rolling . . . when should we start the balloting?
   389. rawagman Posted: November 12, 2008 at 12:22 AM (#3007176)
next Monday?
   390. bjhanke Posted: November 12, 2008 at 05:47 PM (#3007542)
Dan says, "I'd note that my outdated DRA numbers, which account for the GB tendency of pitching staffs, show Wagner as a good, not great, SS, with 8.1 fielding wins above average from 1903-1916. His ranks among the 16 starting MLB shortstops for those years were 5, 11, 3, 4, 5, 6, 6, 9, 9, 6, 10, 4, 7, and 9."

Wow. That is just disorienting. I might have expected mostly 1s, some 2s, a couple of 3s and maybe a 4 on a bad year, because of the grounder tendency of the Pittsburgh Five et al. But nothing like this. You did note that the method is outdated. Everything I've ever heard or seen about Wagner said very clearly that he was a truly top shortstop. If these numbers turn out to be anywhere near right, that's going to be one of the most serious reputation-busters I've ever seen. - Brock
   391. David Concepcion de la Desviacion Estandar (Dan R) Posted: November 12, 2008 at 08:51 PM (#3007767)
With Tinker, Dahlen, and Wallace running around, I'd have expected him to settle into the 2-6 range. But this is obviously lower. That said, until I get complete DRA data from Michael to run regressions against modern PBP metrics, I have absolutely no idea of how much stock to put into these numbers.
   392. Bleed the Freak Posted: November 14, 2008 at 12:45 AM (#3008789)
77. Chris Cobb Posted: December 02, 2007 at 11:34 PM (#2631787)

16. Jimmy Ryan (20). Est. Total = 244.8. The best of the remaining 1890s outfielders. Career shape is a lot like Andre Dawson’s, actually. I’ve moved him up this year, and the numbers would have him on my ballot, but I don’t fully trust them yet. My review of the nineteenth century has started with pre-1893 pitchers. 1890s outfielders will be next.


Chris, have you finished you review for 1890s outfielders or for other positions pre-1893. I am interested to see your analysis, as I have only elected to my pHOM 16 players who played primarily in the 1890's. Anyone know what happened to Jimmy Ryan in 1901?


28. Brent Posted: June 16, 2007 at 09:27 AM (#2405980)

With no data on playing time, based on post # 14 above, I’ll assume 7500 AB (which seems reasonable—I assume his career lasted from 1899–1914, or 16 seasons). Running through the formulas and converting his record to the 1899-1914 NL, we get:

PA AB H TB BB+HBP Avg OBP SLG OPS
8097 7500 2459 3136 597 .328 .377 .418 .795

Avg OBP SLG OPS / LgAvg LgOBP LgSLG LgOPS / BA+ OBP+ SLG+ OPS+
.328 .377 .418 .795 / .267 .332 .350 .682 / 123 114 119 133


Based on this admittedly small sample, it appears that by not being aware of the Cuban League hitting context, many of us may have been undervaluing Monroe.

What about non-statistical information? Riley’s comments on Monroe’s defense are effulgent with praise. Even after taking these comments with more than a grain of salt, it appears that Monroe was a very well regarded infielder. For example, “he excelled at fielding bunts and was considered to be a better fielder and hitter than his white contemporary at third base, Jimmy Collins”; and “he was called the ‘king of the second basemen,’... and ‘the most sensational player on the American Giants’ team.”

What about his teams? According to Gary, he was probably born about 1877, so his prime was spent with the Cuban X-Giants (1900-02), Philadelphia Giants (1903-06), and Brooklyn Royal Giants (1907-10). During his stay with the X-Giants, they were almost certainly the top team in black baseball. A year after his move to Philadelphia, the Giants dethroned the X-Giants as the top team in the east. Rube Foster and Pete Hill also moved to the city of brotherly love, so we obviously can’t attribute the shift in dynasty entirely to Monroe, but at least we see that he moved in the direction of improving teams. The Royal Giants didn’t win the east until his final season there in 1910, but again, they improved when he moved. With the Chicago American Giants (1910-14), he was in his mid-30s and the batting statistics suggest he was slipping, but again he was playing for a dominant team.

