I assume Carey gets a big edge in fielding prowess, because Heilmann is a hell of a lot better hitter than Carey ever was.

How was Heilmann never an elite outfielder? He had an amazing stretch from 1921 to 1930 including a 194 OPS+ season in 1923 when he hit .403 and slugged .632 which was second only to Ruth.

How was Heilmann never an elite outfielder?

He wasn't one of the top three outfielders of his era.

Re: Heilmann

Was he better than Cobb, Ruth, Speaker or Charleston? Not even close.

How about Cristobal Torriente?

Yes, for a few seasons he was a terrific hitter (his defense was suspect though), but it's not enough to place him in the "no-brainer" sections for outfielders, IMO. I honestly don't know where to place him.

He was better than Cobb and Speaker in the early 1920s, by which time they were both past their peak, and nobody knew about Charleston so from 1921-25 he was at least the #2 outfielder in baseball.

He was better than Cobb and Speaker in the early 1920s, by which time they were both past their peak, and nobody knew about Charleston so from 1921-25 he was at least the #2 outfielder in baseball.

From 1914-1932 (the years that Heilmann was playing), Cobb, Ruth, Speaker and Charleston (and there were quite a few players and members of the media who knew who the latter was) were, without a doubt, better than Heilmann.

My point is not to go overboard with the outfielders in proportion to the infielders, that's all. Proceed with caution. :-)

The Bill James HOF question is not "did he overlap with anybody better" but "was he ever the best in baseball at his position?" From 1921 to 1925 Heilmann still wasn't the best, since Ruth played RF, but if you count the outfield as containing 3 players, the top 3 outfielders over that period were Ruth, Heilmann and possibly (we will never really know) Charleston. Cobb was a greater player, but not by the early 20s.

From 1921 to 1925 Heilmann still wasn't the best, since Ruth played RF, but if you count the outfield as containing 3 players, the top 3 outfielders over that period were Ruth, Heilmann and possibly (we will never really know) Charleston. Cobb was a greater player, but not by the early 20s.

Yes, for a few years, he was damn good. But isn't this a form of cherry-picking so that he's not compared to his true contemporaries. He did play 14 seasons with Cobb and Speaker, plus 16 "with" Charleston.

BTW, I doubt that they're any creditable analysts, sabermetric or non-sabermetric, that would place Heilmann near Charleston as a player.

Another James question: Could a team with Heilmann as its best player win a pennant? Heilmann's best season was 35 WS, he also seasons of 32 and 30 (twice). That seems like a yes to me.

Heilmann did serve half a year in WWI, and played only 79 G in 1918 during which he earned 12 WS, so it seems like that in a full season, he might earned 18 to 24 WS to bring the total between 360 and 365 for his career.

The 1923-24 Detroit Tigers got quite close, in spite of having no pitching and the opposition having Babe Ruth. We're talking 5 seasons of dominance, not 1. I have Heilmann at #3, above even Beckley on my provisional 1936 ballot, even though he had fewer hits (though he probably wouldn't have beaten Caruthers.) :-))

He's WAAAAAAAY better than Carey, who appears to me slightly sub-glut, and close to Hooper.

Heilmann was better than Ruth in 1925 and second to Ruth (for RFs) for a majority of the 1920s. being second to the greatest player of all-time should not exclude one from being "elite."

Was Speaker not an elite CF in the teens and early 20s? Cause Ty Cobb was better than he was every one of those years.

I think that's a pretty strict definition of elite if you don't let Harry Heilmann in. Too restrictive. I'm not saying people should go all "Frankie Frisch" on people and say everyone is elite, but come on now. Harry was a monster at the plate for a decade. Let's give him his due.

What about Sisler? Should also be in the discussion. MUCH better hitter than Ichiro!

I have an offense-only system, based on a version of RCAA scaled to the value of the run environment. (The scaling isn't linear, but incorporates elements of the PytPat sliding exponent). The units involved in the following tables aren't runs, they're an arbitrary unit vaguely related to wins. The first column is scaled RC above average. The second column is a bonus for having really big offensive year, years far above average. The third column is scaled RC above 75% of average, and is hence a little friendlier to longer, lower peak careers - even so, it's more a peak measurement than anything else. The fourth column is some arbitrary way of combining the first three together.

Here is the very top of the LF/RF list, with the top two CF thrown in:

Name RCAA Big yrs RC>75% Composite
Ruth 157 241 183 701
Cobb 143 171 178 577
Speaker 109 112 141 419
Crawford 70 51 100 247
Jackson 60 63 75 230
Heilmann 63 52 88 229
Clarke 60 39 88 206
Magee 54 40 78 192
Flick 52 44 70 188
Goslin 49 30 78 173
Wheat 50 28 79 171</pre>

Now here's a slice of the middle of the CF list:

Van Haltren 40 21 67 138
Duffy 38 25 62 136
Ryan 36 20 64 131
Roush 37 21 60 129
Thomas 37 23 55 127
Carey 31 13 58 109
F. Jones 31 13 54 104</pre>

(That's actually Carey with the last few years left off - he looks better that way.)

