hot topics

Some colorful also-rans this year.

Perfect game men Don Larsen and Harvey Haddix. Pull-hitting specialist Wally Moon. The "other" Frank Thomas. Long-time broadcaster Tony Kubek. And one of the ugliest men to ever put on a major league uniform, Don Mossi.

On Wally Moon from baseballlibrary.com:

"'Moon shots' were what they called the home runs that lefthanded-batting Wally Moon golfed over the 42-foot wall a mere 250 feet down the Los Angeles Coliseum's left field line with his inside-out swing. Though he didn't hit many of them, the newly arrived Moon hit them early in 1959, and his ingenuity seemed to capture the spirit of the ragtag bunch that took the Dodgers from seventh place to the World Championship. Appropriately, he scored the last run ever in the Coliseum."

I listened to Haddix' perfect game on the radio. My hometown radio station in Minnesota was part of the Milwaukeee Braves network. I was supposed to be asleep but I could not turn the radio off.

When the redefined no-hitters and perfect games and knocked Haddix off the list, that was when I knew that MLB was no damn good.

I remember having a Don Mossi baseball card from the mid-50's; Indians I think. As a 10 year old kid I was struck by the absense of beauty looking back at me from that card.
A face that undoubtedly caused his own mother to pause and reflect before she claimed she loved him.

Here's something you don't see often: Mossi pitched almost 1,550 innings over 460 games.
During all that time he commited only 3 errors.
That must be close to some kind of record for fielding competency by a pitcher.
I can see it now: during his offseason salary negotiations, the Tigers GM says to Don, "you gotta take a cut in pay this comming year Don; after all, you made an error last year."

As this impacts several pitchers either currently under consideration or to be considered in the future, I'm re-posting this post by Bill Deane on SABR-L:

Subject: Clutch Pitchers

More than 20 years ago, Pete Palmer contributed what I think
was one of the best statistical analysis efforts ever done. The results were published in an article entitled "Do Clutch Pitchers Exist?" in SABR's Spring, 1985 NATIONAL PASTIME. Pete examined the 529 pitchers with at least 150 decisions between 1900-83, accounting for how many runs each pitcher allowed, how many were scored on his behalf, and what his career won-lost
record "should" have been based on that data. He was searching for "clutch" pitchers: men who won significantly more games than they should have, because of some unusual ability to pitch to the score and emerge victorious in the close games.

What Pete found is that most pitchers wound up with about as
many wins as they should have, with variations within the rules of random chance. In other words, if you win more games than expected, you're lucky, and if you win fewer, you're unlucky. His conclusion: "clutch pitchers do not exist."

Pete has updated and fine-tuned his research since then, and
graciously permitted me to share some of the new results. His current study includes all pitchers with at least 200 starts and 200 decisions between 1876-2005. Using Retrosheet data, he was able to use exact scoring figures, rather than the estimates used in the 1985 article. Nevertheless, the results were very similar and produce the same conclusions. Here are some highlights:

* Of 496 pitchers in the study, 99 (20%) came within
ONE WIN of projection. Only four pitchers were more
than 15 wins off projection.

* The two luckiest pitchers were both named Welch: Mickey
(+20) and Bob (+17). The unluckiest, by far, was Red
Ruffing (-24).

* Several pitchers might have made the Hall of Fame, or
at least become more serious candidates, had they only
matched their projected records. They include Bert
Blyleven (287-250 to 299-238; I think somehow he
would've managed one more victory), Carl Mays (208-126
to 217-117) , Jim McCormick (265-214 to 280-229), and
Billy Pierce (211-169 to 218-162). Ed Cicotte
(209-148 to 219-138) would also have improved his
credentials, though he's not eligible.

* On the other hand, Rube Marquard (201-177 to 195-183),
Early Wynn (300-244 to 297-247), Happy Jack Chesbro
(198-132 to 187-143), and Smiling Mickey Welch
(307-210 to 287-230) might not be as Happy or Smiling
any more, on the outside of Cooperstown looking in. I
don't think as many people would be touting Jack
Morris (254-186 to 243-197) for the Hall, either.

Bill Deane

HOM by pct at position, thru 1969 voting


HOM batters by percentage of games played at position (min. 10 pct to be listed)

C (8.90) - Cochrane 100, Dickey 100, Hartnett 98, Gibson 95, Campanella 95, Bennett 88, Berra 87, Santop 75, Ewing 47, Kelly 36, McVey 30, White 28, O'Rourke 11

1B (13.84) - Start 100, Gehrig 100, Mize 100, Terry 99, Brouthers 98, Leonard 95, Connor 88, Foxx 87, Anson 83, Greenberg 83, Suttles 70, Wilson 45, Stovey 37, Charleston 35, Musial 35, McVey 31, Jennings 26, Lloyd 25, Heilmann 22, Ewing 19, Kelley 16, Delahanty 15, Hines 12, Lajoie 12, Spalding 11, O'Rourke 10, Dihigo 10, JRobinson 10, Irvin 10

2B (10.15) - McPhee 100, Gehringer 99, E Collins 98, Herman 95, Lajoie 83, Frisch 77, Hornsby 72, Grant 70, Barnes 69, JRobinson 65, Richardson 43, Ward 26, HR Johnson 25, Groh 20, Hill 20, Pike 18, Dihigo 15, Wright 10, Wilson 10

3B (7.23) - Baker 100, J Collins 98, Hack 98, Groh 79, Sutton 69, White 51, Beckwith 50, Wilson 40, Davis 22, Frisch 20, Wallace 18, Dihigo 15, JRobinson 15, McVey 14, Richardson 13, Vaughan 11, Ott 10

SS (15.18) - Pearce 96, Boudreau 95, Reese 95, Glasscock 94, Appling 94, Cronin 92, Wells 90, GWright 89, Dahlen 88, Vaughan 85, Wallace 77, Jennings 70, HR Johnson 70, Lloyd 70, Wagner 68, Davis 58, Ward 44, Beckwith 35, Barnes 28, Grant 20, Sutton 19, Hornsby 16, Dihigo 15, Irvin 10

OF (40.62) - Carey 100, Clarke 100, Hamilton 100, Thompson 100, Wheat 100, Goslin 100, DiMaggio 100, Averill 100, Doby 100, Slaughter 100, TWilliams 100, Ashburn 100, Snider 100, Simmons 99, Burkett 99, Cobb 99, Flick 99, Gore 99, Sheckard 99, Speaker 99, Medwick 99, Jackson 98, Stearnes 98, Keeler 97, PWaner 97, Crawford 94, Ruth 92, Magee 91, Ott 90, Hines 82, Torriente 80, Kelley 79, Heilmann 77, Irvin 75, Pike 73, Delahanty 72, Hill 70, O'Rourke 69, Rogan 65, Musial 65, Stovey 63, Charleston 60, Caruthers 50, Kelly 47, Richardson 40, Suttles 30, Santop 20, Dihigo 20, McVey 18, Ewing 17, Greenberg 17, Davis 13, Spalding 13, Wagner 13, Berra 13, Ward 11, White 10, JRobinson 10

