I agree with Andrew and karlmagnus that O'Doul should have had a thread a while back.

What's his story? Why the great seasons in his 30s and nothing else?

I think this is common knowledge, and it's not exactly germane to our purposes, but IIRC, Lefty O'Doul was recognized on both sides of the Pacific as one of the major ambassadors for baseball in Japan, especially after the war.

Bold= Led League

Underline = Led Team

1924 -- Salt Lake Cit Bees

.392/ 163 H/ 31 2B/ 4 3B/ 11 HR

(would have tied for batting titled if qualified. Teammate Duffy Lewis also his .392.)
(7-9, 6.54 in 128 IP)


1925 -- Salt Lake City Bees

.375/ 309 H/ 63 2B/ 17 3B / 24 HR

1926 -- Hollywood Bees

.338/ 223 H/ 29 2B/ 3 3B/ 20 HR

1927 -- S.F. Seals

.378 BA/ 278 H/ 43 2B/ 4 3B/ 33 HR(tie)

309 hits, even in the PCL, looks HOM-like to me. I think we need our conversion experts on this one.

How many games?

I don't have time to type his career minor league record in (homework), but here is some quick info.
He started as a pitcher at age 20 in the minors.
At 21, he went 12-8 with San Francisco PCL.
At 22, in 1919 and then in 1920, he is listed as being with the Yankees as a PH / P. In 2 years, he pitched 5 times and got into 32 games total. He only had 1 PO so I assume he never played the field.
1921, at 24, he has a full year pitching with SF in the PCL again.
1922, again with the Yankees, pitches in 6 games and pinch hits 2 times (?).
1923, age 26, with the Red Sox, relief pitcher and outfielder.
1924 - 1927 - outfielder for Salt Lake, Hollywood, and San Francisco in PCL.
1928, age 31 - in majors full-time outfielder for Giants
1929-1930, ages 32 and 33, majors full-time outfielder for Phillies.
1931-1932, 34 and 35, majors full-time outfielder for Dodgers.
1933, age 36, Brooklyn and Giants
1934, age 37, half time outfielder for Giants.
1935 to 1951 manager and pinch-hitter for San Francisco PCL.

Oh, quick games played totals for PCL 1924-1927:

1924: 140 including 31 games at pitcher
1925: 198 as outfielder
1926: 180 as outfielder
1927: 189 as outfielder

From putout totals, it looks like he was a corner OF in the minors. Looks like a strong arm from good assist totals, but not accurate because high error totals.
I'll post more tonight.

We should have looked more closely O'Doul - obvioulsy he was an outstanding hitter, and like Cravath, several of his best seasons were hidden in the minors.

A warning though - PARK EFFECTS!

Salt Lake City was Colorado-like in its effects on hitting statistics. I don't have PCL team data for 1924-25, but for 1923 I estimate Salt Lake City's run environment to have been 6.73 runs per game, 30 percent above the PCL average and 40 percent above the major league average. For 1922 SLC's run environment wasn't quite as outlandish - 5.45 runs per game, which was 17 percent above the PCL average and 12 percent above the major league average.

On the other hand, for 1926 Hollywood's run environment was quite low - 3.95 runs per game, 12 percent below the PCL average and 15 percent below the major league average. For 1927, my estimate of San Francisco's run environment is 5.45 runs per game, 10 percent above the PCL and 15 percent above the major league average.

In 1921 O'Doul may have been the best pitcher in the PCL. He was 25-9, with a 2.39 ERA in 312 innings. (One pitcher had a lower ERA, Aldridge of LA, 2.16 in 283 innings, and I believe O'Doul was tied for second in wins behind Dell of Vernon, who was 28-14.) Unfortunately, the rest of his pitching career wasn't as impressive.

I was a bit surprised at the low run environment for Hollywood for 1926. I just checked an article on sportshollywood.com about LA's Wrigley Field, and it says Hollywood played its home games there from 1926-35, sharing the ballpark with Los Angeles. Wrigley generally shows up as a hitter-friendly environment.

This is a great case for those who say the PCL conversion rate is .95 or .90 or somewhere in that region.

The point being: at that rate, O'Doul's age 21 and 24 pitching seasons are worth some sort of MLEs. Yet at age 22-23, 25-26 he was in the MLs and accumulated almost no value at all.

If I'm inclined to give O'Doul, oh, let's say just for example, I'm makin' this up, 5 and 15 WS for his MLE for age 21 and 24, then I can also say that if he had been used differently (better) while in the MLs at age 22-23, 25-26, he had the ability to rack up 5 to 15 WS each of those years, too. If only he'd gotten the chance.