Conclusions: First, Heilmann is just too much of a hitter to ignore. If I'm comparing him to Crawford, Jackson, and Clarke, the implications for the ballot are clear. With Carey, it's clear that his offense alone, (well, his peak offense) is not enough. He's on my ballot anyway for several reasons: his defense, the length of his career, the fact that this analysis may be undervaluing his years for which CS are unknown, and maybe a little subjective bonus for scoring runs.

Heilmann was better than Ruth in 1925 and second to Ruth (for RFs) for a majority of the 1920s. being second to the greatest player of all-time should not exclude one from being "elite."

I think we can agree that Heilmann won 1925 by default after Ruth became sick (not really a knock on Heilmann, of course).

Was Speaker not an elite CF in the teens and early 20s? Cause Ty Cobb was better than he was every one of those years.

Was Speaker one of the top three outfielders of his time? Without question (in fact, he was #2). How is he similar to Heilmann?

I think that's a pretty strict definition of elite if you don't let Harry Heilmann in.

Who said I wasn't going to place him on my ballot?!? Without working on my ballot, he'll be in my top-ten on my ballot (#5 or #6? I don't know yet...maybe higher). I was only suggesting that he isn't a no-brainer (which does not imply that he is borderline).

We're talking 5 seasons of dominance, not 1.

Certainly, peak voters will love him (as they should).

I think Heilmann is clearly a HoMer. But he was, during his peak in the early 1920s, no better than the 3rd best outfielder in baseball, if you assume he was better than Oscar Charleston (which I find doubtful)

Win shares, 1920-26
Ruth 246
Charleston ??
Speaker 206
Heilmann 190
Wheat 165
Cobb 161
Rice 161
Carey 160
Torriente 159 (my estimate)
Youngs 155
K. Williams 153
Roush 152

If you looked at 1921-25 or 1921-27, he and Speaker would switch places. But looking at the first half of the twenties altogether, Speaker remained the better player, leaving aside the superiority of his total career to Heilmann.

What about Sisler? Should also be in the discussion. MUCH better hitter than Ichiro!
Chance 44 34 58 153
Sisler 36 29 63 142
Fournier 38 28 55 136
Beckley 36 15 68 126
Konetchy 30 15 55 108</pre>

OCF, if you add 1Bs, where are Beckley and Sisler? 1B was clearly more important defensively in the 1890s, but probably not by the 1920s.

Thank you, OCF; interesting. Sisler just above the glut, Beckley in it (but yours is a peakish measurement, and 1890s 1B probably 10% more important than OF.) Need to look at Chance again -- looks there like more career there than I remember. Are you sure about his third factor? -- 58 looks awfully high.

I'm on the Heilmann bandwagon as well. He's got a monster 1921-27 peak. An extension of his peak to 1918-1930 does not embarrass him either. For career, he's got a 148 OPS+ in 8960 PA's.

Is there some defensive liability that would put his candidacy under any doubt?

Also, the early-peaking Al Simmons overlaps quite a bit with Heilmann and would make an interesint comp. Simmons won't be eligible for many years, but he wasn't adding much to his resume after 1934.

karl, one way to look at longevity contributions would be to look at the third number minus the first: Chance 14, Sisler 27, Fournier 17, Beckley 32, Konetchy 25. That makes it very clear that Chance's career was short; he ranks where he does because his rate stats are so high. It's also true that this chart doesn't care about how the career is partitioned into years. Playing 90% of the games for 6 years will come out the same as playing 60% of the games for 9 years.

Also, the early-peaking Al Simmons overlaps quite a bit with Heilmann and would make an interesint comp. Simmons won't be eligible for many years, but he wasn't adding much to his resume after 1934.

I like Simmoms much better than Heilmann. Longer career, better defense and many seasons in CF trump Heilmann's offense, IMO.

MUCH better hitter than Ichiro!

A little off topic....

Just comparing Sisler's best four seasons to Ichiro's only four seasons, it's close. If you adjust their averages over the four years for park (I used the b-r.com pfs and a simple three-year average for Safeco v.2004), then you subtract from that figure the league's averages during the period, Ichiro has hit .097 higher than the league and Sisler .084 higher than his leagues. I may have not done the park adjusting right, but Safeco's been a black hole while STL was mostly a decent hitter's park. Sisler's probably not WAY better...if he's better.

It's very odd, but looking at cahnce again, I still don't see the beef. Only 1200 hits, and fewer than 1 hit per game, not usually a sign of a goodie. Where does Sheckard rank among your outfielders? - also just below 1 hit per game, although 2000 hits. If Chance is that good Sheckard should be up about 160, and it's just an extremely weird enviroment.

Sheckard: 39-23-65-138. The thing about Chance is his very high OBP in a very hostile run environment, plus Chance adds a fair amount to RC with SB.