SP (33.23) - Alexander 100, Covaleski 100, Faber 100, Plank 100, Vance 100, Grove 100, Hubbell 100, Lyons 100, Newhouser 100, Feller 100, Ruffing 100, Rixey 100, Wynn 100, R Foster 99, Brown 99, Mathewson 99, Walsh 99, Williams 99, Young 99, B Foster 99, Paige 99, W Johnson 98, McGinnity 98, WFerrell 97, Lemon 97, Keefe 96, Nichols 96, Rusie 95, RBrown 95, Clarkson 94, Galvin 92, Radbourn 78, Spalding 72, Caruthers 47, Rogan 35, Dihigo 25, Ward 16

INF:: 55.20
OF::: 40.62
P:::: 33.23

1B + OF:::: 54.46

C-2B-3B-SS: 41.46
P + C:::::: 42.13

Caveats: Totals treat all careers as equal. A little off on players like McVey and Sutton due to changing schedule length. Guesstimates on Negro Leaguers. Doesn't sufficiently represent pitching weight of players like Ruth or Caruthers.

P.S. I'd be open to 'improvements' on numbers for McVey/Sutton/Ruth/Caruthers types, and all Negro Leaguers.

er, that's thru 1970

Thanks Howie. I feel good that only 2 of my tentative top 10 are 1B/OFers; they represent over 42% (54.46/(54.46+41.46+33.23) of the HoM's population, which is MORE than a bit high IMHO.

min. 10 pct to be listed

Is Medwick's 1% outside the outfield (to pick a person randomly) counted but not listed, or neither counted nor listed.

As requested, here are the latest Minoso numbers.

These MLEs not only respond to Brent's concerns, but also include the newest data just posted by Gadfly. Thanks, Gadfly! I'm going to post this over on the 1971 thread as suggested by several voters.

Thanks to David Foss for noting: the ages in Minoso's MLE should read 19-24.

Thanks Doc!

It looks like I will be giving him credit for 1948-1950, though I am nto sure how much this changes things really. None of those seasons are real peak seasons, just shoulder type seasons. I still think that Willard Brown was better and I have Minoso only a spot or two behind him.

I now have Minoso as slightly better than Brock (he's got a better peak and prime with a comparable career) and behind Burns (Minoso doesn't have enough peak despite more extended prime and career). I don't see him as a terrible HOF or HOM selection, but I also won't be putting his name on my ballot.

Prelim

PHoM: Spahn and Fox

1. Warren Spahn
2. George Van Haltren
3. Willard Bron
4. Cool Papa Bell
5. Biz Mackey
6. Jake Beckley
7. Dobie Moore
8. Mickey Welch
9. George Sisler
10. Tommy Leach
11. Hugh Duffy
12. Nellie Fox
13. Edd Roush
14. Quincy Trouppe
15. Bobby Doerr

16-20. Griffith, Redding, Ryan, Childs, White
21-25. Smith, Streeter, Minoso, Strong, Willis
26-30. Gleason, Greene, Monroe, Mullane, Kiner

I think the biggest question with Spahn is how will Karl justify placing him below Jake Beckley ;-)

Seriously though, I can't see him not being #1, it will be interesting who did better against the backlog, Berra or Spahn (2nd's for Berra, 1sts for Spahn), I believe that Spahn will.

Seriously though, I can't see him not being #1

I'll be disappointed if he's not.

Here's something you don't see often: Mossi pitched almost 1,550 innings over 460 games.
During all that time he commited only 3 errors.
That must be close to some kind of record for fielding competency by a pitcher.

I suspect he scared the ball away from him towards the shortstops and second basemen behind him. Either that or his massive dumbo ears helped provide better equilibrium than most players.

I am thinking of a move up into the top 20 for Willard Brown, but I have a bit of a problem with him. While it seems that he could have been an absolute monster in MLB, his walk rates are so low that sometimes I wonder if he may not have been a flop, a kind of Glenallen Hill type of player (not saying that Brown was only as good as Hill here so dont' jump on me). I guess I just see Brown as probably the biggest question mark when it comes to translating NeL performance into MLB performance. However, Gadfly has recently given us some walk data that suggests that Brown could at least walk every once in a while when he absolutely had to (like after a bad season). Is this evidence that maybe Brown would have tried to control the strike zone at least a little had he played in MLB? What do our NeL guys think about this?

jschm-

I agree that Brown is a puzzle. Gadfly's given us a lot data about him along the way that's helped, but I feel the same lack of confidence you do.

On the other hand, it's worth looking at the reverse too. Not everyone who walks a lot makes it to the degree we expect them too. Jeremy Brown hasn't made it to the majors yet. Jose Cruz, despite seeming like a walk/power guy hasn't really grown into a monster. Kevin Mench and Gabe Gross didn't become monsters when they showed impressive minor league walk rates.

On the other hand, the MLEs represent what he did, just moved into a different context, and we're supposed to look at a player in terms of his performance.

Short answer: I've got no idea.

Two comments:

1. Willard Brown's success and impact in the NeLs was significant and he should at some level be evaluated on that basis. What he mighta done in the MLs should not be the sum total of the analysis.

2. BBs were not valued in the NeLs as they were in the MLs (and of course they were not valued in the MLs in the 1940s and '50s as they are today). His BB rates ought also to be evaluated on that basis. And given his obvious ability, if he had played in an environment where the BB was valued by his managers and peers, I believe he would have taken more BBs.

If a "poor" BB rate, IOW, is the only knock on Willard Brown, it is to me a bit much to attach too much significance to that.

I just plugged Gadfly's walk data from the PRWL into my MLEs for Brown and they do raise his walk rate a bit. Revised MLEs posted on the Brown thread.

Too busy to post a full prelim, but here are the relevant details:

(1) Spahn #1--duh.
(2) Fox--off ballot, probably somewhere between 25-30; he was a league average offensive player whose defense was only a few runs per year better than guys like Gordon, Doerr, and to a lesser extent Childs; I know all about durability and career length, but the difference in offensive value between those guys and Fox make them players of a different order
(3) Minoso will appear between 10 and 15--with some extra credit, he looks a lot like Medwick; only LF with a similar or higher OWP and a career of even close to similar length who is out is Bob Johnson and all the small pluses go his way when compared to Johnson
(4) W. Brown moves up from the 20's to 14 or 15 with the new walk data

Add me to the list of persons who have not fallen in love with Nellie Fox. His Win Shares came at the expense of plenty of plate appearances, i.e. outs. The defense is nice enough but then again we aren't exactly dealing with the all metal glove society when you think of Doerr, Gordon, Childs, etc.

People are hanging their hat on a positional dominance and in-season durability argument for Nellie. Will it work well enough?