Obviously my point is that MLEs for MiLers make for a remarkably slippery slope. The logic is that he (whomever) was stupidly denied the chance. Well, if being "kept" in the MiLs represents being stupidly denied a chance, then why is being kept on the ML bench not the same thing?

The problems with this are many. We are evaluating ability, not value. We are substituting our judgment for that of people who were there, on the ground, at the time, and saw him (whomever) play, and came to a different conclusion. We are making up scenarios that didn't really happen.

Now, none of these is really wrong, just problematic. I mean, we use them all with NeLers, well, except for the second (substituting our judgment) because that didn't happen. Nobody made the judgment that the NeLers couldn't really play in the MLs, they made the judgment that they were black.

But they are indeed problematic, and vastly more so with MiLers than NeLers. So I say, again, why not MLEs for college or high school ball, or lower MiLs, etc. etc. Once you're on the slippery slope, why not slip all the way?

Besides, in O'Doul's case I only see 4 years of MLE caliber play. I will have to be shown here that his age 21 or 24 years or his 1935-51 years really add up to something. It's true that those 4 years gets him up to 10 seasons as a "regular" hitter, and let's even assume that his OPS+ stays at (or even goes up a little? from) 142. Or even add in all 700 games (2800 PAs?) for a total of 6,000 PAs. Is that enough?

And just to conclude, I have decided to give Gavy Cravath some "extra credit." With Lefty, it's not a question of "giving extra credit," it's what does it add up to?

If you add four ML seasons 1924-27 with about 200 hits each and an OPS+ of 150, O'Doul is around 2000 hits and clearly above Hack Wilson and Chuck Klein. It looks to me as though these seasons at least deserve credit, but what do our experts think?

Nobody made the judgment that the NeLers couldn't really play in the MLs, they made the judgment that they were black.

As the sole remaining Dolf Luque voter, I just wanted to say that I'm giving him extra credit for both judgments.

If there is one thing that drives me nuts, it is the ignorant assumption that players in the Minor Leagues from the 1890s, 1900s, 1910s, 1920s, and even 1930s were there simply because the Major League teams decided that those players were not good enough to play in the Majors.

That's just stupid. The relationship between the Majors and the Minors was totally different then, than it is today. Back then, Minor League teams, especially the top ones, had the power, through the reserve clause (although slowly eroding) to keep their players indefinitely.

There were no modern farm systems that funneled talent upwards (though Rickey was scheming).

Lefty Grove did not stay with Baltimore from 1920 to 1925 because he wasn't Major League caliber. He stayed with Baltimore because he was extremely valuable to Baltimore and the reserve clause kept him there. He stayed with Baltimore because no Major League club would pay the price Baltimore set for Grove's services.

This same principle works for Gavy Cravath and Lefty O'Doul too.

The economic relationship between the Major Leagues and the top Minor League clubs then was much more like today's relationship between a club like the Yankees and a club like the Royals would be today. Players were not 'kept' in the Minor Leagues.

But instead of realizing this, the basic approach seems to just be to denigrate these players' Minor League service.

Lefty O'Doul in the Majors:
AGE-YEARS-WS
31 1928 12
32 1929 31
33 1930 20
34 1931 22
35 1932 33
36 1933 17
37 1934 9

Lefty O'Doul, like a lot of guys, messed up his hitting career by starting off as a pitcher. He pitched from 1917, age 20, until 1924.

He could definitely hit. In 1921, the 24-year-old O'Doul hit .338 with a .502 SA in 136 ABs in the PCL. When his arm went dead in 1924, he finally started his career as a regular player.

In the PCL from 1924 to 1928, O'Doul hit like this:

AGE-YEAR-G-AB-BA-SA
27 1924 140 416 .392 .565
28 1925 198 825 .375 .579
29 1926 180 659 .338 .483
30 1927 189 736 .378 .576

A major league player's career (sans steroids) usually goes in a typical pattern. The player
improves before 26, peaks from 26 to 30, and then fades away from 31 on.

It is quite likely that O'Doul, in 1925 and in 1927, had MLE seasons of 30 or so Win Shares. But Sunnyday above, in a perfect example of Pre-World War 2 Minor League discrimination, deigns to credit O'Doul with, oh say, 15 WS seasons for those years.

The basic argument here, which is ridiculous, is that it's impossible for a Minor League player of that time to have a season of far above Major League average quality in the Minors.

This leads to the absurd conclusion that all these guys who started their Major League careers late, in a complete reversal of the normal aging pattern of baseball, were at the peak of their careers in their mid-30s (Cravath at 35, O'Doul at 35).

The truth is that these players, were, just like most baseball players, at the peak of their talent at the same time as everyone else and their Minor League stats prove it.

