Negro Leagues = full credit
Military Service = full credit
Unfairly Blacklisted = full credit
Pre-farm system minors = partial credit
Major League season shorter = yeah - extrapolation
Reasonable hold-out = partial credit
Got sick and died = partial to full credit
Nasty non-baseball illness = partial to full credit
Nasty baseball-related credit = little to no credit

I'm an easy grader ;D

Negro Leagues = full credit
Military Service = full credit
Unfairly Blacklisted = full credit
Pre-farm system minors = partial credit

Only above average work should be acknowledged.

Major League season shorter = yeah - extrapolation
Reasonable hold-out = partial credit

I'll say no here, but I could be persuaded otherwise.

Got sick and died = partial to full credit

No here

Nasty non-baseball illness = partial to full credit

None from me.

Nasty baseball-related credit = little to no credit</i>

Zip credit from moi.

The parameters are too simplistic. The greatest determining factors revolving around these conditions are what type of season the player had immediately before and/or after said condition, and at what age it occurred. There is no slippery slope. It's very simple. If they had a standout adjacent season, then you extrapolate conservatively with partial credit. To frame this as a simple yes-no proposition doesn't do it justice, but the original question itself is a valid one.

There's no danger of falling into a pit of having to credit Mantle for his nagging injuries, or Puckett or Belle for theirs. In Mantle's case, he didn't miss the majority of any seasons during his prime. In Puckett and Belle's cases, they didn't miss any seasons during their primes, or on the cusp of excellent seasons. Same with Munson, Campanella (at the end), Clemente.

Pre-farm system minors is exactly the same condition as Negro League players. However, I'm not aware of any cases where a player that was kept in the minors "too long" made an immediate impact in the majors, ergo went straight to superstar status. If there are some, then I've missed them, and I'd like to know who they were. Lefty Grove went 10-12 with a 4.75 ERA his first year, and he was only 25 anyway. He shouldn't even be in this discussion. Cravath didn't make a significant impact until his second full season. Extrapolating what he missed is negligible. Pud Galvin's first full season was at age 22. Why would anyone try to project before that early age? No need to.

Major League season shorter than 162 games...
If they played more than one-third of the season, you can't very well give them much partial credit for what was left. If they barely played any of the season at all, then it can be considered an empty season that can be extrapolated. One such example is May's '52 season, where he only played 34 games. Likewise, McGwire played only 27 games in 1993, so on the heels of a fine 1992 campaign, something of value can be extrapolated there.

Baker's 1915 season? It's bookended by a very good 1914 season, so any data set showing trends would indicate some value for his missing season at age 29, regardless of the reason. George Davis' missing most of 1903 would barely make a blip on the radar even if it were assigned. We need better examples for many of these categories. I'm willing to bet that they're straw men, and aren't threats to a slippery slope.

Sisler missed 1923 at age 30, so that's the only season he would need to get credited for. Nothing to worry about. Conigliaro played two full seasons after his beaning, hitting 20 and 36 home runs, though he was an average outfielder. There’s no place to credit him anything extra.

Conservatively, in a typical season you want to assign a value of roughly half the player's value above the league in the surrounding season(s). That's what players average in three successive seasons during their prime. For example, using OPS for demonstration purposes, if a player had a 160 OPS one season and two years later had a 140 OPS, his average production for the year in between is roughly 125 OPS (which is the average of the other two -- 150, and halving its value above the norm). This pattern can be factored in the extrapolation, in conjunction with age patterns.

This list is a good start, but it doesn't have enough examples of what it's trying to define. Many of these areas aren't even issues, because there are hardly any cases that apply to them to any significant degree, which then makes people want to shy away from their use when in fact there are a handful of legitimate cases to where some of them might apply.

We have the template of Ted Williams missing essentially five seasons around his prime, for a very legitimate reason. And Bob Feller and Joe DiMaggio missed significant time in their primes too, all for military leave. I don't know of any other players who missed more than a year in their prime on the wings of a standout season due to health reasons -- with the exception of Sandy Koufax. So the whole health discussion is a red herring. It's OK to allow for seasons missed that are health-related because it's only going to make any substantial difference in just a handful of players' career values. There is no slippery slope there.