Preliminary ballot

1. Spahn
2. Mendez
3. Walters
4. Trouppe
5. C Jones
6. Pierce
7. W Brown: Moves ahead of Duffy based on new walk data.
8. Duffy
9. Bresnahan
10. Mullane
11. Browning
12. Cooper
13. Grimes
14. V Willis
15. C Childs

16. Mackey
17. Williamson
18. Oms
19. Redding
20. Newcombe
21. Latham
22. GVH
.
.
.
31. N Fox: Behind Childs and Doyle, ahead of Doerr, Gordon, and Marvin Williams (remember him???).
.
.
.
42. Minnie Minoso: Behind C Jones and Burns in LF.

People are hanging their hat on a positional dominance and in-season durability argument for Nellie. Will it work well enough?

By my own hand count here's how many times the currently eligible 2Bs were the best (by WS) in their leagues:

Childs 8

Doyle 7

Gordon 6

Fox 5

Avila 4
Dunlap 4
Frey 4
Lazzeri 4
Ritchey 4
Schoendienst 4
Stanky 4

Daly 3
Doerr 3
Evers 3
Gilliam 3

Use that info as you will. For those who discount for war, Gordon led in 1943, and Stanky led in 1945.

IMO anyone trumpeting Fox as a league-leading 2B should be very strongly considering Childs and Doyle because of their dominance at the position. That dominance may well superseed Fox's durability and career length. If dominance at the position is the watchword, C & D also come first before the (also Esheresque) Gordon vs. Doerr merry go round starts up again.

I may be the weirdo who doesn't have Spahn first, though if he isn't he won't be far behind. He's severely testing my will as an extreme peakster; his prime/career are very difficult to ignore. (Mind you, his peak is pretty swell too--I'm just not quite sure yet there aren't one or two better ones in the backlog.)

I have a feeling I'm gonna cave, though.

I have a feeling I'm gonna cave, though.

Just don't cave in from "pressure" from us, Mark. As a peak person, you may have a case not to place him #1 on your ballot, so trust your own instincts and analysis.

Mark - I would say that Spahn's peak is pretty darn excellent considering he would have won 5 Cy Young Awards had the 2 league awards existed throughout his career.

I would think even if career is only a tie-breaker this has to be a better peak than anyone on the ballot. Just curious, who would you consider to possibly have a better peak?

"* On the other hand . . . Early Wynn (300-244 to 297-247) . . . might not be as Happy or Smiling any more, on the outside of Cooperstown looking in."

Great post KJOK.

One nit to pick. I'm going to venture a guess that Early Wynn would have also figured out a way to get 3 more wins, and even if he didn't, 3 wins wouldn't have kept him out of Cooperstown.

"His Win Shares came at the expense of plenty of plate appearances, i.e. outs."

PA do not equal outs.

Fox's OBP was 9 points above league average. He made fewer outs per PA than your average player. A lot fewer than your average 2B.

The guy played 15 1/2 seasons as a league average hitter and a defensive wizard at a key defensive position. That has enormous value. He had a peak too. An OBP heavy 124, 114, 117 OPS+ from a GG 2B is an enormous season, people have won MVP awards (deservingly, I'm not talking Jorge Bell here) for a lot less.

I love this guy. I like him more than Gordon, etc., to me he's easily the best 2B on the ballot. He's got 2 seasons on Schoendienst (even though both played about 15 years, Fox was in the lineup more), and was a better hitter and at worst his equal with the leather.

I'll probably have him just above Kiner at #7 or 8 on my ballot this year, which is top heavy with guys that I like that others don't, like Beckley, Cravath, Easter and Charley Jones.

There is a lot more to fielding a championship club than outfielders that hit home runs. There's a reason this guy played in 12 all-star games and was the best player on a team that contended throughout the 1950s.

Also, WS clearly underrates middle infielders, so that 304 also understates his true worth to the team.

He made fewer outs per PA than your average player. A lot fewer than your average 2B.

Yup, 2Bmen during his career hit .262/.328/.358.

Fox had 139 RCAP for his career. During his peak ('51-'60) he had 185 RCAP and 82 RCAA. (RC aren't park adjusted, but I don't think it matters much for Fox since he played in parks that were pretty neutral.)

Al,

WS already takes into accont howmany outs a player makes. To Doc Fox's WS #'s because it took him more outs to accumulate them would be double counting.

Mark - I would say that Spahn's peak is pretty darn excellent considering he would have won 5 Cy Young Awards had the 2 league awards existed throughout his career.

Indeed, he has five NL WS Cy Young Awards: 1947, 1949(Tie), 1957, 1958 and 1961. That doesn't include his excellent 1953 season which was in the middle of Roberts' big run.

I understand where Mark is coming from, though. It can be a bit underwhelming to see those ten (!) seasons of ERA+ between 120 and 130 staring at me when I look at his page at bb-ref. His top two seasons fit in with the best of them, though ERA+ wise and he was a workhorse. Plus a 125 ERA+ is equivalent to a support-neutral .609 WPct. That's a *lot* of 21-14 seasons.

Fox had 139 RCAP for his career. During his peak ('51-'60) he had 185 RCAP and 82 RCAA.

Joe,

PA do not equal outs.

Fox's OBP was 9 points above league average. He made fewer outs per PA than your average player. A lot fewer than your average 2B.

Fair enough, outs equal outs. And Fox still made plenty of them along with plenty of hits, runs, and RBIs. Maybe what I meant to say is when looking at the WS rates its show he was not gaining Win Shares as quickly as some other candidates. And that same Win Share totals are supposed to account for the great defense he had.

As for Fox beating average players and 2B, well then I elect him to the Hall of Average easily. Problem is I look at other 2B candidates and they all ain't average - they're better than that.

Also, WS clearly underrates middle infielders, so that 304 also understates his true worth to the team.

Is this true? I have no sense one way or another. Then again, if it hurts one infielder it hurts all I guess.

On WS underrating middle infielders:

Well, WARP definitely rates them more highly.

Aside from WS, there is definitely an argument to be made that the HoM electorate is favoring hitters over defenders, by position.

This from Howie's positional count:

1B + OF:::: 54.46

C-2B-3B-SS: 41.46
P + C:::::: 42.13

We've elected 13 more 1B/OF than C/2B/3B/SS.

That's getting to be quite a large gap between bat positions and glove positions. Are we giving glove positions their due? If we are not, is win shares' treatment of the value of glove positions unduly influencing us?

>Are we giving glove positions their due? If we are not, is win shares' treatment of the value of glove positions unduly influencing us?

Yes to #1.

No to #2, at least for me. I make those adjustments. It's just that the guys that I favor--that I've put into my PHoM (Ed Williamson, Dobie Moore) don't happen to be the guys that others favor. I think there will be a general agreement to put Gordon and Doerr into the HoM eventually but it's hard to be optimistic about 3B. The obvious guys coming up--Schmidt, Brett, B Robby, etc.--will not by themselves even things up (and B Robby is perhaps not a slam dunk either).

I was not a Terry backer at all, but otherwise the 1B list is pretty good.
I have far more problems with the OF list, like Ashburn, Carey, Medwick, Thompson, Kelley, and maybe too many all-19th century picks, too.

That said, I am one of the "hitting first, then factor in defense" guys. And I barely use WS.