If O'Doul had gotten a Major League job, rather than a PCL job, as an OF in 1924, his career would have probably looked like this:

AGE-YEARS-WS
27 1924 12
28 1925 28
29 1926 23
30 1927 33
31 1928 24
32 1929 34
33 1930 22
34 1931 23
35 1932 33
36 1933 17
37 1934 9

With a comparatively late peak at 32 because he started his Major League career late due to wasting his time pitching.

I don't advocate O'Doul for the Hall of Fame or Merit based on his hitting stats. There are a lot of guys (Wally Berger, Hack Wilson, Chuck Klein, Buzz Arlett) with very similar cases.

But to denigrate the Man simply because he spent his prime in the PCL is not right.

Gad, I don't know how I could have stated it any more clearly that I was hypothesizing 5 and 15 WS for O'Doul's age 21 and 24 pitching in the PCL. I said absolutely nothing about his 1925 (age 28) and 1927 (age 30) seasons. How did you get from 1918 and 1921 and age 21 and 24 to 1925 and 1927 and ages 28 and 30? Certainly not by actually reading and comprehending the post you deigned or should I say condescended to comment upon. Jeez.

16. Sunnyday-

When reading your post, I took the 25-26 to be 1925 and 1926. Upon rereading it, I admit my error. Admittedly I skimmed over it because I had a problem with your post (and several other posts of yours on the same subject). But the problem I had with your post was not that at all, it was these TWO statements:

1. "The logic is that he (whomever) was stupidly denied the chance. Well, if being "kept" in the MiLs [Minor Leagues] represents being stupidly denied a chance, then why is being kept on the ML [Major League] bench not the same thing?"

The claim that these players (Cravath, O'Doul) were being 'kept' in the Minors is ignorance of baseball history on a grand scale.

2. "We are evaluating ability, not value. We are substituting our judgment for that of people who were there, on the ground, at the time, and saw him (whomever) play, and came to a different conclusion. We are making up scenarios that didn't really happen."

In evaluating Minor League performance, you are very obviously evaluating value, not ability. This statement is simply wrong.

As for the subject of MLEs for Minor Leaguers, I have read several of your posts and you basically dismiss them with sarcasm and irrelevant arguments.

Your sarcastic 'slippery slope' argument about giving credit for Major League bench time is just irrelevant. There is no performance to evaluate.

Your sarcastic and irrelevant argument that: 'if MLEs should be made for Minor League ball, why not College and High School ball?' is also completely irrelevant.

No one has ever argued that here and, if they did, would probably be laughed down.

The particular argument here is this: before farm systems took over (in the 1930s and 1940s), it was quite possible for a player of Major League caliber, even star or superstar caliber, to have a long career in the HIGH minors.'

This is irrefutable.

Your straw man arguments about high school ball, college ball, Major League bench time are all irrelevant.

Lefty O'Doul, after a failed pitching career, was a star Major League caliber outfielder from 1924 to 1934. He spent 1924 to 1927 in the PCL and 1928 to 1934 in the National League.

Of course, there are two ways to evaluate this.

First, you can simply do MLEs for his Minor League time and add them to his actual Major League stats. However, this creates the inevitable adjustment issue in the timeline (In other words, the player is being penalized twice, sometimes thrice, for adjusting to the Majors).

Second, you can try to adjust for that fact. Of course, this creates the inevitable timeline problem that the player would be getting more credit than he actually had in the real world, but gives a truer outline of the career he would have had just playing in the Majors.

I prefer the second approach because it gives a truer indication of the player's real value as compared to other Major League players than the first approach.

But I understand why someone would prefer the first approach.

Perhaps I got a little too testy at your post, but irrelevant straw man arguments and sarcasm always tick me off.

SunnyDay-

On second thought, I will simply apologize for being testy. Guess I just needed to rant.

If a player is not 'allowed' to leave his PCL or AA or IL team then isn't he being 'kept' in the minors? Or are you insinuating that Arlett, Cravath et al wanted to be playing in the PCL instead of MLB?

Schmeg-

Not sure if you were addressing me, but I assumed the term 'kept in the minors' was used in the modern sense. In other words, kept in the minors meant that a Major League team could have brought O'Doul or Arlett or Cravath up to the Majors but decided not to.

This, of course, was not true.

All three of these players had signed a reserve clause contract and had no say over where they played or whether they wanted to play there. They were the property of their Triple-A teams.

Triple-A players of this time could only leave their teams if the Triple-A team itself decided to sell or otherwise get rid of them (there were some rudimentary draft rules, but they were easily gotten around unless, as in the case of Cravath, someone screwed up).

I meant what you said there. That there teams wouldn't allow them to go play MLB ball. Therefore they were 'kept' in the majors.