Addie Joss' final shortened season was barely above average. Ross Youngs' last two seasons were substandard and average. Again, nothing to measure here. We need better examples to call this a troublesome category, if there are better examples.

For shorter seasons, if you analyze them, you'll see that guys like Bagwell and Thomas in '94 benefitted from the shorter time spans. If you don't acknowledge this, then you're conceding that the two best seasons of the '90s were by those two players that year, which would be a rather amazing coincidence. Gwynn and Brett's highest BAvg coming in severely shortened seasons would be quite a coincidence also.

I think this is the wrong question. The correct question is "what kind of player was he?" How does the event that caused the missed games give you information about the player's character when on the field? (when I say "character" I mean the kinds of things that help win ballgames)

I think going to war says very mildly good things about the player's character -- more so Vietnam, Korean War and WWI than WWII because it was expected that every American man would contribute to the WWII war effort (which in no way minimizes the amazing contribution they made to the world by doing so).

I think being banned for life says very bad things about your character. If you're dishonest off the field, I don't trust you to do what's good for the team rather than yourself on it.

I think "normal" physical injuries are part of the player. You weigh that into their value. If McGwire misses that many games, that's part of who he is.

I think "freak" accident physical injuries or deaths are neutral. I'm talking car accidents, plane crashes, tubercular meningitis. The kinds of things that could happen to anyone. You don't account for them at all -- you just use the data you have.

This all helps you paint a mental picture of the man. Stats are by far the most valuable piece of evidence, but even they don't tell the whole story. I ask myself what kind of player the man was -- what kind of player he would have been in different eras.

I guess to some extent this means I give "credit" for everything. When someone missed a lot of games, I ask myself what kind of player he would have been if he'd gotten a full career and, more importantly, how likely it seems that he would have gotten a full career if things had gone differently.

But to some extent I don't give credit for anything. The stats are as they are. Evidence can't lie, it can only be misinterpreted. I don't give a pitcher extra wins or a hitter extra homers -- the did what they actually did.

Of course, this all boils down to the broader question, and one that doesn't have a clear answer: "Who belongs in the Hall of Merit?" Or, phrased differently, "what are we measuring here?"

I tend to use the following mental game to make my selections: Let's say you're the great GM in the sky, and you're presented with every ballplayer who ever lived. They're all rookies, and whoever you draft is going to spend his entire career with you. When that career is done, you'll have to replace him with the folks who are still available, but there are other GM's out there who have been chosing as well, so the pool will be greatly diminished. These other GM's will have teams that play every style of baseball ever played -- Keokuk in the 1870s, Shibe Park in 1913, Yankee Stadium in the 1930s, the Astrodome in the 1960s, Camden Yards today, etc. -- so your team had better be fairly flexible (or if you have specialists, they'd better be good enough to win you all of your games in that particular time and place).

The guys you get are just tendencies -- you don't get their actual production -- you get their potential production as it existed their rookie year. Their potential includes their mental and physical quirks, including tendency toward injury or to do or say stupid things. Albert Belle is still a jerk. John Rocker is still an idiot.

Given that scenario (and the relative scarcity of players of different types), where does any particular player rank?

Great post, Zap. I agree that the concept of "extra credit" is perhaps a distraction, certainly in the global sense you discussed. But also in a very specific sense.

That is that the subject of "extra credit" is clearly two different discussions. For some, it is a question of EXTRA credit, especially for those playes who missed games to serve their country in some war or other, or a player who was especially abused by the "system," whatever that might mean.

And there is the exceedingly rare case of EXTRA credit for players who were affected by racism--that would be Jackie Robinson, Roy Campanella, Monte Irvin and a few others. EXTRA in the sense that this category is players who have a standard-issue major league record by which they can be evaluated, but who also have additional accomplishments that might (or "should" IMO) be considered as well.

But for most black athletes--and similarly for pre-1871 (and to a lesser degree the NA, at least insofar as there are those like Bill James who don't recognize the NA, and to a lesser degree yet the UA, FL, etc.)--the problem is not whether and how much EXTRA credit to give.

