bump!

Not that anyone cares, but here's Keller's stats in 1937+1938 in AAA at Newark:

Keller didn't make his major league debut until May 23rd in 1939. That partially explains the low game total in 1939. Despite the loss of Gehrig, the Yankees were sitting at 21-5 going into that game.

Keller/Selkirk/Henrich were all LHB's, so there was no straightforward platooning amongst the three of them that was going on that would somewhat inflate their rate stats like it would later do for a guy like Woodling.

Oh, we've discussed Keller's Newark MLE's at length in his thread. Maybe I should go bump it.

Hmmm... the Newark MLE's are only mentioned in passing there. I know there was a long discussion somewhere. Its not an easy call. Some people were completely sold on 1938 MLE credit and others were completely against it.

One wonders if Keller mightn't have improved the Yankee OF in 1937, not that anybody could have cared (they won 102 games, scored a league-leading 979 runs and won the pennant by 13). Still DiMaggio was DiMaggio. But:

LF--Jake Powell 94 games (otherwise, ill) 3-45-.263; Henrich 30 games; Hoag 24 games
RF--Myril Hoag 70 games, 3-46-.301, 33 BB, 48 runs; Selkirk 69 games (otherwise, injured) 18-68-.328; Henrich 29 games
Henrich (otherwise injured) 8-42-.320

And in 1938, it was 99 wins, a league leading 966 runs and another 13 game margin.

LF--Selkirk 95 games 10-62-.254 (don't know why the 95 games, maybe just a platoon?); Powell 37 games -20-.256; Hoag 31 games
RF--Henrich 130 games 22-91-.270; Hoag 27 games
Hoag 0-48-.277

So was Keller held back unreasonably? Well, Henrich was a 27 year old rookie in 1937 and Selkirk was a 26 year old rookie in 1934. DiMag, OTOH, was a 21 year old rookie in 1936.

FREE TOMMY HENRICH!!!

Some people were completely sold on 1938 MLE credit and others were completely against it.


I can see the argument from both sides. On the one hand, Keller was only 21 in 1938; lots of guys are stuck in the minors at 21.

On the other hand, he was clearly ready at age 20, was blocked behind 3 all-star caliber OF's, and was playing for an organization with the greatest concentration of talent in MLB history.

Seriously, lets take a moment to ponder how ridiculous the Yankees organization was in 1937 and 1938; the big league club was winning games at a .660 clip, and AAA Newark was winning games at a .700 clip. Basically, they had enough talent in the organization to fill two major league teams. Freakin' incredible.

Basically, they had enough talent in the organization to fill two major league teams.

Were they outbidding the other 16 teams for amateuer talent? Was it just the stars aligning or more of their guys panning out? Makes you understand exactly why Branch Rickey was buying up farm teams. If you can't outspend NY, then you got to see the talent first and lock it down earlier and cheaper.

More than 200 messages and Chet Lemon is still under ten mentions, half them concerning whether he gets his own thread. That's quite a record for someone with his sabrmetric ratings, and number two on the big board up above.

It seems we have no consensus about CF ratings. Chris Cobb's remark is new to me, that WARP rates CF higher than WiSh does (or is that only CF whose defense is highly rated?). Wynn, Lynn, Lemynn. Not quite peas in the pod with, let me say, Ryynn and Van Haltrynn. Go Edd.

I prefer Ynnms ;-)

>On the other hand, he was clearly ready at age 20, was blocked behind 3 all-star caliber OF's, and was playing for an organization with the greatest concentration of talent in MLB history.

'Zop, didja see #206? Jake Powell and Myril Hoag played in 352 games those 2 years. Selkirk played a total of 164. Those Yankees were good but their corner OF positions were absolutely not performing at all-star level those 2 years. Not even Henrich, not even when healthy.

Doc, the Yankees clearly had a big advantage in scouting and money (brains and bucks). They just locked up a lot of good young guys.

Here come the two most fragmented elections back to back and everyone ran out of things to say? I've been looking at the top 10 returnees and only three are going to be on my ballot. The person I'm putting #1 on this year's ballot was 28th in last year's balloting. I'd like to hear the case against Luis Tiant that doesn't involve quotas (only the nth best pitcher of his era, etc).

The only pitchers I see as locks upcoming are Ryan, Blyleven, Gossage, Eckersley. I think Tiant's career has significantly more merit than Tommy John, Rollie Fingers, Dave Stieb, Jack Morris, Bret Saberhagen, Frank Tanana, Dennis Martinez, etc. We're having 12 more elections, are we only going to elect 11% pitchers from here on out in the era of the 12 man pitching staff? As an aside I ran Gossage and he only ended up barely on the ballot. I don't think that's right by a long shot so I'm giving relievers a boost. This boost will get Fingers up to my top 30 from around 50.

Clemens, Martinez, Maddux, Johnson, Rivera, Hoffman, Brown, Mussina, Glavine, Smoltz, etc.

There's lots of pitchers from the 12 man staff era coming up.

None of them eligible until 2011 at the earliest. I think David Cone is a good choice come 2009 and Lee Smith is as good of choice as Fingers. That's still only 5-8 pitchers between 1995 and 2009 out of 45 electees or 11-18%. I foresee 10-12 more between 2010 and 2020 but that's only 30-33% (pretty close to the correct ratio). Historically we're at 28% and trending downward.

DL,
I must admit that I don't see anything special about Tiant. Can you give me the argument FOR him first. It is probably a better way to proceed than have those of us who have him solidly off ballot (He is in the 30's for me) to give an argument against.

Prelim

PHOM battle is currently between Hernandez, Oms, Willard Brown, and Pierce.

1. Keller - The only guy currently eligible that I feel very strongly about, I really hope that this is the year
2. Duffy - WS may overrate his peak a bit, but I am willing to give him some credit for his teams' performances.
3. Redding
4. Walters
5. Wynn
6. Elston Howard - Very similar, though just behind, the recently elected Quincey Trouppe
7. Pete Browning
8. Gavvy Cravath
9. Dizzy Dean
10. Rollie Fingers
11. Keith Hernandez - Only newbie to make my ballot. This is roughly where Mattingly will end up and Sisler would end up as well, it is also the edge of my current PHOM.
12. Alejandro Oms
13. Vic Willis
14. Larry Doyle - 1st time on ballot! He has pushed ahead of Nellie Fox with Dan R.'s arguments against Fox. I do have worries about competition, however.
15. Roger Bresnahan - Back after many, many years off my ballot. Best MLB catcher from 1900-1920 with a nice peak 9even after adjusting for time in CF).

16-20 - George Van Haltren, Urban Shocker, John McGraw, Edd Roush, Al Rosen

Lynn and Quiz are interesting Newbies who could make my top 50. Lynn has a very nice two eyra peak but injuries stopped him from being as good as Jimmy Wynn. I like Quisenberry's peak as well, he may be the first reliever to fail to reach my ballot but still end up in the top 50.

7 seasons in top 10 in adjusted ERA+ plus a 127 in 1964 and 123 in 1966 that weren't top 10. 6 additional above average seasons for a total of 15 above average. Similar PRAR and PRAA as Billy Pierce, Juan Marichal, Don Drysdale, Jim Bunning. 3-0 with 2.86 ERA in 5 postseason appearances (34-2/3 IP). 35th career in strikeouts, 49th in shutouts. 6 top 10 finishes in complete games.