I think that Sunnny's point is that in some ways this possibly unfortunate set of circumstances could be preferable to that player in terms of HOM worthiness. Some of these guys could have been on MLB benches accruing no stats and instead their teams kept them in the Minors, or MLB scouts had some odd reason not to want them in MLB, or they wanted to be on the west coast, or for whatever reason there could possibly be. Not sure I agreee with this stance but I don't think he was misinterpreting baseball history by a large degree.

I still think there is a valid concern/issue here that has not been answered. Obviously I have not been able to express the concern clearly, my bad, here is another tack.

Model A

WS, WARP, etc., posit that value in baseball can be calculated and it equals X. ML seasons = 1X. Other seasons = 0X. (You can calculate WARP or WS for MiL or college or whatever, but those seasonal calculations are never factored into a player's career value, for example, so in that sense they have a value of zero times X.)

Insight

The HoM has rightfully determined and created a consensus that other seasons have value--certainly the NeLs, and now, also, MiL seasons are posited as having value. Hooray for us, I am on board with this.

The question, then, is how much value do such seasons have?

Model B

It has been suggested that AAA seasons, or at least those in the PCL for the era we are now evaluating (the era when players we are now evaluating were active) have a value of .9X or even .95X. This model, as expressed to date, however, posits a value of 0X for all levels of play below AAA (or below the PCL).

Insight 2

If it is arbitrary to say that all value is ML value, it seems arbitrary, too, to say that all value is ML or AAA/PCL value.

Model C

This is what I tried to express earlier--i.e.

ML value = 1X
AAA value = .9X
AA = .8X
A = .7X
B = .6X
.
.
.
College = .2X

I am not arguing whether these are the right numbers, just explicating my model.

Analysis

This is not even so much an argument that Class B or college value should fold into career values for future ML players. It is a way of asking, again, whether .9X is the right AAA value.

It does seem reasonable to me that if .9 is the right conversion for AAA--or, especially, if .95 is the right conversion for AAA--then self-evidently at least some lower levels also have real value for our purposes.

If on the other hand you recoil from the idea that AA play has real value for our purposes, then I can't believe that you would argue for .9 or .95 for AAA. I mean, how much better is AAA than AA? AA than A? And etc?

Insight 3

Finally, the insight that Lefty O'Doul specifically suggested to me what that if .9 or .95 is the conversion rate, then many many many players would accumulate far more value in AAA than they would ever accumulate in the MLs.

I don't have O'Doul's record in front of me, I'm sure I should, but I don't. But so he went 12-8 in the PCL in 1918 at age 21. At .95 this is worth at least 9 WS. The following two years he is in the ML and earns, perhaps, somewhere in the range of zero to one or two WS.

Then in 1921 at age 24 he wins 20 games in the PCL, worth (guessing) 18 or 19 WS. The following 2 years he is in the ML again and earns maybe 5 or 6 WS.

Will the real Lefty O'Doul please stand up? There is something wrong with a conversion factor that makes AAA players out to be better than ML players, when they are exactly the same players.

Problems

1. The conversion is too high? Or,

2. There is some other conceptual problem with the whole model of integrating ML and "other" values into a single analysis. Such as, borderline players don't get enough PA or IP in the ML whereas they get lots of PA or IP in the MiLs. How to deal with that?

That's it.

But so he went 12-8 in the PCL in 1918 at age 21. At .95 this is worth at least 9 WS. The following two years he is in the ML and earns, perhaps, somewhere in the range of zero to one or two WS. Then in 1921 at age 24 he wins 20 games in the PCL, worth (guessing) 18 or 19 WS. The following 2 years he is in the ML again and earns maybe 5 or 6 WS.

In my view, for Negro League/ Minor League play, the first question is "Did he play baseball?" and the second question is "How should that be coverted to major league play."

In 1919 and 1920, he was in the major leagues, sure, but he wasn't playing baseball. He was making sure the bench didn't fly up and smack someone in the face when Wally Pipp got up to bat. So, in my view, O'Doul's years with the Yankees were worth much less than his years in the PCL, because in the PCL he was actually PLAYING BASEBALL rather than serving as a paper weight.

sunnyday2,

I think you lay out the problem very clearly.

I don't have a solution, but it strikes me that we might _narrow_ the problem by focusing on the range of values between replacement level and average.

A non-ML player whose MLE value falls below replacement level obviously shouldn't be receiving credit, even if one's metric gives value to below-replacement play. There are obvious exceptions to this criterion in injury seasons or end-of-career seasons for a very good player.

A player whose MLE value would be above average in the majors obviously deserves full credit for his performance: were he to have been able to play in the majors, any efficient use of talent would lead him to be a starter in some capacity.