Of course that whole system may be a little too peak-friendly and career-insensitive for me to stand, so Chance was last seen at #17 on my ballot, and I would (and did) vote for Sheckard ahead of him.

I do find it interesting that Heilmann did pretty well on MVP/League award finishing 3rd-9th-4th-5th-2nd 1923 to 1927 while Carey did not get any votes in Chalmers/League award vote. Heilmann's ranking should be lowered a notch or two because previous winners like Babe Ruth (1923) were not allowed to win again. Still puzzling about Carey..where writers back then just ignorant of leadoff hitters or is there a mistake in baseball-reference.com?

Thanks, OCF; I think your system shows itself as pretty peak-friendly, but that's helpful because my natural math-major bias leads me to count things, which is why I like Beckley and verge towards career-types. As we get new borderline cases (which I don't think Heilmann is, though I agree he's not top 20 or anything, and have him below whichever of Lloyd or Williams is left over) it will be very good to benchmark them against other such.

Incidentally, OCF, where's Hal Chase? Contemporaries LOVED him, but we don't.

Incidentally, OCF, where's Hal Chase? Contemporaries LOVED him, but we don't.

They loved him only when he was honest.

Chase: 18-7-41-71. Slightly better than McInnis or George H. Burns but well behind Judge and Daubert. Most of the gee-whiz quotes I've seen about Chase concerned his defense - and I'd be tempted to assign negative value to his defense, because whether he caught a throw might have depended on which way the money was flowing. If you want to win games and pennants, the difference between Chance and Chase is ... you fill in the blank.

Thanks once again, interesting because his raw numbers are comparable to Hoopper or Carey

To me there are six players recently or soon to be eligible who are below NB status but definite contenders. It's tough placing subjective candidates like Pike and Torriente, but these guys are all documented and hopefully we can get them right. But of course, quality vs. quantity, offense vs. defense are not such easy issues to decide either.

The big six are Bancroft, Carey, Groh, Heilmann, Roush and Sisler.

Win Shares
Heilmann 356 (#3 player of the 1920s)
Carey 351 (#20 but started way back in 1910)
Roush 314 (#15)
Sisler 292 (#28)
Groh 272 (#26)
Bancroft 269 (#17)

My Favorite Toy (Reputation Monitor--remember anything over a 200 is virtually an automatic Cooperstown HoFer, while most 175-200 are, too, but like the HoF Monitor, this may underestimate defense)
Heilmann 237 (#10 RF)
Sisler 200 (#15 1B)
Roush 184 (#13 CF)
Carey 176 (#14 CF)
Groh 168 (#13 3B)
Bancroft 153 (#22 SS)

So I agree with those who say that Heilmann is almost an NB. Right now I am looking at him at #4 on my 1936 ballot. But then look at Max Carey.

Heilmann 8,643 PA at 148 OPS+, 2,830 R+RBI 1914-32, mediocre RF

Carey 10,367 PA at 107, 2,345 R+RBI 1910-29 (i.e. much more deadball play), great CF plus almost 7X more SB than Harry

For you career voters, I don't see Max too far behind Harry. Then there is Sisler.

Sisler 8,739 PA at 124 OPS+, 2,459 R+RBI, good 1B, not too far behind Harry either IMO.

Peak OPS+
Heilmann 195-179-172-167-161
Sisler 179-169-163-159-153 not too far back
Roush 162-153-149-147-145
Groh 151-159-144-132-123

Meanwhile nobody would base the case for Carey or Bancroft on OPS. But take another look at Carey and Roush, and then the Reds teammates.

Carey 10,367 PA at 107, 2,345 R+RBI 1910-29 (i.e. much more deadball play), great CF

Roush 7,847 PA at 126, 2,080 R+RBI, mediocre CF

Groh 6,770 PA at 119, 1,484 R+RBI, good 3B who played much more of his career in deadball era

XBH
Heilmann 876
Sisler 691
Carey 648
Roush 589
Bancroft 425
Groh 421

All will be on my ballot and I think this is the right order:

4. Heilmann--near NB
8. Groh--deadball peak at underrepresented position
9. Sisler--James trashed him but I think he overstates that case
11. Carey
12. Bancroft--one of the top 5 fielding SS of all-time (ML)

Not sure yet about Roush, but these six deserve a serious look.

Where can you find Negro League statistics? And where can you find a good comparison between the Negro Leagues and the major leagues?

I've always heard that the Negro Leagues were tantamount to Triple-A baseball. Is that correct?

The big six are Bancroft, Carey, Groh, Heilmann, Roush and Sisler.

Heilmann is the only one I consider a "must-elect". (And yes, I do timeline, or, as I would put it, I don't reverse-timeline). Groh is on my ballot, and already has a level of support that guarantees his election unless there is a massive shift of opinion. I would relegate the others to the top of the Hall of the Very Good. Then again, distinguishing between the very very good and the meritorious is the toughest part of this project.