Ryan, Blyleven, Gossage, Eckersley. I think Tiant's career has significantly more merit than Tommy John, Rollie Fingers, Dave Stieb, Jack Morris, Bret Saberhagen, Frank Tanana, Dennis Martinez, etc.

It's not an inspiring lot. I'll take Bly, Steib, and Ryan for sure. I might take Gossage, but I'm very unsure of any reliever at this point. It's unlikely I will vote for Eck or any of the other guys mentioned here.

Clemens, Martinez, Maddux, Johnson, Rivera, Hoffman, Brown, Mussina, Glavine, Smoltz, etc.

It's possible that Kevin Brown will really sneak up on people. Dude was pretty awesome.

On the whole, it feels to me as though the 1980s are to pitchers what the deadball era was to 1Bs. Lots of guys who are either HOVG or just good, and those who are great seem to end up hurt or petering out for some reason. Just a lot of "that's life," just like for Chance and friends.

On the whole, it feels to me as though the 1980s are to pitchers what the deadball era was to 1Bs. Lots of guys who are either HOVG or just good, and those who are great seem to end up hurt or petering out for some reason. Just a lot of "that's life," just like for Chance and friends.

You might be right, Eric, though both groups are still underrated, IMO.

I must admit that I don't see anything special about Tiant. Can you give me the argument FOR him first.

His Ballpark Franks commercials with the Yanks were, by far, the greatest ever recorded.

'Nuff said.

If you like Tiant what about Walters? Similar IP and ERA+. Similar win shares, but Walters has a better peak.

Top 10 in ERA+ 6 times (led 2) for Walters, top 10 7 times (led 2) for Tiant.

Top 10 in IP 8 times (led 3) for Walters, top 10 3 times (led 0) for Tiant.

Top 10 in CG 8 times (led 3) for Walters, top 10 6 times (led 0) for Tiant.

Top 10 in SO 9 times (led 1) for Walters, top 10 7 times (led 3) for Tiant.

Top 10 in GS 9 times (led 2) for Walters, top 10 2 times (led 0) for Tiant.

Throwing in a couple of contenders
BWalters 168 52 46 40 27 23 07
LuiTiant 184 69 32 28 25 20 19 05 02 02 00
BuGrimes 153 44 38 36 31 23 08 08 08 03
TommJohn 154 38 38 37 25 20 19 19 16 14 11 10 09 09 06 03 00
Reuschel 158 57 32 31 19 17 16 16 14 11 05 03

BWalters top 10 in IP: 1 1 1 4 6 6 8 8
LuiTiant top 10 in IP: 6 7 8
BuGrimes top 10 in IP: 1 1 1 3 3 4 7 9 9 9
TommJohn top 10 in IP: 2 5 8 10
Reuschel top 10 in IP: 4 7 7 7 8 9

Tiant's top 2 years look terrific at 1st, but note the terribly weak top-10 IP numbers. We usually elect guys with 4-5 yrs as a contender for best pitcher plus some supporting stats, or longtime workhorses (Grimes is a weak version of the breed).
Tiant was 7th in IP in his great 1968 season (184), but he only threw 179 IP in his 169 season.

So really you only have one season where he's a leader in both ERA+ and IP.
In Walters' two best ERA+ years, he also led the league in IP. Wow.
That's enough to give Walters the edge in best-6 seasons. Tiant may have a better 7th-best season, but after that he just has a couple of league-average seasons without a lot of IP. Pretty irrelevant, mostly.

Tiant may make my top 25, but I'm not knocking myself out to go to bat for him. Reuschel also was 7th in his best season and non-top 10 in IP in his 2nd-best.

You may want to show me how Tiant blows Reuschel out of the water, career-wise.

Here's my RA+ Pythpat year-by-year records for Tiant, Walters, and Reuschel. The ones for Walters have been adjusted for his offense, which is significant. None have been adjusted for defensive support, which is likely significant in the case of Reuschel. The years have been sorted from best to worst by equivalent FWP. The third number after the W-L record is the equivalent single-year FWP.

Walters' 1940 was totally a product of his fielders. The Reds' defensive efficiency that year was nearly 2 standard deviations above the NL average. That said, none of these guys are serious contenders in my book. The only pitcher on my ballot these days is Cicotte, and he's way down there.

"totally a product of his fielders" seems like a stretch, but I'll look into it. I barely vote for Walters anyway.
And what if Walters was inducing hitters to hit grounders/fly balls/whatever works? Obviously no pitcher can consistently make a guy hit it to his 2B, for instance, but with a team full of vacuum cleaners in the infield, inducing grounders might be possible - and sensible.

> a couple of league-average seasons without a lot of IP. Pretty irrelevant, mostly.

I think league average pitching seasons are being underrated here.

If "nobody" is worth voting for, perhaps we're holding pitchers to an unreasonably high standard.

Being the anti-establishmentarian that I am, I still maintain that Lefty Gomez and Tommy Bridges are very ballot worthy pitchers.

Howie, Walters also had exceedingly low BABIP relative to team in '41--but then went the other way and was *worse* than his teammates in '41 and '43 (he was average for his team in '42). I should have said that his freakishly low .227 BABIP in 1940 was the product of phenomenal defense *and* great luck. He didn't have a consistent ability even in his peak period to outperform his teammates in BABIP, and his K/BB/HR were pedestrian.

Someday we may even have Bucky Walters' pbp, which would be very, very cool. I'd love to know if he was inducing tons of grounders or what.

I have to say that I have no problem giving a pitcher credit for a low BABIP. It is not a great sign going forward, but we are not going forward and such a season provides great value to a team.

Now as for how much of that is defense, I don't know (I am willing to take away some credit because of that).

Now as for how much of that is defense, I don't know (I am willing to take away some credit because of that).

Is it different than the Cravath/Klein/Rice/Ott question?

Walters was a sinker-ball specialist. It isn't uncommon for a sinker ball pitcher to have better results one year than he does another and it has little to do with luck. It has everything to do with good defense and how good the sink is that particular year.

I'm a lot bigger fan of Bridges than I am of Walters.

I'm rethinking Walters though. He's being adjusted down 22 PRAA for defense. Some of that value is because Walters was a really good defender though. I'm going to give him the benefit of using his defense to his advantage. That's pitching. Still, he only gets up to 68th. He had a really good bat and a really good glove for a pitcher which will enhance his peak value. His pitching career is only medium length, 3100 innings, due to the late switch to pitching from 3B. He deserves no war credit and should probably get a bit of a war discount. If you're looking for peak I like Dean and Redding better.

it has little to do with luck. It has everything to do with good defense and how good the sink is that particular year.

In my opinion, it has a lot to do with luck. If you let the other team hit the ball, sometimes those balls will find holes and sometimes they won't - there's a limit to how much you can control it. That's why strikeout pitchers are more consistent - there's no chance for a hit if there's no contact.

Strikeouts have nothing to do with luck? Whether or not the batter makes contact has as much to do with luck as whether he makes good contact.

Sinkerball pitchers are able to induce weak contact but most aren't able to do it year after year.

Note that Walters allowed over 15% of unearned runs, while Tiant was at about 8.5%

Strikeouts have nothing to do with luck? Whether or not the batter makes contact has as much to do with luck as whether he makes good contact.