In between those two thresholds, though, we have the O'Doul sort of case, where a player might be good enough to _play_ in the major leagues but not good enough to _start_ (or be in the staring rotation) in the major leagues, especially for a very good team, whose bench players might actually be better than the starters for a very bad team.

Fortunately for our purposes, few players that we look at as serious candidates spend much time in this gray area, but I'd suggest that it's here that the most difficult problem rises.

It's a problem, I note, for measuring career value, not peak value.

PS. To complete my thought, I should have mentioned the other issue, which is selectivity. I mean, it's one thing to look at Earl Averill or Lefty O'Doul, though even that can be problematic. But it's an entirely different thing to be voting for Buzz Arlett.

The point is, if the conversion is .9 or .95, I can easily imagine a dozen or more career MiLers who might have racked up in excess of 300 MLE WSs. I mean, all that would take is 15 years @ 25 WS X .9 or .95 ? 337.5. How rare can this kind of career really be?

John said elsewhere that we couldn't very well declare a moratorium until we had the time to explore every career MiLer. I disagree. Of course, I only speak for myself, but I have a moratorium on Buzz Arlett until there's at least been some level of good-faith effort to determine whether he is the best career MiLer, or not.

Of course, if the right conversion factor is really .75 then it is probably moot. None of the career MiLers is a contender. But I don't think you can have it both ways--i.e. .9 or .95 for Buzz Arlett and not .9 or .95 for Ike Boone and Nick Cullop and Joe Hauser, etc. etc. etc.(not .9 for Ike Boone in the sense that nobody is willing to test for him).

Now OTOH if somebody who really believes in Buzz Arlett was willing to test the other 10 obvious career MiL candidates, and if he turned out to be #1, then I'd be willing to consider him, just as I have (belatedly)come around on Gavy Cravath.

Then, as I said, to a lesser degree, I worry about Earl Averill's ML + MiL MLE. His ML + MiL might be greater than Joe Blow's ML, but are we sure it's greater than Joe Blow's ML + MiL MLE? Not until we've looked at Joe Blow's MiL MLE, we're not.

It's just a little nagging worry that keeps me from getting excited about those candidates whose case depends to any degree on that MiL MLE.

The conversion factor of .9 or .95 applies to the stats, not to the Win Shares, whcih must have a much lower conversion factor. Batting .330 doesn't get you 10% more WS than batting .300, it gets you 50-100% more.

Since this is true all the way down, it becomes very difficult to rack up serious MLE win shares in Single A.

I agree with Chris Cobb about the need to compare with replacement level (more than half of all Triple AAA regulars would be below major league replacement level).

I also agree with Karlmagnus that a different conversion factor applies to win shares. My guess (I'm just making this up) is that for win shares the formula would look something like

ML WS = .8 * AAA WS - 10

So a player near the AAA average would translate to major league replacement level or below.

The case of Lefty O'Doul sitting on the bench with the Yankees is a good example of why I think career value is a less reliable measure than peak or prime value. I know this may seem counter to our intuition, since career value is based on a larger sample of seasons. But there are just too many random factors over which a player has no, or limited, control that can cause one player to have 3 or 4 more seasons as a regular than another player who is equally good. I think we generally can measure peak and prime value much more reliably than career value - especially since no one really knows precisely where replacement value should be set prior to the development of farm systems. I give only a small weight to career value (and set my replacement level quite high), and base my ratings more on peak and prime. Consequently, I don't worry a lot about how exactly many seasons of minor league credit a player gets, because only the truly outstanding ones are going to matter much in my ratings.

I should also add that we never actually calculate minor league win shares and convert - rather, we convert basic statistics to major league equivalents and then calculate win shares.

Sunnyday2 wrote:

To complete my thought, I should have mentioned the other issue, which is selectivity. I mean, it's one thing to look at Earl Averill or Lefty O'Doul, though even that can be problematic. But it's an entirely different thing to be voting for Buzz Arlett...

John said elsewhere that we couldn't very well declare a moratorium until we had the time to explore every career MiLer. I disagree. Of course, I only speak for myself, but I have a moratorium on Buzz Arlett until there's at least been some level of good-faith effort to determine whether he is the best career MiLer, or not.

Of course, if the right conversion factor is really .75 then it is probably moot. None of the career MiLers is a contender. But I don't think you can have it both ways--i.e. .9 or .95 for Buzz Arlett and not .9 or .95 for Ike Boone and Nick Cullop and Joe Hauser, etc. etc. etc.(not .9 for Ike Boone in the sense that nobody is willing to test for him).