Fox is an interesting conundrum given the borderline rankings of Doerr and Gordon. Tough, tough comparison. Maybe all 3 belong, eventually. But I like Childs just as much or more, sigh.

"His Win Shares came at the expense of plenty of plate appearances, i.e. outs."

PA do not equal outs.

Not sure I follow this, Joe. Yes, outs are the key divisor by which we can make many rate stats, and OBP is still undervaklued by many, but my understanding of Win Shares is that
30 WS in 600 PA = 30 WS in 600 PA, regardless of how many outs are used,
since WS already takes into account that outs cost runs. If I have this wrong, someone enlighten me please.

3 wins wouldn't have kept him out of Cooperstown.

No, but 13 thirteen might...just ask The Nasty Dutchman.

Just curious, who would you consider to possibly have a better peak?

Possibly Méndez, possibly Waddell. Also possibly no one. (You have to remember that short but extreme peaks--even as short as three-year ones--work for me. And that I rely heavily on PRAA these days. And that I'm not as in love with workhorses just for their own sake as most here are.)

But the Cys are impressive, and the consistency is impressive and unmatched, and if I decide to go with Spahn above those two, those factors will be the main reason why. He was obviously the second-best (I'm very high on Roberts) pitcher in baseball for quite a while. I like to mostly compare pitchers head to head, but it would be remiss of me not to notice that the numbers are dropping across the board for the '50s guys, and that's probably not because all the pitchers in that era were worse than the earlier ones.

Regardless, Spahn will easily make my pHOM this year, of course.

I'm really confused about Fox as well. A befuddling election for me.

I have Spahn in the no-brainer #1 slot. All the data I have looked at has Fox about the same as but clearly below Schoendienst. Red isn't near my top 30.

DL,

How is Fox cleraly belwo Schoenkienst?

Mark,

I must say that as a peak voter I wont' think twice before putting Spahn at #1. I also have both Dean and Walters ahead of Mendez and Waddell.

I'm sure Mark is well-meaning and conscientious and all. But any system that would result in Spahn being any less than an overwhelming #1 on this ballot has major shortcomings. He should take it as a red flag that his system has weaknesses that ought to be addressed.

How Warren Spahn could ever be NOT considered the #1 no-brainer on the ballot is puzzling.
I think Spahn is a true no-brainer "inner-circle" pitcher.
THe only guys better than him all-time are named Johnson, Grove, Paige, Clemens and maybe Alexander.
There is nobody else on this ballot who is within 30 or 40 slots on the best 100 all-time players list.
We're talkin thin air here.

Spahn is Cy Young cloned 50 yrs later.

You can argue for Maddux and Seaver >= Spahnie as well, but the point is well taken.

He should take it as a red flag that his system has weaknesses that ought to be addressed.

Now that's pressure! I'm tempted to respond with "Yeah, well, you know, that's just, like, your opinion, man..."

Look, Spahn will still most likely be my #1 (and if I hadn't said anything about my process on this thread, that probably would have been the end of it). But it just isn't as much of a slam-dunk when you're looking primarily at extreme peaks, in my opinion. I've explained why, and it seems one or two people don't necessarily think I'm entirely out of my mind, given my emphasis on those peaks.

I imagine, of course, that you think I shouldn't be looking primarily at extreme peaks, and that my doubting Spahn proves it. But the thing is, I'm not doubting Spahn. If it were up to me, the guys I have the temerity to consider as possibly having better peaks would have been elected years ago, and Spahn would be my no-brainer #1. There's no question in my mind that he should be in the HOM--he is, in fact, a no-brainer.

Personally, I think it's bewildering that you (and others) have Wally Schang within a mile of your ballot, let alone on it. To me, that's just a difference of opinion, though, not a sign that your system has "weaknesses." We just differ (significantly) on how to measure merit in baseball players; to me, extreme dominance, even for short periods, trumps all. The fact that few voters agree with me on that is not proof that I'm crazy or wrong.

So talk me out of my extreme-peak emphasis--I'm willing to listen--but don't just throw comments like this at me, please. They just make me wonder if I should stop commenting on ballot discussions and vote quietly every "year."

So talk me out of my extreme-peak emphasis--I'm willing to listen.

While I'm enjoying a brief moment as the voter at the top of the consensus scores, let me make a case for moderation over extremes :-) . (I think I was last at the top of consensus back in the 1940s, so I expect I'll be there again in the early 1990s.)

To my mind, these are the significant arguments against placing excessive weight on a short set of peak seasons.

1) Small sample size. The fewer seasons you consider, the more weight you give to apparent peaks that could be affected by distortions in the metric or by fluky conditions. A small sample will not only include some players who might not deserve inclusion, it might also exclude some deserving players whose best seasons are depressed by distortions in metrics, fluky conditions, or ill-timed injuries. The counter-claim is that value is value, but if you make that claim then you run into argument two, which is

2) By helping their teams to win many times, players with good long careers can contribute more to their teams' successes than players with great short careers. This claim is supported by the Pennants-Added metric, which calculates the number of pennants that would be added to an average team (I think) by a player over the course of his career. It's hard to make a purely value-based case for merit and focus only on a few peak seasons.

So it seems to me that the argument for judging players based on peak comes back to the "we want to honor the best players, the players who reached the highest pinnacle of performance." And if you want to do that, then you need to look at enough seasons to make sure that you are really seeing the results of a player's achievement and not a run of luck or a metric's blip.

I certainly don't argue against weighting peak heavily in a ranking system, but I think placing very heavy weight on any measure of less than five years, esp. when using Win Shares, which are affected in several ways by contextual factors, will lead to dubious conclusions.

There's a third argument that is not against being an extreme peak voter per se, but that may apply to any ranking system

3) Any system that routinely ignores accomplishments that are obviously meritorious is a system that can be improved. Now "ignores" is a strong word here, and I'm talking about extreme cases. But if one has a system that says, "I look at three-year non-consecutive peak value in win shares" and you have one guy with a career of 30, 30, 30, 15, 15, 15, 10 win-share seasons and that's it and a guy with a career of 28, 28, 28, 28, 28, 28, 28, 27, 26, 24, 23, 22, 21, 20, 20, 18, 17, 16 win share seasons, and your system says, "well, the 30-30-30 guy tops the 28-28-28 guy," then I think the system has a serious flaw. An example demonstrating the narrowness of a purely career system could also be readily devised, I think.

So there's the case.

Chris--

Thanks for that post, seriously. I actually agree with most of what you say.

We can dismiss point 3 right away, as I most definitely don't do that, or anything like it (at least not intentionally...I'm often finding flaws in my estimations from year to year). On the other hand, if the numbers were something like 39-38-38-35-24-23-22-21-15-12 for the first guy, I might be more inclined to him.

As for 1 and 2, that a good pair of crosshairs you have me in. And I should take another look--should always be taking looks--at my favorite candidates with particularly short peaks (Rosen, Dean, Willis, guys like that). For the record, I do fall into the first category you mention, not the second--I'm not a big fan of Pennants Added, as I feel there are just too many variables that go into winning pennants that they don't (and can't) measure. (Even the literal "wins" part of Win Shares creeps me out a bit.) So the sample size problem is the main thing I need to be concerned about. I feel I have been, in fact, though it's true that I need to continue questioning myself here.