According to Riley, Monroe batted third with the Cuban X-Giants and the Philadelphia Giants, leadoff with the Brooklyn Royal Giants, and cleanup (behind Pete Hill) with the 1911 Chicago American Giants.

Overall, this information suggests that I’ve probably significantly underrated Monroe. I wish we had more statistical information, but a principle I try to follow is to not hold a paucity of statistics against a Negro League player—to downgrade individual players based on the uncertainty due to small samples would have the effect of biasing against the entire class of players.

Is there a case for not placing him high on my ballot?

29. Brent Posted: June 16, 2007 at 10:17 AM (#2405995)
The MLEs shown above come from 1906-08 when Monroe was probably ages 29 to 31. Therefore, assuming a normal aging pattern, they are probably a little higher than what we'd get from his entire career, but perhaps a little lower than his peak (which typically comes at ages 25-29, i.e., 1902-06).


The above quote was taken from the Bill Monroe thread. Has anyone else performed MLE's on Monroe since Brent's post? Is the data still too sketchy? 2B is the position I have the fewest pHOM electees, and Monroe strikes me as the best candidate among eligible 2B's.
   393. Howie Menckel Posted: November 14, 2008 at 01:02 AM (#3008793)
Yeah, let's start balloting on Monday.
   394. Paul Wendt Posted: November 14, 2008 at 01:10 AM (#3008796)
> Anyone know what happened to Jimmy Ryan in 1901?

At the end of the 1900 season he was benched(?) or sent home for the last road trip(?) or demoted from captain(?). I don't recall which but the handwriting was on the wall. He requested his release to become a minor magnate, player/manager/president or partner (a la Kid Nichols one year later), and he got the both the release and the ballclub. That was in St Paul, Minnesota, the city Comiskey had abandoned for Chicago one year earlier.

I don't know whether the 1901 season in St Paul was a failure, or whether Ryan was frozen out by the parties who put the AA together in 1902 (including St Paul MN), or whether the Washington Senators made him an offer he couldn't refuse.
   395. Chris Cobb Posted: November 14, 2008 at 01:45 AM (#3008808)
Chris, have you finished you review for 1890s outfielders or for other positions pre-1893. I am interested to see your analysis, as I have only elected to my pHOM 16 players who played primarily in the 1890's.

I have finished doing a review of the top remaining 1890s candidates. I don't have time tonight to post my full analysis, though I should be able to do so over the weekend.

Here's a summary.

My approach was to come up with estimates for pre-1893 value that would fit with Dan R's WAR2 numbers for 1893 on. I came up with what I thought were plausible league adjustment factors for the period 1885-1892, extrapolated positional replacement levels from Dan's trend lines in the 1890s, and used BP's BRAA, FRAA, and a little bit of estimated baserunning value to come up with estimated WAR2 careers for Ryan, Duffy, Griffin, and Long, who are the top four players in my 1890s queue, setting aside John McGraw, who has no pre-1893 play to estimate, and Ted Breitenstein.

None of those four are close to the HoM as my estimated WAR2 sees it.

Of that group, Mike Griffin does the best. He looks comparable to modern centerfielders Cedeno, Lynn, and Pinson and to near-contemporary Roy Thomas.
Hugh Duffy is a step behind Griffin. He is highly comparable to Tony Oliva among moderns and Fielder Jones among near contemporaries.
Ryan is quite a bit below that. He shows up as comparable to guys like Dave Parker and Sam Rice among right fielders.
Long is a bit farther down than Ryan. There are a lot of modern players with numbers in his range whom I have never studied, I am sure, but of players I have run through my system using Dan R's WAR2, he is closest in value to Bill Madlock and Al Oliver, although the value he provided was of a completely different sort than Madlock's.