I understand your concern that we may be missing equally qualified candidates. To some extent, I think we all have a similar concern about candidates from the Negro leagues, where there aren't comprehensive data--especially for the periods when the leagues in the U.S. were weak and many top players were going to Cuba, Mexico, or the Dominican Republic to play. How can we really be sure we're fair to everyone and not missing a more qualified player?

One way, obviously, is to do conversions for more players. I'd like to do them for Boone and Hauser and several others. But calculating MLEs takes quite a bit of time, so obviously we won't get to everyone. We need to be selective and focus on the few players who are likely to be among the best.

What should we look for? A lot of the characteristics that Arlett and Cravath have: A long career spent in the majors and/or highest minors; power hitters; good on-base percentage. I didn't start with Buzz Arlett randomly - several experts have pointed to him as perhaps the greatest minor league star.

And if we miss a few players we should have evaluated, so what? The whole HoM exercise is designed to learn things about baseball's history that we didn't previously know. In 10 or 20 years, thanks to researchers like Gary A and Gadfly I expect we will know much more about the Negro leagues than we know now. Once minor league statistics get put into databases and can be systematically analyzed, we will have a much better handle on minor league players. But we're not holding off on this project until the data are perfected. We're making the best choices we can based on what's available and the best analysis we are able to perform, given the limited time we each have to spend on this activity.

But we're not holding off on this project until the data are perfected. We're making the best choices we can based on what's available and the best analysis we are able to perform, given the limited time we each have to spend on this activity.

Exactly, Brent. Besides, why should Arlett (who I don't have high on my ballot, BTW) get penalized because we haven't looked at every single player? If there were better players than Arlett in the minors, so what? Obviously, it would be nice to know about them, but whether we ever find them or not is not going to change my placement of Arlett on my ballot, regardless.

Nobody is obligated to do anything they don't want to do. I'm not going to do MLEs on the next ten career MiLers.

But as long as nobody does the work then, yes, we can always say, "So what?" because we'll never know what we don't know. But life, as they say, is unfair.

The other half of the equation, however, is that for those who are reluctant to support Buzz Arlett, there are two things that might reduce that reluctance, and both of them would be outputs of some further study of career MiLers.

1. Further testing the conversion factors by cases.

2. Building a comfort zone that Arlett really is the best of the breed.

Sans some further progress on these two fronts, I can't see supporting the Buzzer.

Most of this simply seems to me to be a tempest in a teapot.

A couple of points:

1) Lefty O'Doul's three seasons wasted on the Major League bench as a spot and warm-up pitcher are interesting but basically irrelevant. There are very very few players, basically none, of any merit who have anything like this problem.

The only two instances, other than one season lost as a bench warmer, that come to mind are McCovey and Miguel (Mike) Gonzalez. But McCovey, of course, was platooned and could easily be given credit for that. Gonzalez, on the other hand, spent his prime as the New Giants' bullpen catcher. But he was hardly a superstar.

The bench warming argument basically boils down to the same thing as credit for college play. it's just irrelevant.

2) While MLE for Negro Leaguers or, say, Japanese players are one thing, Minor League MLE's are a whole different ball of wax. There really is a much smaller window here.

No Hall of Fame worthy player has been stuck in the Minors since the Minors came completely under Major League domination. A process that started in 1903 and was pretty much complete by 1950.

[Hector Espino may be a lone exception.]

So basically, from the start of Minor League play in 1878 to 1950, there was a window where some players, of Hall of Fame caliber, could get stuck in the Minors. And, basically, all of these players, that did get stuck, got stuck in the high Minors: AA, IL, and especially the Pacific Coast League.

In fact, if one was to construct the perfect model for a forgotten Minor League superstar, he would probably fit the following profile:

1) Started career as pitcher.
2) Was one hell of a hitter.
3) Played in the PCL.

Lefty O'Doul, though he got his shot in the Majors, fits this profile. Buzz Arlett, of course, is the absolute prototype for the profile.

There are not a lot of Minor League superstars. SABR actually published a book that has pretty much all of them in it.

There is really only somewhere between 10 to 20 of these guys to evaluate (disregarding guys like Earl Averill or Gavy Cravath who had substantial Major League careers); because, basically, if you were a Major League star caliber player and were not a little too tanned, you were in the majors before 1950.

From everything I know, the two guys who should be considered the top of the heap are Buzz Arlett and Perry Werden, granting that there are some guys (Ike Boone comes to mind) who may have been even better hitters, but not by much, in shorter careers.

{Which reminds me, did Perry Werden ever get a thread? I think of Werden every time I see those elect Beckley posts. I think Beckley would get Werden on career, but certainly not peak.)

"The only two instances, other than one season lost as a bench warmer, that come to mind are McCovey and Miguel (Mike) Gonzalez."