Some other notes: for pitchers, at least, I've been relying more on PRAA than Win Shares lately. Waddell and Dean are both peak monsters there. As far as Méndez goes, maybe I'm crazy, but his MLEs have always just wowed me, and I'm convinved by his record against major leaguers that they're for real. And the looming specter of Koufax is part of what convinced me, after I had demoted Dean, at least, due to sample-size issues, that my initial feelings about what constitutes enough of a peak--for a pitcher, at least--were correct. To me, these guys did prove they were for real in that short period, before injuries ended their careers. And they were so good during that period, so dominant, that I'm left with no doubt they deserve inclusion, regardless of their lack of any career to speak of.

Of course, Dean and Koufax are outliers, and more fringy guys (Willis? Rosen?) may be slipping through undeservedly. I'll keep checking myself on that point.

And despite all that, as I said, I think Spahn's record is impressive enough to convince me, regardless of his lack of a completely-blow-me-away peak, raw data-wise. He's a no-doubt HOMer, there's no question about that.

This from Howie's positional count:

1B + OF:::: 54.46

C-2B-3B-SS: 41.46
P + C:::::: 42.13

We've elected 13 more 1B/OF than C/2B/3B/SS.

That's getting to be quite a large gap between bat positions and glove positions. Are we giving glove positions their due? If we are not, is win shares' treatment of the value of glove positions unduly influencing us?

Another way to look at the issue is by counting HoMers according to their primary positions (as listed, for example, in the HoM Plaque Room). By this measure the difference between offense and defense is less severe: 1B + OF = 53; C+2B+3B+SS = 46; a difference of 7.

The tendency for Howie's measure to show more offense should be expected. If primary positions were equally distributed across the defensive spectrum, the distribution by games played would still show a concenration on the 1B + OF positions because that's the direction players move as they age and can no longer play their primary positions. But is it relevant? Cronin was elected to the HoM based on the 92 percent of his career spent at shortstop, and the fact that he played 49 games at first base at age 37 really had nothing to do with why he was elected. This is a case where the less precise measure (the one that assigns each player to a primary position) may be more meaningful than the more precise one that weights the positions by playing time.

Some additional comments:

- Much of the shortage of infielders can be attributed to the underrepresentation of third basemen, which is something I think we've all expected; many writers have noted that there was a lack of great third basemen pre-1950. We'll start seeing better third base candidates soon, but as Bill James wrote in his first Historical Baseball Abstract, "Nobody makes up rules for the Almighty. If he chooses to make three right fielders who are greater than any catcher, you can't tell him not to."

- Although the offensive positions are slightly overrepresented relative to the defensive positions, within each category it is the most important defensive position (shortstop for the defensive group, center field for the offensive group) that has the highest representation -- 17 players each according to the list of primary positions shown in the Plaque Room. This suggests that it is unclear whether defensive positions are actually underrepresented.

- Personally, I'm more concerned about a possible shortage of pitchers -- the 36 pitchers constitute 26.7 percent of the HoM, which seems low to me; I'd prefer to see at least 30 percent. But the voting record suggests that my opinion is not widely shared.

Brent,

I am not sure if it is necessarily that people don't agree with you but that there is disagreement amongst most of us as to which pitchers we should support. I think we are all a little more comfortable with position players than pitchers. So much of a pitcher's value can be decided in how one weighs pitching vs. defense. Also, how do we balance pitchers across eras when their counting stats will look SOOO much different? Then again I guess that could be said for any position huh?

Also if Griffith and Spahn are elected this year (the favorites I would presume) how does this look? Seriously, we could have 40 pitchers by the end of the balloting in 1972.

Interesting, Brent.
Of course, the "92" at SS for Cronin is not counteracted by an "8" anywhere else - the minor alternatives are not listed or credited at all.

I agree with you on the pitchers; we collectively seem more reluctant to elect a Clark Griffith vs a Joe Kelley. I think maybe in part it's because in the early stages we had a lot of 'no-brainer' SPs, so it wasn't as obvious that we also should honor the more modest SP HOMers. I had a heckuva time carrying Eppa Rixey in, for example, lol.

Brent, I agree that a count by primary position shows a less severe glove deficit and may be the better measure. I raise this deficit as a concern partly because it looks like it will grow substantially over the next decade unless we change direction somewhat on infielders. I think, actually, that some change has been happening as both Doerr and Gordon have made substantial moves upward in the last couple of elections. It will be interesting to see if a few other borderline infielders follow suit.

With pitchers, on the other hand, I agree with jschmeagol that we are going to be electing a lot over the next 5 years and, as I favor 27.3% representation for pitchers, I think we are closing in on appropriate representation for them :-) .

I'm guessing, in fact, that when we get to the tail end of the "Greatest Generation" of pitchers in the early 1990s, we'll be up to or slightly above 30% representation for pitchers, and then we'll drop down as the 1980s pitcher wasteland of scragged arms arrives for consideration.

...as I favor 27.3% representation for pitchers, I think we are closing in on appropriate representation for them :-)

That's why you lead us all in consensus score. ;-)

Hey, I usually 'chase' right behind Chris Cobb on consensus!

The hard part is that I have no doubt that even though I am convinced I have had some clout on certain players (leading to a raise in my consensus scores), I'm also sure that Cobb very deservedly has swayed more voters more often - so that HIS consensus score is in larger part a result of voters following in his footsteps...

Mark wrote: So talk me out of my extreme-peak emphasis--I'm willing to listen--but don't just throw comments like this at me, please. They just make me wonder if I should stop commenting on ballot discussions and vote quietly every "year."

We certainly DO want voters to contribute to ballot discussions. We all learn this way.

My style tends to be rather hit and run, "toss in the smoking grenade and watch the fun" sort of contribution, at the risk of being too blunt. We all contribute what we can.

I conscientiously try to avoid calling people crazy or wrong; the idea, as you discerned, was to further discussion. Chris picked up on this and give you a quality critique.

I'm also sure that Cobb very deservedly has swayed more voters more often - so that HIS consensus score is in larger part a result of voters following in his footsteps...

Given that I've been contributing to discussions rather less in 2006 than in previous months, I think my rise in consensus score has little to do with my swaying other voters. . . . unless the lesson is that my guys do better when I keep quiet :-).

I favor 27.3% representation for pitchers


I favor 31.642% representation for pitchers. I hate it when people like Chris Cobb and Albert Belle posit vague numerical generalizations about the appropriate level of representation for pitchers in the Hall of Merit.

27.3% equals 3/11. Three pitchers and eight position players.

... not that everyone must agree with the positional ratios that Chris favors, but that three digit percentage looks much less convoluted and random written as a simple fraction.

I know -- I just thought it was funnily exact.

I've kind of thought that 30% pitching makes the most sense. That's 3.3/11 players. In a HOM of 200, that's 60. I once had a good reason why that was, but I'm not sure I remember it now.