Sinkerball pitchers are able to induce weak contact but most aren't able to do it year after year.

Do you have any evidence of this? I've never seen any.

Let's say there are three possible outcomes - a batter hits the ball hard or in the air, a batter hits the ball on the ground softly, and a batter doesn't hit the ball at all. Let's assume that one pitcher has an ability to get the opposing batters to miss the ball (a strikeout pitcher), and another has the ability to get the opposing player to hit the ball on the ground (a sinker ball pitcher).

When the strikeout pitcher succeeds what is the opposing batter's chance of getting a hit? 0%. What is the margin of error? 0. Once a pitcher succeeds in getting the opposing batter to swing and miss, how much luck is involved in terms of whether the batter still gets a hit? None.

When a sinker ball pitcher succeeds in getting the opposing batter to hit the ball on the ground, what is the opposing batter's chance of getting a hit? Greater than 0%. What is the margin of error? Greater than 0. Once a pitcher succeeds in getting the opposing batter to swing and miss, how much luck is involved in terms of whether the batter still gets a hit? Some.

So while the ability of a sinker ball pitcher to get the opposing batter to hit the ball on the ground may be no more dependent on luck than a strike out pitcher getting the opposing batter to swing and miss, there is substantially more luck involved in whether the groundball leads to a hit than when the swing and a miss leads to a hit.

Ummmmmm sinkerballers have a HIGHER BABIP allowed than flyballers, guys. It's just that because those hits are singles rather than doubles and triples, and because they get double plays and don't allow sac flies, the extra hits they allow come out in the wash.

I understand the point you're making but there IS luck in getting the batter to swing and miss (or as I prefer, externalities the pitcher the cannot control). There is also lots of data showing that some pitchers repeatably induce flyballs v groundballs. Major league pitchers in general have a good ability induce weak contact more than those who don't stick in the majors. There isn't a lot of standard deviation between major league pitchers in this ability and the external noise dominates the numbers but there are some consistent outliers.

If the consequences for a batter aren't so different between striking out and a batted ball out, why is it so much better to have a strikeout pitcher instead of a pitcher who gets batted ball outs?

If the consequences for a batter aren't so different between striking out and a batted ball out, why is it so much better to have a strikeout pitcher instead of a pitcher who gets batted ball outs?

Ah... the Batter/Pitcher Strikeout paradox! :-) One of the gurus of sabermetrics should write an eloquent essay on this. (Bill James?, Tango?, someone else?)

My impression was that part of it had to do with splitting credit for BIP outs between defense & pitcher where for a K out it was almost all pitcher.

Also, much of the perceived value of a high-K pitcher is not current value, but predictiion of future value. I think high-K pitchers age better. That predictive part doesn't really affect what we're doing here in the HOM though. All the guys we are considering here are retired.

Still, its an interesting paradox and I'd love to see a nice essay explaining it. :-)

His Ballpark Franks commercials with the Yanks were, by far, the greatest ever recorded.

It's great to be a wiener!

Sigh. Just not the same without the accent...

It's great to be a wiener!

Sigh. Just not the same without the accent...

lol

I think it is similar to Hugh Duffy's Win Shares. If you basically believe in DIPS (or, more precisely, believe that the signal-to-noise ratio on batting average in balls in play for pitchers is so small as to be generally better off ignored), then seasons with flukishly low BABIP's are difficult to value: you don't want to give credit for the "missing" hits on balls in play to the pitcher and you can't give them to the fielders because other pitchers on the staff had much higher BABIP's. So, like the "extra" runs scored by the 1890's Boston teams, you either pretend that value never existed (my approach), or you have to devise a system to give credit for it (James divvies it up over a whole team's performance; you'd have to give part to the pitcher and part to the fielders).

I asked John Holway a couple questions about Bus Clarkson and Luke Easter. Here's his response:

"In 1947 Easter hit .339 with ten homers, which tied him for fourth in
the
league. In '48 he hit .363, tied for third place; i have no home run
data.
I dont have any numbers on Clarkson and dont consider him an important
player.

How can I become a member of your group?

john"

How can I become a member of your group?

john"

He wants to join us? Wow. That's pretty cool. I don't think any of us have said a bad thing about him either, which doesn't hurt. :-)

If you can, DL, just send him a hyperlink to this thread so he can introduce himself to our group. Joe or I can handle it from here. Of course, if you want to fill him in on the details, that would be great. Thanks for bringing this to our attention.

I told him to set up an account on btf and gave him a link to this thread.

Thanks, DL!

ALERT: This does not sound to me like an authoritative analysis.

>I dont have any numbers on Clarkson and dont consider him an important
player.

I'm still trying to help him get registered but he indicated that he views Dick Lundy as our biggest omission among Negro Leaguers.

I'm still trying to help him get registered but he indicated that he views Dick Lundy as our biggest omission among Negro Leaguers.

I think it's Oms, though Lundy is close to the end of my ballot.

Every now and then I realize we've overlooked a great or near-great player.

Connie Marrero became eligible in 1960. My compliments to Dan for including him on the list of players, even though he had only 44 win shares. I believe the only comment on the discussion thread came from Dr. C, who asked "Was there a rest of his career somewhere south of the border?" The fact that Marrero's rookie season with the Washington Senators came at age 38 should have been a tipoff that indeed there was.

I've been putting together some information over at Wikipedia. Marrero had a long and illustrious career in Cuba, running from about 1938 to 1957. Unfortunately, he has the Luke Easter problem--close to half his career (everything before 1946 or age 35) took place in the Cuban Amateur League for which we don't have statistics.

A few points:
- As a major leaguer from 1950 to 1954, ages 38 to 42 (or, according to other sources, 39 to 43) Marrero was pretty good. His career ERA+ of 108 actually underestimates his pitching because he received poor defensive support from the Senators.
- His professional Cuban League record, which began in the fall of 1946 and overlapped with his major league career, indicates that he was even better during 1946-49 (ages 35 to 38) than he was as a major leaguer. He spent the summers from 1947-49 dominating the Florida International League, which seems like a waste--except that playing for the Senators during that period perhaps would have been an even bigger waste!
- The big mystery is how good he was from 1938 to 1945 as an amateur.

A little bit of background--Cuba had an "amateur" league that played baseball during the summers while the professionals were playing in the American major, minor, or Negro leagues. Although the league was nominally "amateur," the teams were sponsored by commercial enterprises that would provide comfortable jobs for the better players (sort of like how Cuba's current "amateur" system works under Castro). The league was quite popular and according to Cuban baseball historians, the quality of the amateur league sometimes rivaled that of the professional winter league. (There were a couple of famous series where the amateur champions beat the professional champions.) Furthermore, the amateurs represented Cuba in the Baseball World Cup (also known as the Amateur World Series) competitions, which brought a lot of attention to Marrero and made him a national hero well before he went pro. Until the late 1930s, the amateur league was all-white; the government forced it to desegregate when they began participating in World Cup competitions in 1939. Ironically, because many of the best white Cuban players remained amateurs, their statistical records are often sketchier than those of the black players.

I'm not going to try to argue that Marrero is a HoMer. He's one of those players--like Luke Easter or Perucho Cepeda or Carlos Moran, about whom we just don't have enough information to determine whether he had an HoM-quality career. But it's interesting that everything I've learned about him suggests that he might have been that good. He should be right up there with Dihigo, Mendez, and Luque--one of the top Cuban pitchers.