How was Heilmann never an elite outfielder? He had an amazing stretch from 1921 to 1930 including a 194 OPS+ season in 1923 when he hit .403 and slugged .632 which was second only to Ruth.

The system I've been describing on this thread does create a number for each year. Heilmann's 1923 is indeed a monster year. Out of my the 2000 or so player-years I've worked up so far, who ever had an offensive year as good or better?

Babe Ruth 11 times
Lou Gehrig 8 times
Ty Cobb 7 times
Honus Wagner 5 times
Rogers Hornsby 4 times
Nap Lajoie 1901, 1904, 1910
Joe Jackson 1912, 1913
Tris Speaker 1912, 1916
Eddie Collins 1909, 1914
George Stone 1906
Sherry Magee 1910

Heilmann's 1927 wasn't that far behind. Just as a reminder of the effect of adjusting for run context, I have both of these Heilmann years clearly ahead of Hugh Duffy 1894. Of course, the gap between Ruth and Heilmann, especially in 1923, was enormous.

Anybody have an idea what Heilmann's OPS+ of 148 would be today? 135? 140?

I'm trying to see what hitter today resembles him for comparison.

Heilmann's OPS+ of 148 would be 148, or very close. This is the 1920s, not 1875; samples sizes are large enough and variances small enough that the standard deviation of OPS+ is close to modern levels.

Think Pujols or Ramirez, having a good year.

CAREER
1905-1940

RUNS CREATED RATE PLAYER LEAGUE
1 Babe Ruth 249 2910 1168
2 Rogers Hornsby 204 2108 1031
3 Joe Jackson 202 1164 577
4 Ty Cobb 200 2757 1376
5 Lou Gehrig 199 2367 1187
6 Jimmie Foxx 193 2055 1065
7 Tris Speaker 183 2353 1284
8 Mel Ott 177 1731 978
9 Honus Wagner 173 1296 747
10 Arky Vaughan 165 1051 636
11 Harry Heilmann 161 1726 1074
12 Sam Crawford 159 1252 785
13 Chuck Klein 158 1379 872
14 Eddie Collins 155 2024 1308
15 Joe Medwick 155 995 644
16 Paul Waner 151 1721 1137
17 Hack Wilson 151 1050 695
18 Nap Lajoie 149 934 625
19 Sherry Magee 149 1274 857
20 Babe Herman 147 1146 779
</pre>

CAREER
1968-2003

RUNS CREATED RATE PLAYER LEAGUE
1 Barry Bonds 211 2471 1174
2 Frank Thomas 183 1664 910
3 Manny Ramirez 175 1229 704
4 Mark McGwire 173 1504 867
5 Larry Walker 173 1478 854
6 Edgar Martinez 170 1543 907
7 Jim Thome 170 1272 748
8 Jason Giambi 166 1064 641
9 Jeff Bagwell 164 1650 1005
10 Willie Stargell 163 1138 696
11 Mike Schmidt 161 1735 1080
12 Gary Sheffield 158 1479 935
13 Carlos Delgado 158 1039 657
14 Alex Rodriguez 158 1121 710
15 Chipper Jones 157 1149 733
16 Ken Griffey Jr. 156 1523 975
17 Joe Morgan 155 1532 991
18 Willie McCovey 154 930 602
19 Reggie Smith 152 1193 784
20 Mike Piazza 152 1130 745
</pre>

There... two equal-length periods... ranked by top RC+ (vs non-pitchers only). Note some players have had their career's clipped a bit. I used a 5000 PA cutoff.

Heilmann looks a lot like Mike Schmidt with the bat, although Gary Sheffield looks like a better subjective match.

Anybody have an idea what Heilmann's OPS+ of 148 would be today? 135? 140?

Karlmagnus is basically correct, but if you want a fine-tuned timelining, you could use BP's.

They drop Heilmann's .316 EQA to .309, which is a drop of about 2.5%.

Applying that penalty to Heilmann's OPS+ yields a competition-adjusted OPS+ of 144.3

They drop Heilmann's .316 EQA to .309, which is a drop of about 2.5%.

Applying that penalty to Heilmann's OPS+ yields a competition-adjusted OPS+ of 144.3

That's what I was looking for, Chris. There had to be some decrease due to competition. Thanks!

I don't see Heilmann and Carey as remotely comparable unless you don't believe in peak at all. Similar career value--slight edge to Harry once you factor in war time--and Heilmann had a monster peak (4 30+ WS seasons topping at 35), while Carey never broke 30. WARP grossly overstates the value of guys like Carey because of the absurdly low defensive replacement level. Heilmann is, I believe, a "no-brainer" HoM'er; he's just not an inner-circle one. Sheffield is a very good comp. Heilmann could easily be a consensus no. 1 in a middling year. Carey is on the bubble.

I don't see Heilmann and Carey as remotely comparable unless you don't believe in peak at all.