Wouldn't Elston Howard and Jorge Posada fit into this category? Both were kept on the bench and if they had not might have very good HOM cases instead of merely interesting ones.

Schmeag-

Howard and McCovey are guys who didn't get full playing time, but they were playing enough to be evaluated as if they had (Howard also has some Negro League and Military Service issues) but that's a different story).

I don't see Posada at all, he had to fight for playing time a little, but none of these three guys have what O'Doul does.

I stated above three seasons, but it's actually four. Four seasons of virtual inactivity.

YEAR AB IP
1919 16 05
1920 12 03
1922 09 18
1923 35 53

I'm surprised his career survived and these's really nothing like it that I can think of other than Mike Gonzalez' time on the bench as John McGraw's scout, fourth catcher, Cuban contact.

One of the things I like best about here is that I learn something new all the time. As far as I know O'Doul wasn't injured, he was just inactive. It seems pretty unique.

Which, of course, is why it's relevence is simply minimal, although really fascinating.

Gad,

I too am anxious to learn about Hector Espino.

But I don't think the questions about O'Doul are irrelevant. Maybe to O'Doul's case, if you think he doesn't have one. But as a test case of a methodology.

If Lefty could earn 5 WS (just a guess) in the PCL at age 21 and 15 more (just a guess) in the PCL at age 24, then it would stand to reason that he would be worth 5-15 WS or more at age 22-23, 25-26. But he was not.

That raises questions about how to convert PCL value to ML value, at least in my mind.

And that in turn, raises questions about the Buzzer. O'Doul is extremely relevant to Arlett's case.

There are not a lot of Minor League superstars. SABR actually published a book that has pretty much all of them in it.

If anyone on the list has ready access to a copy of this book, could you post a list of the players profiled in it? It'd be handy, and I must confess to being quite curious about this unusual group.

sunnyday2 wrote:

If Lefty could earn 5 WS (just a guess) in the PCL at age 21 and 15 more (just a guess) in the PCL at age 24, then it would stand to reason that he would be worth 5-15 WS or more at age 22-23, 25-26. But he was not.

That raises questions about how to convert PCL value to ML value, at least in my mind.

But doesn't the same issue arise with major league players? I assume that a number of the regulars for the 1930s St. Louis Browns would have been sitting on the bench if they had played for the Yankees. Few HoM candidates spent much time on the bench, but several players (Max Carey comes to mind) were probably able to extend their careers a few more seasons than would a player on a more competitive team.

It seems perfectly natural to me to say that O'Doul was worth more playing for San Francisco than for New York, because SF actually used him. I'd bet that this difference in value was also reflected in prices paid for players during that era. That is, my guess is that an analysis of player trades and sales would show that minor league teams of that era paid more for players they intended to play regularly than the other major league teams paid for bench players.

There are actually several different editions of that book, Chris, each of which has different players in it. I have (I think) volume 3, the green one with Chuck Connors on the cover.

There's actually a better book with comprehensive year-by-year stats on the best minor league players... but for the life of me, I can't recall the title or author. I believe it was published by Baseball America about 10 years ago.

OK, I found in my files a Xeroxed page from the book I referred to in #38... it's called The Minor League Register, by Lloyd Johnson. Was originally published by Baseball America in 1994 and apparently reissued in paperback in 2000 (although I have not seen that edition). Anyway, it's definitely a superior reference source to the SABR books of minor league stars. It also has biographical info on each player, in addition to notations such as records of note, MVP awards, all-star teams, etc.

SunnyDay-

I always wondered if Espino was a Major League caliber star or just a good slugger who decided to spend his whole career in Mexico. But, since he retired in 1984, it'll be a while for the Hall of Merit to get to him (I think I also read that he passed away relatively recently but could be mistaken).

As for your comments on O'Doul, I'll try to express my thoughts logically (without ranting and raving for no reason).

1) Methodology

I don't think O'Doul's case raises issues about the methodology. The basic methodogy is to find the different strengths between two leagues (in this case, the National and the PCL), and then use the difference to express a player's statistics as if he had played in the other league.

There are problems with this approach, park effects, playing time, adjustment factors, etc. But the basic method, which is a simple conversion, seems to be completely sound.

2) Lefty O'Doul's Inactive Years

Of course, one thing I champion is adjustment factor. In other words, when you adjust one player from one league to another, you have to allow for the time it takes the player to adjust. Otherwise the player is being penalized for having to adjust and will look worse than he actually is (my whole Cravath thread thing).

But some people are uncomfortable with this, though I think it's totally logical, because it is not a straight conversion. In other words, it begins to stray from reality, i.e. what actually happened.