If the consequences for a batter aren't so different between striking out and a batted ball out, why is it so much better to have a strikeout pitcher instead of a pitcher who gets batted ball outs?

Does anyone here directly credit pitchers for strikeouts? Using widely available pitching data on runs and earned runs we can easily look up whether Jack Chesbro or Tommy John yielded unearned runs at a high rate, so that ERA and other statistics derived from ER are distorted in his favor, and how much distorted. We have measures such as RA+ (by OCF?) that count runs instead of unearned runs. But we no comprehensive and no widely available batting data for first base on errors. We know about Bobby Bonds strikeouts and can easily check their precise number but we can only infer that Bonds reached base on error less often than average and only guess wildly(*) how much OPS+ and other statistics derived from Hits are distorted in his favor.

(*For Bonds and other batters in the Retrosheet era, some of us can do much better than guess wildly, but not so well as almost anyone can do with earned and unearned runs for John and other pitchers.)

I'm not going to try to argue that Marrero is a HoMer. He's one of those players--like Luke Easter or Perucho Cepeda or Carlos Moran, about whom we just don't have enough information to determine whether he had an HoM-quality career. But it's interesting that everything I've learned about him suggests that he might have been that good. He should be right up there with Dihigo, Mendez, and Luque--one of the top Cuban pitchers.

Brent, I'll set up threads for Marrero and Moran (who never received one, either) sometime this weekend.Thanks for bringing their potential HoM cases to our attention.

John I got your E-MAILs I'll deal with it as soon as posisble

if it's not to much trouble at a conveinent time can you
on the important links thread put all are HoF voting threads in a seprate HoF thread to shorten the list

Just throwing this out there because I did the work. I updated through 1990 my ranking of 1Bmen. This is where I take the Win Shares for the #1 first baseman for each team, find the average for each year, and how much each player is above/below the average. Then add it all up. I started with 1888 (way beck when, this was Beckley-related), and here are HoMers, candidates, and other players of interest through 1990. The number in parentheses is their ranking if you only included the "positive years" - more than 2.5 WS above the average.

1. Lou Gehrig 235.91 (1)
2. Jimmie Foxx 146.09 (2)
3. Johnny Mize 116.43 (3)
4. Eddie Murray 115.88 (4)
5. Keith Hernandez 90.67 (6)
6. Willie McCovey 86.62 (5)
7. Dan Brouthers 82.58 (8)
8. Bill Terry 82.04 (11)
9. Orlando Cepeda 81.41 (13)
10. Gil Hodges 79.7 (10)
11. Stan Musial 77.38 (9)
12. Frank Chance 76.28 (15)
13. Roger Connor 72.79 (17)
14. Hank Greenberg 70.54 (14)
15. Norm Cash 69.9 (19)
16. Jake Beckley 67.82 (16)
17. Ed Konetchy 66.71 (12)
18. George Sisler 58.17 (7)
19. Harry Davis 56.9 (22)
20. Don Mattingly 55.49 (20)
21. Will Clark 53.91 (23)

25. Mickey Vernon 43.94 (18)
26. Cecil Cooper 42.06 (28)
27. Steve Garvey 41.08 (26)

40. Boog Powell 25.85 (29)

76. Luke Easter 11.14 (112)

117. George Kelly 4.75 (58)

374. Bill Buckner -12.09 (176)

473. Ed Kranepool -52.62 (NA)
474. Phil Todt - 54.26 (NA) (Ironically, Todt and Gehrig are on adjacent lines in the database, which was entered in the order of the Win Shares book)

And just a couple of notes, the average of the averages for Gil Hodges' 12 years was 14.36. The average for Hernandez' 12 non-strike years was 15.62. Hodges' era was probably the weakest since the deadball era. The strongest was the 30s, no surprise there.

And here's the top 6 individual seasons in the period:

Lou Gehrig 1927 27.81
Stan Musial 1946 27.12
Jimmie Foxx 1933 23.87
Will Clark 1989 23.11
Jimmie Foxx 1932 22.94
Norm Cash 1961 22.74

I finally completed my double checking of all candidates in our collective top 20. Essentially, I looked for differences in my positional rankings between guys on my ballot (either in the consensus top 20 or not) and the guys I have ranked lower.
In general, my initial rankings didn't change too much here. Tony Perez and Charlie Keller each went up a handful of spots but neither nearly enough to challenge for ballot spaces.
However, I have one question that remains unresolved. I may be able to answer it myself, but I am short on research time right now.
Alejandro Oms: I have him listed as a RF but I recall that some remark on his CF play. Does anyone know offhand if his positional splits were discussed in his thread, or if there is any consensus thought on how his time was divided?

if it's not to much trouble at a conveinent time can you
on the important links thread put all are HoF voting threads in a seprate HoF thread to shorten the list

Good suggestion, yest. It's already done.

Since Holway hasn't logged in I'll quote his e-mail:

"Clarkson is a nobody. Lundy is one of the Big Three shortstops. Hit
30 points higher than Wells. Took (an older) Lloyd's job away from him.
Won three flags as a manager. Considered the smoothest fielding shortstop.
A travesty that he didnt make the Hall of Fame. I'd have named him in the
first five on the latest ballot."

There are two HoM arguments for Lundy:

(1) We are underrating the Bancroft-Rizutto-Long excellent glove, ok bat SS's, and Lundy belongs at the top of the list of those guys.

(2) We are underrating Lundy's hitting by a rung or two and he really would have been a 105 OPS+ guy in the bigs which with his fielding and his career length would make him an obvious HoMer.

I have no particularly reason to believe either argument is true, but either or both certainly could be.

I do have Dick Lundy ahead of Nellie Fox...

I guess this is what we have to answer:

06) Nellie Fox 348pts
34) Phil Rizzuto 109pts
39) Dave Concepcion 99pts
42) Dave Bancroft 72pts
50) Vern Stephens 51pts
62) Fred Dunlap 62pts
72) Luis Aparicio 24pts
76) Rabbit Maranville 23pts
82) Herman Long 16pts

Dick Lundy - no votes

One of the things that really hurt Lundy's MLEs in my original study of him was his _really_ anemic walk totals, according to the best data we had at that time. I have not re-done MLEs using the HOF data, partly due to lack of time in my life for that work and partly because the borderline HoM NeLers have had their cases depend more on MxL and minor league data.

A quick trip over to the HoM site to check Lundy's stats suggests that his walk totals during his prime may be higher than the old data indicated. I'll have to check that when I am at home and have access to my MLE spreadsheets. He's no walking machine, but a revised view of his OBP might help his OPS+ significantly. Given how badly his hitting trailed off in the mid-1930s, though, I have a hard time seeing him sustaining a major-league career of more than 13-14 seasons, and it seems unlikely that he was a better hitting/fielding package than Dave Bancroft, who is just hanging on in our balloting. We are overrating Fox, but if Lundy's comparables are the Rizzuto/Concepcion, Bancroft/Aparicio/Maranville/Long group, then it seems unlikely that our undervaluing of Lundy (if there is any), has been so great as to keep him from deserved election to the HoM.

I'll post again on this matter once I have had a chance to look at the data I used for Lundy's MLEs.

Lundy is close enough to the end of my ballot that a revised MLE for him could push him onto my ballot.