I agree. That's why Heilmann will be high on my ballot, while Carey probably will never make it.

Ever notice those context-dependent ads that appear near the top of the page? As I write this, one of the two ads is for bats, and the other one is for gloves. That seems appropriate.

Heilmann could easily be a consensus no. 1 in a middling year.

Judging by posts like #41 and #42, Heilmann vs. Carey doesn't seem like much of a debate.

It seems likely that Collins and Lloyd will be elected in 1935 and Williams and Alexander in 1936. So, for 1937, who's the highest ranking outfielder: Heilmann or Torriente? That's the debate I'd like to see. Dan R. - that's your quote, and it seemed hypothetical. Make it a year with Torriente available - what is Heilmann in that case?

I think Heilmann and Torriente seem quite comparable, and if there's a tie between an ML and Negro League player I will vote for the ML player since the record is better documented. But Torriente would be right on his heels.

Heilmann was just a historically great hitter. Ya gotta give the man his due--10 more hits in his career (I think) and he woulda hit .400 four times!

I think Heilmann and Torriente seem quite comparable, and if there's a tie between an ML and Negro League player I will vote for the ML player since the record is better documented. But Torriente would be right on his heels.

I'm going to have Torriente ahead of Heilmann, but it's close.

Torriente over Heilmann.

Similar hitters, but Torriente was a fine, probably an outstanding centerfielder, while Heilmann was an average right fielder at best.

Reconsidering my omission of Carey for 1926 . . . I agonized for a while before submitting my 1925 ballot, and now I'm having second thoughts.

One thing I like to look at is "Pro Bowls" made. The Pro Bowl is like the All-Star Game, but it's after the season. My pro-bowl teams consist of the Top 2 at each position, and the 5 best pitchers, for a 23 man Pro Bowl roster. It lets me gauge how well a player performed against his peers.

Max Carey makes 11 Pro Bowls in the NL:

He was the best left fielder twice, in 1912 and 1915.
He was the best center fielder 8 times, each year from 1916-1918 and and 1921-1925.
He was also the #2 centerfielder in 1913 (to Leach).

Harry Heilmann makes 9 Pro Bowls in the AL:

He was the best right fielder four times, in 1921-1922 and 1925-1926.
He was the second best right fielder in 1916 (to Hooper), 1923-1924 (to Ruth), and 1927-1928 (to Ruth).

Considering the relative importance of CF and RF, I think I have missed the Carey boat, by at least a year.

Reconsidering my omission of Carey for 1926 . . . I agonized for a while before submitting my 1925 ballot, and now I'm having second thoughts.

Why were you having second thoughts back then?

:-D

My pro-bowl teams consist of the Top 2 at each position, and the 5 best pitchers,

Interesting concept. Considering that teams have required 4-5 "median aces" to cover their IP for a season after 1898, having only 5 pitchers on the Pro-Bowl team shorts the credit for the pitching staff relative to the position players. 8-10 might be more appropriate.

Chris Cobb #47
Heilmann was an average right fielder at best.

Who are the worst outfielders by Fielding WS?

Here are the worst outfielders by Fielding Win Shares (WS/1000ip) among the 260 outfielders with 10000+ career innings --according to Bill James, Win Shares (thru 2001), p618-20.

Outfield WS/1000ip (10000+ innings played)
1.69 Jeff Burroughs
1.73 Greg Luzinski
1.79 Harry Heilmann
1.8x --
1.9x Frank Howard, Dave Winfield, Bruce Campbell, Elmer Valo, Tommy Griffith, Rusty Staub, Patsy Donovan

That is, I count 10 outfielders with WS/1000 < 2.00 among the 260 leaders by innings. (Winfield, Staub, and Donovan are among the Top 100 outfielders by innings, 14186+. Heilmann is #110.)

Here are the leading and trailing outfielders by Fielding Win Shares (WS/1000ip) among the 100 outfielders with 14186 career innings --according to Bill James, Win Shares (thru 2001), p618-20.

Outfield WS/1000ip, 14186+ innings played
(parenthetical interpolations: 12500-14185ip)
(Curt Flood)
4.93 Tris Speaker
(Marquis Grissom, Paul Blair, Jimmy Piersall)
4.37 Max Carey
4.33 Lloyd Waner
4.29 Fielder Jones
4.25 Amos Otis
(Sam West, Garry Maddox)
4.13 Willie Wilson
4.11 Willie Mays
4.01 Joe DiMaggio
. . .
86 others, > 14186ip
(116 others, > 12500ip)
. . .
(Chuck Klein)
2.11 Ken Griffey Sr
(Bill Nicholson)
2.06 Gary Matthews
(Ruben Sierra)
2.01 Del Ennis
1.99 Patsy Donovan
1.99 Rusty Staub
1.92 Dave Winfield
(Harry Heilmann)

stats comparison
player ..years games OWP RCAA
Heilmann all ....2198 .706 663
Carey 1910-26 ..2205 .569 183

I chopped Carey's last 2 mostly worthless years off, so they come out at 16 years each, and same playing time. Heilmann is offensively 30 runs / year better (30 * 16 = 480 = 663 - 183)

Was Carey 30 runs / year better with the glove? That is 9 win shares per year. Seems too much - gotta go with the stick.