I haven't done Van Haltren or Lave Cross, but I think it is pretty unlikely either of them would do better than the above group. Dan R has mentioned how a lot of Cross's apparent value in WARP1 comes from its crazy fielding replacement level, and Van Haltren is unlikely to look much different than Ryan.

The only unelected player from the 1890s I am supporting for the HoM at this point is John McGraw.

16 players from the 1890s for a pHoM does not seem like an unreasonable number.
   396. David Concepcion de la Desviacion Estandar (Dan R) Posted: November 14, 2008 at 01:49 AM (#3008811)
McGraw did play in 1891 and 92, but didn't accumulate any significant value.
   397. Chris Cobb Posted: November 14, 2008 at 03:02 AM (#3008845)
Hey Dan, would you have time to run Gavvy Cravath's minor-league MLEs through your system?

Brent did very solid offensive MLEs for him, which appear in post 176, and his likely fielding value could be estimated fairly readily based on surrounding seasons for his Minneapolis sojourn. His PCL fielding would be a bit more of a guess, but there's some data around that one could eyeball.

Filling in rough estimates for Cravath based on MLE OPS+, he looks on the cusp of making my ballot, so I could really use a more precise evaluation.
   398. David Concepcion de la Desviacion Estandar (Dan R) Posted: November 14, 2008 at 11:44 AM (#3008934)
Yes, but not for about a week, as I'll be crazy busy with a story I need to finish. Is next weekend (the 22-23) OK?
   399. Chris Cobb Posted: November 14, 2008 at 01:24 PM (#3008962)
Sure. If the election starts Monday, the polls won't close until the 24nd at least (maybe we'll have 2 weeks to vote?), and once I have your numbers, it only takes me a few minutes to run them through my system. I'll try to pitch in by getting fielding win share estimates together for you to use.
   400. Ken Fischer Posted: November 14, 2008 at 02:30 PM (#3009016)
Here's my prelim...it still needs a lot of work

1-Rickey Henderson
2-Dick Redding
3-George Van Haltren
4-Mickey Welch
5-Carl Mays
6-Vern Stephens
7-Wally Schang
8-Bob Johnson
9-Tony Mullane
10-Tony Perez
11-Luis Tiant
12-Lou Brock
13-Jim McCormick
14-Reggie Smith
15-Bucky Walters
Page 4 of 5 pages ‹ First  < 2 3 4 5 > 

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
dirk
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

2015 Hall of Merit Ballot
(100 - 9:57am, Dec 22)
Last: rawagman

Herman Long
(11 - 9:22pm, Dec 21)
Last: Joey Numbaz (Scruff)

2015 Hall of Merit Ballot Discussion
(107 - 9:08pm, Dec 21)
Last: Joey Numbaz (Scruff)

Most Meritorious Player: 1960 Ballot
(11 - 2:37pm, Dec 21)
Last: bjhanke

Most Meritorious Player: 1901 Discussion
(29 - 4:50pm, Dec 18)
Last: John (You Can Call Me Grandma) Murphy

Ben Taylor
(81 - 11:04am, Dec 15)
Last: DL from MN

New Eligibles Year by Year
(962 - 10:09am, Dec 15)
Last: Chris Fluit

Cristóbal Torriente
(82 - 11:34pm, Dec 14)
Last: Brent

Brian Giles
(59 - 4:08pm, Dec 14)
Last: Joey Numbaz (Scruff)

Most Meritorious Player: 2014 Discussion
(38 - 12:16pm, Dec 13)
Last: lieiam

Larry Doby
(95 - 10:00pm, Dec 12)
Last: Howie Menckel

Pedro Martinez
(159 - 6:44am, Dec 11)
Last: bjhanke

Most Meritorious Player: 1960 Discussion
(25 - 2:12pm, Dec 09)
Last: DL from MN

Most Meritorious Player: 2014 Results
(7 - 9:24pm, Dec 05)
Last: Chris Fluit

Most Meritorious Player: 2014 Ballot
(29 - 4:20pm, Dec 05)
Last: DL from MN

Syndicate

Page rendered in 1.3046 seconds
43 querie(s) executed