More on Lundy:

Lundy's walk rate presented in the HoF data is significantly higher than the rate used in his original MLE projections, which was based off of limited data from 2 seasons.

In the MLE, Lundy's walk rate was .0483 BB/PA for his career. That is a rate adjusted upward due to the low BB environment of the NeLs in comparison to the majors.

In the HoF data, Lunday career walk rate is .0818 BB/PA for his career. That rate should be adjusted upward to approximate a major-league rate. That's a very significant change, and a change that is supported by a large data set.

Even if it were not adjusted upward, that rate will move Lundy substantially up as a hitter. In the MLE formulas, BBs, hits, and total bases are interconnected, so I can't do revised estimates quickly, and I won't have time to work through the formulas until later this week, but, in the last set of MLEs I posted, Lundy showed up with a career OPS+ of 92 through 1935. My guess is that this change will move him up to around 100, depending on how expanding his numbers to full major-league seasons and regressing them affects the whole workup. It seems a safe prediction that he will look like the offensive equivalent of Dave Bancroft (career OPS+ of 98). I think it very unlikely that he will show up as the equivalent of Joe Sewell (OPS+ of 109, and OBP-heavy), or Willie Wells (who, iirc, had a career MLE OPS+ in the 110-115 range -- my data on hand don't have the career value), but I won't categorically exclude the possibility of a larger rise until I have actually put the data through the full process -- there are too many factors involved to make an eyeball estimate trustworthy, except insofar as to say that his MLEs are going to rise significantly.

Hey guys I've just read this tread from start to finish (I may have skimmed a touch here and there) . . . I realize I've been pretty AWOL lately, trying to get back into the swing of things.

I took some notes:

Completely agree with 'zop on war credit. It was a major mistake not forcing it from the beginning. I was too much of a people pleaser there, and not enough of a leader on that one. C'est la vie . . .

Sunnyday - I don't see how anyone could have Quisenberry on the same planet as Reuschel, even a peak voter. Reuschel's 1977 was outstanding, better than any season Quis had.

The Stats All-Time Handbook RC totals will add up to team R for all of the players on the team for any season, all the way back to 1876 - they normalize them.

I don't like using standard deviation to evaluate players. Period. It came up with Dan R's system and it came up with Bucky Walters too. Just because the spread for teams or players is wide or small doesn't mean players are better or worse. The two are unrelated. They just coincidentally track for a lot (but not all of) baseball history. Chris Dial told me that he's looked at it (taking his word for it here) and standard deviations are going up since 1993 - has the talent gotten worse? That's just one example.

Bill James also pointed out an example at one point - I can't remember where, that basically debunked Neyer/Epstein's use of "SD Scores" in their Dynasties book. Something about comparing an offense in the 1972 AL to NL IIRC.

Juan - Bucky W. gave up more UR than Tiant because UR were much more common in his era. Need to compare each to his peers there.

Great thread guys. More catching up to do later. I'll chime in on Reuschel, who I have ranked pretty high later this week.

Joe Dimino, I disagree that there is no relationship between standard deviation and quality of play. I just think that the relationship between them is misunderstood. Stephen Jay Gould argued that standard deviation moved in inverse proportion to quality of play, period, which I disagree with. In any given year-to-year change, standard deviation and quality of play are *definitely* related--if the best players remain the same, and the worst players get worse (which happens when the league expands), standard deviation will go up. Similarly, if the best players remain the same, and the worst players get better (which happens with integration), standard deviation will go down. However, in the long run, the best players will *not* stay the same--their quality will improve as well. If the improvement in quality of play over time is evenly distributed across the major leagues (so stars and scrubs improve at the same rate), then the standard deviation should not change.

I've done a regression of various variables onto standard deviation, and found that the vast majority of its accounted-for variance is caused by just three factors: integration, expansion, and runs per game. All make perfect sense. Integration was a one-time shock to the system that replaced white scrubs with black stars; you would expect the standard deviation to go down as it did. Expansions are one-time shocks to the system that create jobs for scrubs, which both allows preexisting major leaguers to beat up on weaker competition and drags down the league averages they are compared to. You would expect the standard deviation to go up as it does. And the more runs per game that are scored, the more plate appearances each team and player has, which means that the best players have more opportunities to distinguish themselves from the worst. You would expect the standard deviation to be higher in high R/G leagues, as it is.

Standard deviations were indeed much higher in the 1990's and early 2000's than in the 70's and 80's, which is *exactly* what you would expect--those were high-scoring leagues that expanded twice in five years.

I firmly believe that if you do not account for standard deviations, which corrects for these factors, your Hall of Merit will have an overrepresentation of players from high-scoring years, expansion years, and pre-integration years. Joe Dimino, please let me know if you disagree, and why.

Juan - Bucky W. gave up more UR than Tiant because UR were much more common in his era. Need to compare each to his peers there.

I do. According to my UER trend, Walters was 1-1.5 percentage points above the average, while Tiant was about 2.5 percentage points below the average. I should have said that before...

DanR, the standrad deviation should go down not up with expansion, because assuming major league ballplayers are all at the right end of the Gaussian distribution, it gets fatter (i.e. more players of that ability) as you move leftwards. The extra 50 players in an expansion are much closer to the next worst 50 players than the 50 best are to the next 50. In other words, you've expanded the number of observations, but expanded the dispersion by less. Conversely, adding the Negro Leaguers adds stars and forces out scrubs, but the stars are further from the median than the scrubs, so the standard deviation should go up.

What does reduce the standard deviation is the gradual increase of population, and imporved potential scouting, which produce more A qualityu ballplayers from the broad pool of superb natural athletes, but not necessarily more Babe Ruths. Otherwise one could say that the years with the highest standrad deviation were those with the best quality.

Pity you can't really. Levi Meyerle! .491, baby!

Karlmagnus, you are indeed right that the gap between player no. 750 and 800 is lesser than the gap between player no. 1 and no. 50. Nonetheless, the effect is real--you're still adding 50 players who previously were not good enough to play in the major leagues. That does improve the nominal performance of preexisting major leaguers (MLB hitters can now beat up on ex-minor league pitchers, and vice versa), and it also drags down the league averages. You would expect a spike in standard deviation in the years after an expansion, and you see it, consistently. This isn't just theorizing, it's empirical. I've done the research and am happy to send you the data to show you the correlation. You can even see it anecdotally, though, just by looking at the proportion of career years resulting from expansion--Cash/Mantle/Maris/Gentile '61, Reggie/Petrocelli/Aaron/Staub/Wynn '69, Carew '77, McGwire/Sosa '98.

And you have to look at the whole picture. By the time the leagues are fully integrated in the 1960's, you don't just have four black superstars, you have black utility players, black pinch-hitters, etc., all of whom have replaced inferior white players. Each of those replacements, as well as the entrance of the balck stars, increases the league mean, so all stars are closer to the mean than they were in the pre-integration era. Again, this is empirical, not just theoretical--take a look at standard deviations in the integrated 1950's National League versus 1935-42 and 46. The stdev is 0.12 wins per player per year lower. And this is pre-baby boom, so you can't attribute it to population factors.

What's your evidence that increasing the player population finds more A quality ballplayers but not necessarily more Babe Ruths? I'd believe it, but you'll have to show me some proof...