Was Carey 30 runs / year better with the glove? That is 9 win shares per year. Seems too much - gotta go with the stick.

Why does it seem too much for a very good center fielder to be worth 30 runs more with the glove than one of the worst rightfielders in history?

Consider, for example, 1921-1925, when Carey was making 400-450 put outs per year, and Heilmann was recording 200-350 per year (including put outs at first base!). Heilmann never had more than 279 outfield assists in any single year. Carey bested that by 100 put outs in 6 separate years.

Now, I didn't have Carey on my 1925 ballot either, but it seems to me that, to be fair, while Heilmann's 148 to 107 OPS+ advantage is impressive, Carey's 6363 to 2794 outfield put out advantage and 339 to 183 outfield assists advantage (there's 10 of your 30 runs right there!) are at least equally impressive.

Interesting concept. Considering that teams have required 4-5 "median aces" to cover their IP for a season after 1898, having only 5 pitchers on the Pro-Bowl team shorts the credit for the pitching staff relative to the position players. 8-10 might be more appropriate.

I tend to look at the "bell curve" of talent more than strict numerical accuracy. To pick a year when Heilmann and Carey were both "Pro Bowlers" at their position, consider 1921.

Here's the AL right fielders by WARP-1 (my ranking does not strictly follow WARP-1, but it is based on it). Top 2 make the "Pro Bowl":

Harry Heilmann -- 9.6
Bob Meusel -- 8.0
Jack Tobin -- 7.6
Harry Hooper -- 5.1
Whitey Witt -- 4.7
Elmer Smith -- 4.2
Shano Collins -- 3.3
Clyde Milan -- 2.4

Here's the NL Centerfielders

Cy Williams -- 8.3
Max Carey -- 8.2
Ray Powell -- 7.8
George Burns -- 7.8
Ed Roush -- 7.2
Les Mann -- 4.6
Hy Myers -- 3.9
George Maisel -- 3.2

Now, consider the AL Pitchers:

Red Faber -- 15.0
Urban Shocker -- 11.1
Carl Mays -- 10.3
Sam Jones -- 9.4
Stan Coveleski -- 8.6
George Mogridge -- 8.5
Eddie Rommel -- 7.9
Walter Johnson -- 7.4
Waite Hoyt -- 7.3
Joe Bush -- 7.3

And the NL Pitchers:

Burleigh Grimes -- 10.0
Dolph Luque -- 7.3
Pete Alexander -- 6.9
Eppa Rixey -- 6.6
Joe Oeschger -- 6.6
Wilbur Cooper -- 6.2
Babe Adams -- 5.0
Whitey Glazner -- 4.8
Jesse Barnes -- 4.5
Clarence Mitchell -- 4.5

So you see that, in general, the "Pro Bowl" requirement is above 8.0 WARP-1 points. It works out comparably by including 5 pitchers, irrespective of the theoretical equivalent number of "median aces".

You can see, for example, in the AL that stopping at 5 pitchers keeps out the likes of Eddie Rommel, who let the league in losses, and was a good "innings eater" (8th in IP) while putting up an ERA+ of 113. Meanwhile, only one "8 WARP" player gets excluded.

Meanwhile, in the NL, you see a completely different story. Eppa Rixey makes the Pro-Bowl with only 6.6 WARP. That is important for me to see, because -- while lumping all the pitchers together puts Rixey well outside the Top 10 -- he was actually one of the Top NL pitchers that year! Expanding the list would put in mediocrities like Glazner or Mitchell, who don't really deserve slots.

What the Pro-Bowl roster show me is who were the dominant members of their league at their position. It shows me, for example, that Dolf Luque was likely relatively more important to his team's success than Eddie Rommel or Stan Coveleski, even though his "raw" numbers are lower, because he was farther along his league's bell curve.

"Was Carey 30 runs / year better with the glove? That is 9 win shares per year. Seems too much - gotta go with the stick."

"to be fair, while Heilmann's 148 to 107 OPS+ advantage is impressive, Carey's 6363 to 2794 outfield put out advantage and 339 to 183 outfield assists advantage are at least equally impressive."
--
just shows how unsure we are of measuring defense, does it not? By win shares, Carey is better by maybe 10-12 runs/year. By BPro cards, it is almost 30 runs/year. I wouldn't bet my mortgage on either measure. My best guess is that the "true" answer is somewhere in between,

Phillybooster,
That interpretation makes a very big deal of rank order and, in particular re Luque and Coveleski, the difference between #2 and #5 on two different rank orders.