DanG--

Fair enough. Sorry if I freaked out there. I guess I need to get a thicker skin--I need karlmagnus lessons, clearly!

My rationale for the representation of pitchers being around 3 of 11 (as Daryn correctly noted, I went with the percentage to one decimal place for comic effect, not because I seriously believe that the number is exactly right) is that, in my still-unsystematic view of pitching staffs, it seems like there is most usually a significant drop-off in quality from the #3 to the #4 starter on a team, leading me to believe that the #4 and #5 starters on a team are more or less equivalent to the #4 outfielder and #4 infielder: players whom you would like to be as good as possible, but players who are very seldom going to be an above-average player. The rare team may have 4 starters who are all average pitchers or better, just as the rare (great) team may have utility players who could be average players or a little better if they were starters for another team, but that is not generally the case.

In any case, I use the ratio of 3/11 as a guideline only. I don't use a positional quota system in ballot construction. When I do see positional groups gettting far out of balance in my win-share-based system, though, I revise the system. This has happened eight times in the course of my 68 seasons of voting.

1) Shortly after I began, I added a 10% career bonus for infielders excluding 1B.

2) About ten years after that I settled on 30% career and seasonal bonuses for catchers.

3) Around this same time, I increased fielding win shares for pre-1930 players, beginning with 30% for pre-1900 seasons and gradually scaling down to 0 in 1930.

4) About ten years after that I developed my own win-shares system for pitchers because I wasn't confident that WS was handling pre-1920 pitchers correctly.

5) Around that same time I added a peak rate measure to my system because I thought I was undervaluing peak performance.

6) Shortly after that I raised the fielding bonus for first-basemen pre 1930 so that it started at 50% and scaled down from there.

7) About twenty years ago, i removed the fielding bonus provision for pre-1930 outfielders.

8) About ten years ago I added a 10% peak bonus for infielders excluding 1B.

At present I am concerned that infielders are still being underrated, so I'm working on a way to incorporate WARP into my ranking system, since it seems to give more weight to defensive value than WS does. I haven't completed that project yet, and I may decide to chuck it, but I'm concerned about trends in infielder representation on my own ballot and in the HoM, so I'm going to keep at it for a while.

Chris,

I didn't realize how numerical your system was -- does it spit out a global acore for each player? If so, do you give youself the discretion to submit a ballot that is not consistent with that global score?

I'm wondering how many people in general use a global score (with or without discretion) and how many eyeball it. I'm a bit of a hybrid -- in 1915 when I joined I assigned a global score to all of the candidates who had ever received votes and then rank ordered them with small modifications based on feel. Since then I have basically just tried to insert the new candidates into that list, comparing them to each person we have considered in descending order until I get a positive answer to the question "Is he better/do I prefer him as a candidate than X"? Now, if they don't make the top 75, I just discard them.

Every once in awhile new information about a particular candidate or my acceptance or rejection of a general proposition that I had previously disagreed with or agreed with affects this order.

2B WARP3BRARBRAAFRARFRAAG
Fox 83.5276-3846765 2367
Red 82.0 2842442101 2216

What's the difference? Fox has slightly higher replacement value but Red has the better peak.

I didn't realize how numerical your system was -- does it spit out a global acore for each player? If so, do you give youself the discretion to submit a ballot that is not consistent with that global score?

My system does not quite spit out a global score for each player, but it does provide numbers that I apply in a purely mathematical process to place each player within an all-time, non-period-adjusted ranking list. The main numerical part of the system gives me three measures: career win shares, total peak, and peak rate multiplied by the # of seasons (usually five but sometimes higher) in the peak. Because the peaks are of differing lengths, I don't quite have a single global score, so I have to compare players head-to-head, and among a group of players with very similar value, it may happen that because of differences in peak height and length, Player A may beat Player B and Player B may beat Player C, but Player C will beat Player A. Then I have to make a judgment call among the three. But these cases are very rare, and have never had an effect on ballot order, so far as I know.

That is not the only way my discretion enters the process, but the next step is also systematic: I break the all-time list down chronologically to generate a rank order of players for each decade, sometimes making judgment calls about the decade to which a player belongs or when to split a player between two decades, though I have formulas for guidance in each case. A player's rank order within his decade--his merit relative to his immediate peers--then becomes the basic criterion for his ballot placement.

Before I use rank order this way, however, I re-weight each player's rank order based on the competition-adjusted "quota" that I have set for HoMers for each decade, and I make the initial arrangement on my ballot based on the players' adjusted rank order within each decade. The quota is just an analytical tool, not decisive element in the rankings. It's basically a judgment of how many HoMers a decade would have, if great players were distributed evenly year to year, accounting for changes in population and access to the major leagues. It's basically a way to take some account of improving competition without having access to a really reliable direct measure of competition levels.

I then use my own discretion to modify this list, based on a variety of factors. If Player A from a decade with stronger competition happens to be farther down in rank order than Player B from a decade with weaker competition but nevertheless ranks higher in the all-time list, I will rank Player A higher in most cases. If my system has evaluated a player very differently from one or both of the comprehensive metrics and from the consensus, I may move the player up or down somewhat. If I think that my system is misjudging a group of players, I may shift them them somewhat in the rankings prior to finding a way to tweak the system to make such a shift automatic. No system is perfect.

We're at the point now where my system requires much discretion from me in ballot-construction, because most of the players in the backlog are very similar in decade-by-decade rank order and within the all-time list.

Players who are 1-3 spots above the quota-based in-out line for their decade include:

1880s: Mickey Welch
1890s: George Van Haltren, Herman Long
1900s: Rube Waddell
1910s: Spotswood Poles, Harry Hooper, Ben Taylor
1920s: Edd Roush, George Sisler, Buzz Arlett
1930s: none
1940s: Bobby Doerr, Bobo Newsom, Bucky Walters
1950s: Billy Pierce, Nelly Fox.

That's a whole ballot's worth of people whom my initial sorting mechanism finds to be very near to equal. Cross-era comparisons spread these players out between spots 11 and 48 in my actual rankings, but there's certainly a lot of discretion involved.

What I really rely on the "global score" element in the system to do is to properly weight career, prime, and peak value and to tell me how the player ranks relative to his immediate peers. It's in the cross-decade comparisons that my direct judgment comes into play.

Sievers not close.
1)Spahn--ERA+ not as good as I expected, but still he’s an easy choice. Gets some war credit too.
2)Fox--Not to sound like Karl Magnus, but the man had 2663 hits (#61 all time) and was a great fielder. If you’re going to take one of the second basemen he should be the one.

Regarding Fox vs. Schoendienst, I see them as having similar quality hitting and fielding, then career length/durability goes clearly to Fox. I guess the way Fox would end up behind Schoendienst is a league quality discount, I don't do that but maybe I should.

My system does not quite spit out a global score for each player... <u>[snips out several paragraphs of intricate and well reasoned ranking process]</u> ...It's in the cross-decade comparisons that my direct judgment comes into play.