Another reason that there would be an exceptional spike in standard deviation in expansion years is not just that there is a larger number of qualitatively inferior players entering the league, but there is a larger number of new players entering the league. Whatever their actual talent level, all players go through a period of adjustment when they enter a more competitive environment. During the adjustment period, they don't play as well in the new league as their projections based on earlier play would suggest that they should. Once they have made the adjustment, they then rise to around the level of expected potential. This is clearly the pattern for Negro Leaguers who entered the majors, and Gadfly posted thorough discussions of this effect back in the day. It seems obvious to me that it would have a strong effect on expansion years, and it would also explain why SD values spike in expansion years. Once the new group has made the adjustment, the fact that (as Karl pointed out), they really are not so much worse than the marginal players who were already in the majors appears in the return of SD toward pre-expansion levels.

I've done a regression of various variables onto standard deviation, and found that the vast majority of its accounted-for variance is caused by just three factors: integration, expansion, and runs per game. All make perfect sense. Integration was a one-time shock to the system that replaced white scrubs with black stars; you would expect the standard deviation to go down as it did. Expansions are one-time shocks to the system that create jobs for scrubs, which both allows preexisting major leaguers to beat up on weaker competition and drags down the league averages they are compared to. You would expect the standard deviation to go up as it does. And the more runs per game that are scored, the more plate appearances each team and player has, which means that the best players have more opportunities to distinguish themselves from the worst. You would expect the standard deviation to be higher in high R/G leagues, as it is.

Am I wrong, or does this mean that better pitching eras are penalized in your system by virtue of a smaller runs per game average?

Hitters from low R/G eras get a boost in my system, since they have fewer PA and therefore fewer chances to accumulate WARP, while hitters from high R/G eras are penalized since they have more PA and therefore more chances to accumulate WARP. I haven't done WARP for pitchers yet. Once I do I will obviously do a separate regression on the standard deviation of pitching WARP and report the results.

Speaking as a less statistically abled person, I want to say that it makes sense to me that:

The SDs are affected more by the utterly random fluctuations in the performance of the top 5-10-50 players at the extreme end of the distribution...than by the fairly small and predictable fluctuations in performance represented among the bottom 5-10-50 players at the fat end of the distribution who are the ones who enter the equation in times of expansion (whether 1882 or 1901 or 1961 or whatever).

As I've always said, HoMers are outliers. Statistical averaging doesn't account for the random nature of their gifts and therefore their performance.

It's funny sometimes to think that pitchers are victorious more than 2/3's of the time, but we recognize them less than 1/3 of the time.

Agree in principal with sunny... it may make more sense to calculate SD's of either the bottom half of all MLB players, or at least skim off the top 10% prior to the calculation.

Also, expansions are not necessarily efficient. In addition to the adjustment factor that Chris Cobb mentions, the expansion teams are usually weaker and have less depth than the last place teams were the year before or would be the next year.

sunnyday2, that is definitely true on a year-to-year basis. But that's why we do regression--to separate out the real causes from the noise. I don't use the actual league standard deviation to make my adjustment for a given season--if I did, Eddie Stanky in 1950 would be worth as much as Luis González in 2001, which would clearly be wrong. Instead, I use a regression-predicted standard deviation, which is based primarily on whether the league was integrated, how long it had been since it last expanded, and how many runs per game it scored. This correctly adjusts for how easy the league *actually* was to dominate, while still accurately reflecting years where either very many or very few players randomly happened to have great seasons. HoM'ers outlier-ness does *not* count against them in my system. All I do is adjust for the *conditions* that they played in, which enables us to compare the greats from eras that were easy to dominate and those from eras that were hard to dominate on a level playing field. Is that clear?

Dan - but you say yourself that you haven't run the rule over pitchers yet. Pitchers win more than 70% of the time. Don't you have to adjust for that with your batters at all times?
The batters aren't swinging off tees.

TomH--the league conditions (integration, expansion, run scoring) affect all players, stars and scrubs alike. To repeat again: by using a regression-projected standard deviation, years or eras where there simply happen to be very many or very few great players will still show up that way. All I am doing is correcting for those three factors. Why should I take the top 10% out of the calculation? Their league-relative production, just like everyone else's, will be further from the mean in high-scoring expansion years than in low-scoring years far removed from expansion. I really don't understand this objection at all.

rawagman--I don't understand your point at all...what are you suggesting I should do? I suspect that the stdev for pitchers will be affected differently, and perhaps by different factors altogether, than the stdev for hitters. (It might be easier for pitchers to accumulate WARP in low-scoring eras, for example, since they face fewer batters per inning, for example). Once I have the data for pitchers, I'll report my findings. But I don't see why that would change my results for hitters.

"It might be easier for pitchers to accumulate WARP in low-scoring eras, for example, since they face fewer batters per inning, for example"

I don't think this is the case, because it is a zero sum game, and we are comparing them to a baseline. It's going to be easier for all pitchers of that era to accumulate WARP, and we adjust for this by adjusting stats to era norms.

Because sometimes batters don't hit very well for the simple reason that they are overpowered/overmatched by the pitchers.
There may very well be a highly defined correlation between standard deviations and run scoring environments. But it seems that you lump the run scoring environment among the causes and not as the effect of anything else.
For example, many baseball fans know of 1894 and 1968. The mound was moved back, so the batters were able to let loose more. The mounds were raised and the strike zones were swollen and pitchers dominated. In both cases, there was no real change in the professional baseball playing populace, but the conditions of the game changed greatly, something easily seen within makeup of the run scoring environment.

Joe Dimino, that may be true; I haven't done the study yet so I have no idea what I'll find.

rawagman--I agree with everything you say in #282. And because of the correlation between standard deviations and run scoring (alongside other factors), you will find that the spread of season length-adjusted WARP scores among hitters in 1894 is far greater than the same spread in 1968, simply because the run environment gave them more plate appearances. If you just look at WS or BP WARP, you'll see the best hitters of 1894 with more season length-adjusted WS/WARP than the best hitters of 1968, even though they weren't any better or more dominant. They just got more PA due to the run environment. You have to correct for that.

There are hundreds of things that affect talent levels, quality of play, and ease of dominance. Number of Japanese players, introduction of South American players, increases in technology, introduction of the draft, changes in scouting, changes in the way the minor league system works, changes in the population, changes in interest in the game among youngsters, the bonus baby system, World War II, and on, and on...

As far as I can tell, Dan has selected three factors - integration, expansion, and runs scored per game, and thrown them into a regression equation. The assumption is that this provides for a more accurate result. This may not be the case. The three that Dan chose may be biasing the results more than adding to them. There are many, many ways to set up regression equations - Not all of them are going to yield useful or meaningful results.

I'm not arguing that I could come up with a better regression equation. Rather, I would treat with extreme skepticism what any regression equation indicates about quality of play.

This correctly adjusts for how easy the league *actually* was to dominate,

It is more accurate to say that this attempts to adjust for how easy the league was to dominate. Whether it actually does so is entirely another question.

Dizzypaco--Actually, I stuck around 15 variables into the regression stew, but those were the only three that showed a statistically significant correlation to standard deviation.

Ease of dominance and quality of play are not the same thing. They are very similar on a year-to-year basis--a shock to the system, like expansion or integration, will show a very strong relationship between the two. But there's no conceptual reason to think they would correlate over the entire history of major league baseball.