Yet, each rank order is derived from a quantitative rating. The rating incorporates playing time and quality and its distribution is nothing like a bell curve. The rating conveys more information than the ranking, not less.

5 pitchers keeps out the likes of Eddie Rommel, who let the league in losses, and was a good "innings eater" (8th in IP) while putting up an ERA+ of 113.

1921. (Is that the wrong year?)
Dolf Luque: 2d in losses, 2d in innings, ERA+ 106.

The rating conveys more information than the ranking, not less.

To a certain degree, of course, they are both important. My point, though, was to counter the argument that I was including too few pitchers on my Pro Bowl teams. The implicit criticism of Dolf Luque argues for fewer pitchers in many cases, not more, which bolsters my argument somewhat.

I am not sure that I agree, however, that rating is more important that ranking. With catchers, for example, we give a "catcher bonus" specifically because so few catchers reach the rating stats of, for example, left fielders.

We do the same thing across era: with pitchers, for example, giving Joe McGinnity an era bonus for being a "short career" pitcher in a short career era. If he were to put up the same stats in 1970s that he did in the 1900s, he might not make a single ballot.

Rankings do the same thing on a micro-level. Did one league use its pitchers differently from the other? Did one league's hitting "suck" points away from its pitchers? Would Luque have performed better in the AL than he would have in the NL?

One interpretation of the data (by ratings alone) is that, by skill or coincidence, 10 out of the top 11 or 12 pitchers in 1921 were in the American League. That strikes me as largely implausible, absent a compelling reason (better scouting? better use of Cubans?) but I don't see any evidence of it.

The alternative is to look at rankings, which show us that Luque was more valuable to his team that Joe Bush was to his, even though their rating was the same. Absent a more holistic way to compare across leagues, I find it more plausible that the quality was no so widely skewed.

One interpretation of the data (by ratings alone) is that, by skill or coincidence, 10 out of the top 11 or 12 pitchers in 1921 were in the American League. That strikes me as largely implausible, absent a compelling reason (better scouting? better use of Cubans?) but I don't see any evidence of it.
. . .
I find it more plausible that the quality was no so widely skewed.

It isn't time to look for a reason, yet. The WARP rating doesn't measure quality; it combines playing Time and playing Quality. The league-season correlation between T and Q may be volatile, and it doesn't even need to be volatile to generate occasional large interleague differences in the number of players with ratings above some threshold.

#34-40, including
John Murphy #34 "Anybody have an idea what Heilmann's OPS+ of 148 would be today? 135? 140?
I'm trying to see what hitter today resembles him for comparison.

David Foss #37 listed the top 20 batters 1968-2003. That list supports analogy between Heilmann and Mike Schmidt or Gary Sheffield

Does this distribution of debuts trouble anyone else?
pre1968 four (#10, 17-19)
1968-73 Mike Schmidt (#11)
1974-79 --
1980-85 --
1986-91 several \
1992-97 several / including all of #1-9

...and it doesn't even need to be volatile to generate occasional large interleague differences in the number of players with ratings above some threshold.

Indeed. Consider 2004, with rapid interleague mobility in place. Choose some high threshold for some grand stat like WS for position players. Hmm... the top 5 all seem to be in the NL. Would you want to say that Gary Sheffield is better than Jim Edmonds because he ranks higher in comparison to his own league?

Does this distribution of debuts trouble anyone else?
pre1968 four (#10, 17-19)
1968-73 Mike Schmidt (#11)
1974-79 --
1980-85 --
1986-91 several \
1992-97 several / including all of #1-9

It does a little bit. I'm not so worried about the pre-68 guys because I clipped their careers off. I'm a little worried about the 92-97 guys but most of them should have decline phases coming.

I guess it is tricky using Lee Sinin's tool. I was trying to show an "era" but it was difficult expanding the size large enough to include full-careers, but small-enough to be meaningful.

Still, where is Reggie Jackson (token 1967), Rickey Henderson, Tim Raines, George Brett, Wade Boggs, Dave Winfield?

I'll double check the list when I get home.

Here's the NL Centerfielders

Cy Williams -- 8.3
Max Carey -- 8.2
Ray Powell -- 7.8
George Burns -- 7.8

There's also very little difference at this position.

If you lump the players into populations of similar sizes, you will probably see similar properties in each population. Each team requires 4 starting pitchers, and also 4 starting 'bats' (1B,LF,CF,RF), and 4 starting 'gloves' (SS,3B,2B,CA). If you look at the league in these groups, you will probably see a few outstanding bats and gloves, quickly moving into a less distinguishable cluster, not previously noted because they played at slightly different positions.

Populations at a position of 6 starting players (+ 2 replacement level multi-player situations) are too small to really work well statistically when comparing with the much larger pitcher population.

Does this distribution of debuts trouble anyone else?
pre1968 four (#10, 17-19)
1968-73 Mike Schmidt (#11)
1974-79 --
1980-85 --
1986-91 several \
1992-97 several / including all of #1-9

I have been monitoring this trend and have been adjusting my rankings accordingly (I'm up to 1972 players as of right now).