Well, Chris, if you are going to be that simplistic about it, I don't even know why you bother to vote. :)

Ok, you guys, I'm thick skinned and simple minded, proud of it, and intend to remain that way. :-)

Silly question: if we love Fox so much, why aren't we all voting for Lave Cross? 2645 hits, only 18 less than Fox, even before you adjust (so about 2900 adjusted) OPS+ 100 instead of 93, 3B, which is equivalent to a 1950s 2B and played largely in the 1-league 1890s?

Actually, I do like Lave Cross more than most, he used to be on my ballot. The 100 OPS+ is padded by taking advantage of some weak leagues, and I don't believe 3b was ever the equivalent of a 1950s 2b.

39)Lave Cross—great fielder. Caught some too. Only hit well in weak leagues, but still that’s a lot of career value…2645 career hits with a lot of defensive value. All time leader in Win Shares / 1000 innings at 3B.

For example I don't think you'll find 1890 3b's with comparable careers to:

Don Blasingame, 79 OPS+, B- win shares fielder, 1444 games played
Julian Javier, 78 OPS+, B-, 1622
Bobby Richardson, 77 OPS+, C+, 1412

I think the closest we'd get might be
Arlie Latham , 92 OPS+, 1627 GP
Billy Shindle, 88 OPS+, 1422 GP
Joe Mulvey, 84 OPS+, 987 GP
Charlie Irwin, 82 OPS+, 989 GP
(Mulvey and Irwin adjust to about 1100-1200.)

(Don't have the WS fielding ratings handy.)

Got 'em.

Latham and Shindle are A's. Irwin a B+, and Mulvey a C. Mulvey's our man if they'd have played him another year. ;-)

I like Lave Cross too; he hasn't made my ballot, but he's always been in my top 100. However, a big difference in comparison to Fox is that Cross would generally miss about 20 games a season (even after he stopped catching), whereas Fox would miss maybe 3 or 4.

There are arguments for Cross over Fox: I don't think I buy them, but they're not unreasonable.

I believe I would take Cross over Schoendienst.

Looking at the Lave Cross's record, another thing I'm struck by is the depressing effects on his statistics of playing in the single league NL of 1892-1900 compared with the two-league environment that preceded and followed.

From 1892-1900 (ages 26-34) Cross's OPS+ was 96, while from 1901-1907 (ages 35-41) it was 106. Based on just this one player, it looks like the effect of the one-league environment on OPS+ might have been as large as 15-20 points (figuring that the normal expectation is a decline in OPS+ with age). Has anyone ever systematically looked at the statistics to try to figure out the effects of the 1892 contraction on player statistics?

This also could be relevant to evaluating other periods of contraction -- the contraction of the Negro Leagues in the 1930s and the contraction associated with integration in the 1950s. I believe that our methods are not adequately accounting for the effects these events had on player statistics, implying too little representation for the 1890s, 1950s, and possibly the Negro Leagues of the 1930s.

I'm not much of a timeliner; I probably like Latham better than I like Fox.

My system likes infielders, but Fox is held back a little compared with his WS totals by playing in the "wrong" league. Ample WWII credit puts Gordon and Doerr as the top 2 eligible 2B in my evaluation.

1971 prelim
1. Spahn
2. Schang
3. Sewell
4. Gordon
5. Doerr
6. Hodges
7. Keller
8. Cool Papa Bell
9. Mendez
10. Billy Pierce
11. Nellie Fox
12. Sam Rice
13. Mackey
14. Bridges
15. Schoendienst

My system likes infielders, but Fox is held back a little compared with his WS totals by playing in the "wrong" league.

Is your contention that integration and the contraction of the Negro Leagues caused the quality of play to drop in the AL? That doesn't make sense to me. I see it it increasing, just not as much as the NL.

Our second child is due any day now, so, if I am MIA on the ballot thread, please copy this one over:

(1) Warren Spahn (new)-Duh.

(2) Dobie Moore (2nd)--I started this project as a peak voter and quickly realized that there were just too many guys who were astounding for a year or two to make that a viable strategy. Since then, I have really focused in on a combination of career on the one hand and value over a roughly seven-year period on the other hand. On a seven (or six or eight or nine) year measure, Moore is the best player we have ignored.

(3) Oms (3rd)--The objective numbers back up the subjective comments--Sam Crawford with good but not remarkable longevity is probably the best comparison.

(4) Van Haltren (4th)--My position on him remains consistent: the small plusses add up. There are many position players who are quite similar to him in the next 20 spaces and he has a small but signficiant plus factor on each. Another words, he is one notch above borderline.

(5) Roush (6th)--Missed a lot of games due to injuries and holdouts or he would have had career stats that make him a low-interest no-brainer like Clarke, Wheat, or Goslin. Still, was a consistent top 10 (usually top 5) OPS+ guy who was also a solid CF. 135-140 games of that is awfully valuable.

(6) Beckley (7th)--If you do a major era adjustment, he's Palmiero without the steroid issue. If your adjustment is more minor, he's Gil Hodges, Steve Garvey, or Keith Hernandez with a record-long career. Either way, he's a bottom-quarter HoMer.

(7) Duffy (8th)--I need to do a study of why his offesnive WS are so high throughout the early 1890s. Until I do that, I will split the difference between WS assessment of his offensive ability and other metrics. When I do the study, he may move to the top of the ballot or off ballot.

(8) Childs (9th)--Now come the solid 2B with the bats of good OF's. They are both underrated.

(9) Gordon (10th)--Like I just said.

(10) Mendez (12th)--Looks an awful lot like the Lemons, Coveleskis, and Vances of the world.


(11) Trouppe (11th)--I wish our info on him was better, but my consistent policy has been to evaluate the excluded players as best as I can and then vote that evaluation without a deduction for their larger confidence interval. Similar to Frank Grant in that there is a chance he wasn't one of the top 1000 players of All-Time but that there is a better chance that he was one of the top 200.

(12) Sisler (13th)--Had Charlie Keller's career, then another one that was of some small but genuine value. Might be a bit overrated here but the bunching is extraordinary.

(13) Minoso (off)--On further review, he looks at least as good as Medwick who would rank here.

(14) Elliot (15th)--Elliot, Doerr, and Sewall are essentially tied on my ballot, but I'm letting position scarcity break the tie.

(15) Willard Brown (off)--Moves up with the new data saying he would take a walk if it was waved in hi face.

See above for why I don't like Fox and previous ballots for the other perennials. On further review, Pierce is more like number 19 than number 14.

OPS+ game
400+ PA seasons, at least 100 OPS+, 90+ in parentheses

NelFox 124 17 14 07 05 (99 97 94 92 91)
BDoerr 165 31 28 28 17 16 16 16* 14 05 03
Gordon 155 35 35 26 24 21 20* 17 10* 08 03 (98)

Doerr * for a war-year credit 116 - and the 165 comes in weak 1944
Gordon *s for war-year credit of 120 and 110


I just don't see how Nellie beats these guys out, with a bunch of extra sub-90 OPS+s. At that point, he's not helping all that much.