Sure, I'm not saying I've solved standard deviation by any means. But 44% R-squared ain't bad.

Joe Dimino #266
I don't like using standard deviation to evaluate players. Period. It came up with Dan R's system and it came up with Bucky Walters too. Just because the spread for teams or players is wide or small doesn't mean players are better or worse. The two are unrelated. They just coincidentally track for a lot (but not all of) baseball history. Chris Dial told me that he's looked at it (taking his word for it here) and standard deviations are going up since 1993 - has the talent gotten worse? That's just one example.

Michael Schell relies heavily on standard deviation of performance within the player population in Baseball's All-Time Best Hitters and, with some twists, in the sequel, Baseball's All-Time Best Sluggers (I've looked it over but haven't read it). In the first book his use is essentially SJGould's so DanR will not generally defend it. But I wonder about one particular point, the relation between SD and interleague quality differences. Anticipating critics in the book, or responding to critics (I don't remember which), Schell points to the relatively high SD sustained in the AL for decades as evidence for the relatively high NL quality. That is, from 1901 expansion to WWI the AL never catches up in quality. The AL has not only Eddie Collins (who would be a star but not so dominant in the NL, on this view) but George McBride (who would not have a job in the NL).

I definitely would not agree with that interpretation. Does anybody really think Ty Cobb, Tris Speaker, Eddie Collins, Nap Lajoie etc. would have been exposed as mere mortals if they had played in the NL? I think the AL just had a deadball "star glut," pure and simple. This is precisely why you have to use a regression-projected stdev, reflecting the actual factors that make a league easy or hard to dominate, rather than the actual stdev for a given season--otherwise you can't distinguish between the best of a bad lot (Nellie Fox) and the best of a great one (Collins/Lajoie).

Just looking at the top-ten backloggers and looking at where my ballot is at right this second, and, well, it's possible that we could elect a group of players that includes no one on my ballot. If Hernandez and Roush don't make it, both of which are possible, my consensus score will plummet to new lows, heretofore unseen since the Leroy Matlock days. And, in fact, I'm considering putting Leroy back on, so why not?

I just have to say that I'm happy that we have that extra week. Besides being able to do some extra work in the Plaque Room, I have a decent backlog of plaques already made.

my consensus score will plummet to new lows,

Funny, but I actually have a shot at having the highest consensus score for once.

Funny, but I actually have a shot at having the highest consensus score for once.

Over Howie's cold, dead log-in!!!!

Over Howie's cold, dead log-in!!!!

Heh.

Well, I do have the top-three candidates (I think) higher up on my ballot than Howie does, so it should be interesting.

John I'm still trying to work on my ballot and I might not have time to post the final version if I don't post a ballot by the end off the election can you please count this

1996 ballot
Hernadez , Buddy Myer, and Covelski make my PHOM this year


1. Pie Traynor most 3B putouts 7 times (made my personal HoM in 1942)
2. Nellie Fox led his league in putouts a record 10 years in a row (made my personal HoM in 1971)
3. Chuck Klein 4 hr titles 1 triple crown (made my personal HoM in 1951)
4. Tony Oliva most hits 5 times (made my personal HoM in 1983)
5. Mickey Welch please see his thread (made my personal HoM all the way back in 1898)
6. Sam Rice imagine if he would have started earlier (made my personal HoM in 1940)
7. Pete Browning 13th in career batting avg. (made my personal HoM in 1906)
8. Jake Beckley 30th in hits (made my personal HoM in 1915)
9. Hack Wilson 4 hr titles RBI season record (made my personal HoM in 1940)
10. Hugh Duffy had 100 runs or RBIs every full year he played (made my personal HoM in 1908)
11. Addie Joss 2nd in era (made my personal HoM in 1918)
12. George Kell very good hitter and fielder at important and under elected position (made my personal HoM in 1963)
13. Bill Madlock 4 batting tittles (made my personal HoM in 1984)
14. Al Oliver 1 batting title (made my personal HoM in 1991)
15. Jim Rice hit 300 7 times (made my personal HoM in 1995)
16. Rollie Fingers best HoF speech ever (made my personal HoM in 1991)
17. Harvey Kuenn led AL shortstops in putouts twice assists once (made my personal HoM in 1972)
18. Heinie Manush 330 batting avg. (made my personal HoM in 1957)
19. Edd Roush 323 batting avg (made my personal HoM in 1937)
20. Hilton Smith see his thread (made my personal HoM in 1964)
21. Ray Schalk the best catcher ever (made my personal HoM in 1938)
22. George Van Haltren 31st in runs (made my personal HoM in 1925)
23. Jimmy Ryan 30th in runs (made my personal HoM in 1926)
24. Luis Aparicio being a better offensive player then Rabbit puts him here (made my personal HoM in 1979)
25. Jake Daubert 29th in triples (made my personal HoM in 1930)
26. Bobby Veach most doubles twice (made my personal HoM in 1931)
27. Bill Mazeroski probably saved on average around 90 runs a year (made my personal HoM in 1984)
28. Roy Thomas most times on base 6 times (made my personal HoM in 1984)
29. Keith Hernandez 11 gold gloves (makes my personal HoM this year)
30. Gavvy Cravath most active HRs 1918, 1919 and 1920 (made my personal HoM in 1928)
31. Lou Brock like the steals more then most (made my personal HoM in 1984)
32. Kiki Cuyler 2299 hits (made my personal HoM in 1968)
33. Lloyd Waner had the most OF putouts 4 times, finished 2nd once and finished 3rd twice (made my personal HoM in 1968)
34. Ginger Beaumont 1902 batting title (made my personal HoM in 1932)
35. John McGraw 3rd in on base percentage (made my personal HoM in 1930)
36. Steve Garvey 200 hits 6 times (made my personal HoM in 1984)
37. Jim Bottomley 2313 hits (made my personal HoM in 1968)
38. Levi Meyerle best rate season ever (made my personal HoM in 1975)
39. Eddie Yost most walks 6 times most times on base 3 times (made my personal HoM in 1987)
40. Rabbit Maranville best shortstop before Ozzie moves down do to reading accounts on how his drinking hurt his team more then the numbers show(made my personal HoM in 1939)
41. George J. Burns most walks 5 times (made my personal HoM in 1932)
42. Orlando Cepada 297 batting avg 379 HRs (made my personal HoM in 1987)
43. Stuffy McInnis led in fielding% 6 times (made my personal HoM in 1987)
44. Deacon Phillippe best walks/9 IP in the 20th centaury (made my personal HoM in 1988)
45. Babe Adams led in WHIP 5 times (made my personal HoM in 1992)
46. Buddy Myer 1935 batting title (makes my personal HoM this year)

explanation for players not on my ballot
Dick Redding barring new evidence not one will make my ballot (the HoF vote has absolutely no bearing on my vote )
Jimmy Wynn don’t buy the Houston logic
Charley Jones no black list points
Charle Keller I’m not giving him WWII or minor league credit of MVP seasons like others.

John I'm still trying to work on my ballot and I might not have time to post the final version if I don't post a ballot by the end off the election can you please count this

Will do, yest.

That's ok, I come in 2nd a lot, and in the 1940s-50s I wasn't even a consensus contender..