I do not advocate Dummy Hoy as a candidate --or Max Carey or anyone else. But "Max Carey" seems a good place to note Hoy's rookie season.

William Ellsworth"Dummy" Hoy, Washington NL 1888

- rookie, age 26 (debut 25.11) - some career obstacle?
- played every game in CF (the Baseball-Reference data implies every inning in center but also shows one appearance in right)
- OPS+ 134 (career high)
team leader (* by a wide margin)
G 136 (probably 1179 innings)
PA 583
*R 77
*H 138
*BB 69
HB 11
*SB 82
*Bat .270
*Onbase .374
Slug .338

Hoy produced at stereotypical leadoff rates OBA/SLA .374/.338, with only 20 extra base hits in 218 H+BB+HB. He led the team by 104 On-Base points and 17 Slugging points.

.274 .374 .338 - Hoy
.239 .286 .327 - League average (adjusted)
.225 .270 .321 - Team runner-up
.208 .248 .271 - Team average

My limit is one exclamation mark and I don't know where to use it!

(Concerning Hoy):
- rookie, age 26 (debut 25.11) - some career obstacle?

The career obstacle was that society had a program for how a deaf person would live his life, and playing professional baseball wasn't part of that program. Hoy was schooled as was typical for the deaf, and was apprenticed in to some profession stereotypically associated with the deaf (cobbler, perhaps?) His opportunities to play baseball were limited until he was an adult, and it took some persistence (and a misunderstanding about his age) for Hoy to catch on with a minor league team.

It seems fair to assume that Hoy lost a portion of what could have been his prime to that general societal restriction of opportunities for the deaf - after all, Paul has pointed out that he was a very good player as a 26-year-old rookie. However, he doesn't have a record of sustained great play in the minors; he started late there as well.

Hoy also lost a year of major league play to the contraction of 1900. The merged Louisville/Pittsburgh team had both Hoy and Ginger Beaumont. They gave the CF job to Beaumont. Hoy played (and played well) for the 1900 Chicago White Sox, in a league not yet considered a major league.

Hoy started late, peaked late, and played until he was 40. As a possibly related item, he lived to the age of 99.

On the other hand, as a candidate, I can't see putting him ahead of Roy Thomas, and Thomas has fallen off the bottom of my ballot. Carey has a big advantage in career length over both Thomas and Hoy (and Beaumont). Defense and baserunning are strengths for Hoy, but Carey has advantages there.

Win Shares on Heilmann in the Outfield see #75

Win Shares rates Harry Heilmann nearly a match for Jeff Burroughs and Greg Luzinski; far outside and below the range of all his rough contemporaries who played many innings in the field.

--
Does this software enable linking to a particular numbered article in the blog? Not by using the following URL.

http://www.baseballthinkfactory.org/files/primer/hom_discussion/23655/P100#75/

Max Carey played 68% of his outfield time in centerfield.

For some discussion see "Ranking the HOM CFs" #44 by Paul Wendt and the prompt reply by KJOK.

Of course that thread includes more on Carey. How much do his fielding and baserunning add to his batting resume, which is mediocre for a longtime regular outfielder.

Cool info on Carey: Retrosheet has virtually complete data for the 1922 Pirates (and 1921 as well, but I didn't look into that yet). I asked Dan Fox is he would be kind enough to run EqBRR on that year, and he generously obliged. It turns out that Carey's superficially stunning 51/2 SB/CS ratio is a bit misleading, since he was also picked off ten times (including once by the catcher on first base with the bases loaded and one out!). That said, he still comes out at a terrific +9.6 runs/+0.9 wins above a player with league-average SB attempt and success rates and league-average advance rates on hits and outs. (I have him substantially higher, at 1.5 BRWAA1, because I didn't have the pickoff info). No way to know whether that was fluky or a consistent pattern over his career. His teammate Carson Bigbee was +6 (despite a poor SB showing, he was terrific at advancing on hits and fly balls), and Frankie Frisch comes out around +4-+4.5 (with 21 Giants games missing from the record). On the flip side, the Pirates' Charlie Grimm was an atrocious -9, adding a bunch of pickoffs to his 6/10 SB/CS line and stinking it up on the other types of advancement as well, while Charlie Hollocher was close behind at a heinous -8, although that much is pretty much evident from the 19/29 SB/CS line. (In fact, I was already giving him precisely -.8 BRWAA1).

He also gave me some limited data for 1911. What's clear is that the Giants won the NL pennant in spite of McGraw's managing, as he just about ran the team out of an inning every game. Josh Devore was a ghastly -10, and Fred Snodgrass was right behind at -6. The best runners were Miller Huggins (+6/+.7 for the half of St. Louis's season that's available), Buck Herzog (+5 in a half-season), and Jimmy Sheckard (+6 with nearly all the games counted).