While I agree that Fox was not as good as either Doerr or Gordon, remember that Fox's OPS+ are low because of his low slugging. If he was the other way around (good SLG and low OBP) those OPS+'s would look at little bit better.

Point granted, schmeagol.
But adding 10 to each season (and isn't that a little much of an adjustment?) still seems to leave him short.

Actually, that's a good question, since this is an extreme case. What IS a fair upward adjustment for Fox here?

Not much.

Fox lifetime OPS+ = 93.9, based on OBA and SLG compared to lg&park; avgs of 348/339 and 363/398.

If one pt of OBA is worth 1.8 times as much as SLG (many would say that's a bit of exaggeration in Nellie's favor), change Nellie's stats from 348-363 to 333-390. Now his lifetime OPS+ would be 100*(333/339 + 390/398 - 1)= 96.2. So you can add 2 to his OPS+ totals.

I don't think that does him justice either.

He's got 4.55 RC/27 in the All-Time HB, vs. 4.39 for his leagues. That shows him as an above average offensive player.

But wait, OPS+ removes the pitchers from league averages - could that be the cause of the discrepancy? He wasn't a great base-stealer, so that's not it.

The ATHB uses different park factors too, that may be a part of it.

He's got a ton of fielding black in, wow.

Led league in putouts 1952-61. Every year. For a decade.

Led league in assists 1952, 1955-57, 1959-60.

Led league in DPs 1954, 1956-58, 1960.

Led league in FPct 1952, 1954, 1956, 1959, 1962-63. Pretty amazing considering how many chances he was handling.

Led league in 2B games, 1952-59, which could be why he's got the black ink, except that . . .

Led league in Range ((PO+A)/G) 1951, 1953, 1955-57, 1960.

I've got to break this book out more often.

By the way, as I look ahead to next year's ballot, I expect that very few players will create as mcuh hand-wringing and soul-searching and plain ol' arguin as Mr. Koufax. As a career-centered voter, I'm already arguing with myself over what is "obvious" (Sandy was positively great) versus what the numbers say. I think it's poetic that he comes on the ballot the same year as Robin Roberts, forcing us into a head-on analysis of another guy with a great peak and more career. I note that Koufax rates above Roberts in the NBJHA, even though the ##s appear to point toward Robin - could be one o' them fudge factors :)

I say this NOT to encourage early analysis and arguin, but to say we each ought to gear up for a fun food fight scheduled to open at a discussion board near you next Tuesday.

Koufax #1, Roberts #2-3-4?

But wait, OPS+ removes the pitchers from league averages - could that be the cause of the discrepancy? He wasn't a great base-stealer, so that's not it.

I think so. The Sabermetric Encyclopedia lists his OWP as being below 0.500.

On the other hand, Baseball Prospectus lists his EQA at slightly above .260 (.261). I'm not sure if they remove pitchers batting or not.

The conclusion I draw from the discussion of the differing representations of Fox's career offensive rates is that, in the upside interpretation, he was slightly above average, and in the downside interpretation, he was slightly below average. Regardless of his ambiguous placement relative to an overall league average, we know pretty conclusively that he was an above-average hitter for his position, but that he is not nearly the best hitter eligible at his position.

What I take issue with is an idea that league-average offensive production is a necessary benchmark for a player at a high-defense position to be HoM-worthy. "He was a BELOW AVERAGE hitter?!? We can't have THAT in the HoM!" Not that anyone has put the matter in precisely those terms, but I think that there is too much emphasis on "career league average hitting" in the consideration of the high-defensive value players.

That said, Fox will probably not make it on to my ballot this year, though he will be close. I am arguing back and forth with myself on his merits and Doerr's merits, win shares versus WARP, better hitting versus more durability, weaker league versus pre-integration league.

It's a tough call.

What I take issue with is an idea that league-average offensive production is a necessary benchmark for a player at a high-defense position to be HoM-worthy. "He was a BELOW AVERAGE hitter?!? We can't have THAT in the HoM!" Not that anyone has put the matter in precisely those terms, but I think that there is too much emphasis on "career league average hitting" in the consideration of the high-defensive value players.

For what it's worth, in my adjusted RCAA system, Fox comes out at about +7 wins above average offensively over his career - hence a little above average. But I'm not removing pitchers, so that's a little optimistic. If I isolate out a 10-year stretch in the middle of his career, he's +16 over those 10 years, with a peak year of +4.5 in 1957. Of course, some of our other second basemen on this scale: Lazzeri +29, Childs +28, Gordon +27, Doerr +22. But I am willing to call Fox an above average hitter, especially in the middle of his career. Fox's career rate stats are dragged down by his early and late career, in which he was less of a hitter but presumably still had defensive value.

This is a run-up to talking about Ozzie Smith. Of course, Ozzie's career is even longer and his defensive value higher. I have already run Ozzie through this system, and I get +1 win-valued RCAA for his career (the same scale on which Fox is +7.) If I do the same as I did for Fox, I get a 10-year stretch in the middle of his career in which he's +13, with a high point of +3.6 in 1987. (And that 10 years of Ozzie's career compares well to all of Rizzuto's career.)

A question addressed to any voter(s) who wants to answer it:

For those of you who constantly tinker with your systems, or who have periodically overhauled it, how much do you think your PHOM would change if you had employed your current system from the start of the project? Have you elected anyone to your PHOM whom you do not think would get in now?

I tinker pretty much every year but I don't think it affects much for two reasons.

1. People aren't moving much in either direction. The most I've moved anyone up is 10 slots. Most moves are 5 slots or less.

2. The people who move the most tend to be deeper in the backlog where a slight change means you move 10 places. I've never moved a guy off ballot to #1.

I may hypothetically alter the last 10 to 20 slots in my PHOM but it should remain about 95% the same.

For those of you who constantly tinker with your systems, or who have periodically overhauled it, how much do you think your PHOM would change if you had employed your current system from the start of the project? Have you elected anyone to your PHOM whom you do not think would get in now?

As someone who has overhauled his system numerous times, I don't bother with a PHoM because of the many upheavals. Too embarrassing, in a few instances (like Tip O'Neill :-( ).

When I tinker with my system, I go back to the candidates for the first election and apply it to everyone who has a legitimate chance of being in my consideration set.

It's a pain, but it's kind of fun, particularly if the changes were significant or represent a breakthrough (in my own mind).

I haven't had to do it more than about 10 times. LOL

What DL said.

Biggest exception is Pud Galvin, who I had championed for some time and then back-tracked on, after we had elected him. He's now at the bottom of my rankings on our HoM list, although I can see with our bunch of 'elect 3' years upcoming that I would have eventually advocated for his election anyway, albeit 100 years later.

Two ballot procedure questions (I've lost track of what the answers are):

1. Are we still required to include 15-20 on the ballot?
2. Are we still required to mention something about the people who made the top 10 consensus but are not on our ballot?

I'm not playing ballot cop here. It's just that I've seen some ballots without those disclosures, and since I voted against those amendments anyway (b/c I like shorter ballots), I'd like to remove them from my ballot if they are no longer mandatory.

If they are still required, then I'll keep everything the same. :)