Bringing this over from the ballot thread:

ronw wrote:

Chris, I respect your opinion and baseball knowledge, but I think that you are being way too self-righteous with your ballot policing. Criticism is one thing, but obliging a relatively stable participant to reconsider a ballot is just plain wrong.

As I said in my comment, I understand that my view about ballot consideration is not the majority position, and I don't seek to impose it on the electorate. I disagree that my view is just "plain wrong." It is not the way we have done things, and I respect the community's practices. _I_ cannot obligate anyone to revise their ballots, and I hope I did not appear to claim to be doing so. I think it is valuable for the process for us to consider carefully what our standards of reasonableness are, and I think that higher standards of reasonableness would lead to better ballots and better overall results.

From above:

However, it seems a vote that _no other_ voter can see a rational justification for.

Chris, I note that you used to feature Bobo Newsom pretty prominently on your ballot when no one else saw it. karl is the only one to vote for Sam Leever. yest had a Tony Oliva fiasco a few years ago.

I think rico gave a reason. He said Buckner had a long career, good career stats, led the league in a couple of different stats, accumulated some solid seasons along the way.

All of us may disagree with the reason, but I think that the original ballot should stand. (I realize Rico revised his ballot).

There's a difference between being the only one to vote for a candidate and voting for a candidate whom no one else can see a rational justification for. The group's standard for "rational justification" seems to be "offering a comment as a reason." I would prefer a higher standard of proof. If two players are near contemporaries, playing positions of similar defensive value, have careers of similar length, and one player has a career OPS+ that is 25 points higher than the other, I think that a serious argument for how the difference in value is made up ought to be offered, or the vote should not be counted. That seems to me to be the responsible way to proceed. We have decided to encourage but not require people to adhere to that sort of standard in order to participate in the project. That has advantages for inclusiveness, but diminishes somewhat the seriousness of our efforts to arrive at good answers about questions of value and merit.

If we are going to bring up the "Hall of Merit ballot committee" again lets actually form it instead of just arbitrarily deciding that In practice, we operate as a committee of the whole on this one... In other words,if we're going to cite the Constitution against a ballot, lets actually follow it. I want to serve on that committee, particularly so I can defend voters from having their ballots disregarded.

I was not "abitrarily deciding" what we do: I was describing what we do. This is what we do: if somebody strongly objects to a ranking on someone else's ballot, that somebody calls the voter on it. If others feel the same way, they speak up, too. The voter either responds to the criticism, accepts the criticism, or ignores the criticism. Then, if there is a lot of objection and/or disagreement about the objection, we may have a discussion about the Constitutionality of the ballot. Then we look at the Constitution, and discuss what it means. Either a clear body of opinion emerges about the Constitutionality of the ballot, or Joe eventually gets involved and makes a decision. This has happened many times. I went to the Constitution to show that it does address standards of reasonableness, as well as the principle of fairness to all eras. I don't see how this ballot dispute is qualitatively different from any other, except that it turns on the issue of reasonableness.

If this sounds a little bitter, it is because I am a bit upset about this practice. Something really bothers me when a long-time voter is "obligated" to reconsider his ballot. Criticizing the ballot is fine, but forcing someone who has participated for several months or years to reconsider is insulting to the voter, because it assumes that the voter had not actually considered his ballot. In a way, it is insulting to all of us. Do you assume that I don't consider my ballot? What about Rob Wood, Ken Fischer, who post limited explanations and/or don't post often.

I am sorry that you are upset. As it turns out that no long-time voter has been obligated to reconsider a ballot, I'm not sure whatt you are upset about, however, except perhaps that the Constitution provides for an (unexcercised) process for requiring any voter to reconsider a ballot, and that I have gone so far as to suggest that we might actually use that process. I don't make any assumptions about whether or not someone has considered his ballot. I study each ballot and the explanations on it, consider what I know about the voter's idea of what merit it, and try to discern the rationale for a player's rankings. I have no trouble discerning the application of clear, consistent, and sensible criteria in the cases of Rob Wood and Ken Fischer. If either of them had voted for Buckner, I think it would have provoked a similar reaction from the electorate, if they had not included a justification. If I see a case where a voter seems to be applying his own criteria inconsistently, I might ask him a question about it, as I've done a few times. If a case has arisen where a vote seems completely incompatible with the information provided by basic measures, I might point that out. I didn't initiate the questions about Rico's ballot; I did bring in the matter of the Constitution because I thought it was relevant.

I think the voters have an obligation to submit a carefully considered ballot. If they do not, then it undermines the integrity of our project. I think upholding the integrity of the project is more important than worrying about insulting someone by requiring them to change their ballots because their votes don't meet a minimum standard of reasonableness. We don't in fact, do that, and that's ok with me, but I'm not going to stop bringing the idea up for fear of insulting long-time voters because they might think that I am assuming that they don't consider their ballots. And that's probably more than enough said about that.

Chris:

As I said, I don't mind you criticizing a ballot. I mind you suggesting that a ballot that complies with the rules is unconstitutional because the person has an "unreasonable vote" for a player. I also mind you suggesting that someone is "obligated" to reconsider his vote. I don't think that either is right and both seem too controlling to me.

Questioning a ballot is one thing. Disallowing a ballot should be used only if someone refuses to provide comments and/or for new voters who do not fairly consider all eras.

Having participated in this project since 1911, I know that we have never disallowed an existing voters ballot for a weird selection. I would venture a guess that the first time a long-time voters ballot were disallowed because of a controversial vote, that we would suddenly have a drop in participation.

Perhaps your comments on rico's ballot were meant only as criticism. My reading of your comment on rico's ballot was that his ballot should be disallowed unless he reconsidered Bill Buckner. If that wasn't your intent, then I should not have been worked up and I apologize.

However, I think that the ballot of an existing voter (say 5+ years) who fills out his ballot with 15 names, provides some minimal comment on each one, is not admittedly engaging in strategic voting, and comments on the top 10 returnees should never be disallowed. The person can be freely criticized, but he should not be obligated to make any changes to his ballot. If he wants to, he can submit a new ballot.

Finally, I think that ballot disallowance actually undermines integrity of the project. Each one of us trusts the others to carefully consider each ballot, to be fair to all eras, and to come up with their own independent ideas. I think that the threat of ballot disallowance could curtail free-thinkers from participating in the group. One should never refrain from voting for a player because that player is unpopular with the majority (or all) of the group. By banning the disallowance of long-time voter ballots, I am upholding that integrity.

Hey, Doc's ballot was a lot more weird than Rico's!

;-)

Hey, Doc's ballot was a lot more weird than Rico's!

;-)

I agree. I just write more stuff after the player(s)' name than Rico does. ; )

I've been keeping very quiet because as a notorious Beckley supporter I would doubtless alienate everybody. I have to say however that if Rico wants to stick to his original ballot I think he should be allowed to, even though it's also appropriate for others, particularly long standing highly respected others like maybe-Grandma and Chris Cobb to question it vigorously.

This project is full of sad, sad people who will never vote for Beckley,. They divide into two distict categories: those who put him 150th and make cracks about "Boy, anyone who votes for that bozo is from Mars" and those more maddening ones who put him 17th every election, but always find some new and exquisitely qualified candiadate to move above him so he never actually makes it onto the bottom of their ballot. Buckner was one such, but as one who feels personally every wound received by the noble Jake, I can tell you there have been LOTS of others.