Giving O'Doul credit for his inactive years is a step beyond even this. For one thing, he was inactive, there is no conversion to do. It would be kind of like giving Pete Reiser credit for his lost career. It is totally possible and logical that O'Doul could have converted to a hitter in the Major Leagues in 1919 and had a completely different career. But it did not happen.

A conversion that gave him credit for this would be opening the door to too many conversions about what would happen if Tony C didn't get hurt or if Bill Terry didn't quit baseball to go work for a couple of years before coming back, etc. etc.

In post 37, Brent has an excellent take on what a can of worms the approach of giving O'Doul credit for time he's inactive would be.

3) Buzz Arlett

None of this, as far as I am concerned relates to Buzz Arlett. Arlett, as he proved beyond a shadow of a doubt in 1931, was a star caliber Major League hitter. He played full time in the Minors, mostly the PCL, for 20 years or so and straight conversions can be done on these stats. The only real discussion is what those conversion rates are.

Chris Cobb-

Eric Enders has it. The Minor League Register is actually a compendium of the three SABR books with some other guys thrown in. There are probably more than 700 guys in the book with complete career stats, so listing them isn't possible.

Of course, a lot of these guys had Major League careers and a lot more are just in there for fun like football player Ken Strong, but a list of the very best of them who 1) played in the High Minors (eliminating the Joe Baumans of the world) and 2) didn't really have Major League careers of note would not be long.

If I am not mistaken Sunny is not advocating credit for Lefty O'Doul because of the years he spent on the bench, but instead saying that maybe some credit shouldnt' be given to buzz Arlett and others because it is enitrely possible they could have been on a bench had their teams let them go to MLB. Something like that.

Also, hasn't every player ever had an adjustemnt period to MLB? As far as I know there aren't many players whose first game of baseball came in MLB, so there would have to be an adjsutment period. Those that didnt' have much of an adjustment period (Ichiro, Pujos for modern examples) are just freaks of nature who desrve credit for being such. Giving Cravath credit for this period seems to be giving him some extra credit I am not sure he deserves.

Schmeag-

I think we're saying the same thing here, just differently. The contention that Buzz Arlett should not be given MLE credit for his actual play in the PCL because he may have simply stayed inactive on the bench in the Majors, using the single example of O'Doul, is pretty silly.

A hitter of the caliber of Arlett would have forced himself into the line-up. O'Doul is just unique. Can you name any other hitter of his (O'Doul or Arlett) caliber that sat on a Major League bench, pretty much TOTALLY INACTIVE, for four years (or three or even two)?

This goes without mentioning the most important fact, which is that O'Doul was not a hitter but a pitcher during those four years. O'Doul's career is simply odd, odder, and oddest in this respect and can't be used to generalize anything.

Like you, I am not really sure sometimes how or why SunnyDay is taking O'Doul's extremely odd career path and applying it to question Arlett or MLEs; but I am pretty sure that it is wrong and inappropriate as an argument.

As for adjustment periods, the argument is simple. Players obviously take time to adjust to whatever League they enter. A player's career and career evaluation can be extremely effected by this problem, but every player is different.

In Cravath's case, there is a huge problem with this as he bounced around from League to League quite a bit. I am sure that you will admit the probability that Cravath's career, if he had simply played the whole thing in the Majors, would be quite differently shaped.

However, the choice to make an allowance for this is really personal preference. I make an allowance for Cravath because what happened to him is very much a result of the time (and that timeframe's Major-Minor League relationship) in which he played. It would never happen to a hitter of his caliber today.

There are very obviously other players that fit this criterea too (most obviously some of the Negro League guys like Jackie Robinson and Monte Irvin).

As for Ichiro and Pujols, as I said every player is unique. The adjustment differences between Ichiro and Hidecki Matsui are very interesting and fascinating to me. Pujols, on the other hand, is simply on a classic Hall of Fame career path with two things really standing out:

1) Is his actual age correct?

2) Has any hitter ever improved as much as he did between his last (only) Minor League season and his first Major League season?

In single A in 2000, Pujols hit
.324-.389-.565 in 395 ABs.

At the end of the season, he got a shot at AA
.284-.341-.481 in 81 ABs, and AAA
.214-.267-.286 in 14 ABs.

Then in 2001, in the Majors, he went
.329-.403-.610 in 590 ABs.

So basically the man hit better in the Majors than he did in A ball just one year before. As far as I know, that's unique and also simply odd, odd almost beyond belief.

But just because a player has an odd or unique adjustment period doesn't mean that it can be completely disregarded. The odd (Pujols) and weird (O'Doul) really don't tell us much about the normal and expected.

I guess I just think that every player has an adjustment period except for a few and that Gavvy Cravath would have had one as well. Therefore I dont' give Gavvy credit for that part of his carer like you do.