It is simply whether and how much CREDIT to give in the first place. For many players there is relatively speaking almost no standard-issue conventional everyday statistical record of credit to add extra credit to. Yes, we have some minimal data for the Negro Leagues and certainly a bunch for the NA. But what does it mean? How does it compare to the same data at a different time and place. We have a sense that we don't really know what a .300 BA in the NA or the Negro Leaguers really means, while we think we know what it means in other contexts.

So how much credit? That's where stuff like "discounts" come in along with other evaluation strategies. But those have little or nothing to do with EXTRA credit. It's not adding to an established record, it's a question of guesstimating what the basic record really is.

I DO give CREDIT, though not extra, to pre-'70, the NA, UA and FL heavily discounted, and Negro Leagues, regardless of whether there is a statistic base or not. To me, the issue of EXTRA credit is 5 percent, the issue of CREDIT for non-statistically documented achievements in 95 percent of what we should be talking about.

So having said all of that, the only players I give any EXTRA credit to are the guys who served in war--95% in WWII and a few in WWI and Korea (I'm not aware of anybody we will be talking about who served anywhere else)--and those few blacks who jumped from the Negro Leagues to the majors (I should have mentioned Sam Jethroe who will be something of an acid test).

I give no extra credit for the various mishaps represented by Charley Jones, Addie Joss, Frank Baker, Joe Jackson, Kirby Puckett...if so, why not Pete Reiser, Ken Hubbs, Lyman Bostock? Yes, I think that is a slippery slope.

PS. I haven't entirely figured out the following, so this is a question. Yes, I extrapolate short seasons pre-1904 (and 1918 and 1919) to 162 games, though when working with WARP I use their measly ^1/2 factor.

Not sure about strike years. It seems to me that 1884 and 1984 are quite a bit different scenarios???

Wow... a thread based on one of my posts!

Yeah, sorry if you think a couple of the categories are strawmen. I was just brainstorming trying to think of all the reasons players get extra credit. I understand no one has argued to give extra credit in some of those situations. If I knew that it was going to be posted at the top of a thread, I might have tried to word it more carefully.

That said, I have read that a couple of people are considering doing some extrapolation of the numbers for Joss's career. I wouldn't do that. I judge him only on his nine seasons.

Also, at the top of my head, I only listed the memorable examples. Memorable examples tend to be great players (Davis, Grove) who aren't going to need extra credit to get into the HOM anyways. Same goes for WWII credit to Williams & DiMaggio. I've heard some argue that Feller should not get war credit because the war probably saved wear and tear on his arm. Also, Dimaggio spent one extra year in SF as a stipulation of his acquisition by the Yankees. His rookie performance would certainly imply that he was MLB caliber in 1935 as well. Interesting debates... but certainly moot because these guys aren't going to need extra credit anyways.

Anyhow, my post was directly in response to giving extra credit for Gavvy Cravath. In my opinion, his resume needs the extra credit to move above the bar. WWII players like Phil Rizzuto are going to need the extra credit. Benny Kauff just got a vote... he's certainly going to need extra credit (and until a couple of days ago, I was unaware of the strange circumstances of his ban).

Cravath brings up another question: Who else is there who would need pre-farm-system minors credit? Earl Averill is twenty "years" away but also spent too much time in the PCL. Anyhow 1900-1930 was a time when the minor leagues thrived but weren't affiliated with the majors yet. Are there more Cravath-like cases out there?

- reasonable credit based on their minor league stats, if the player was obviously of Major League caliber

I'm not sure that this should apply to pre-farm system minors, but not to players who were kept in the minors when they could have been productive major leagurers. Being blocked by another player or being unrecognized by the orgagnization for what he could do are a couple reasons this happens. The first name in my mind is Edgar Martinez. So give this credit to players of any era, or don't give it at all.

What about the Japanese players?

I should add that I don't use "final" numbers for people. I don't care that Frank Chance had 237 career win shares -- I don't think it's possible to boil a player down to a single number like that. It's my preference instead to look at their year-by-year stats.

I'll pull up Frank Chance's "matrix" of annual stats and hold it up next to Jake Beckley. I don't think one is obviously better than the other -- Chance had a shorter and more typical career (ramping up from average as a rookie to great at his peak ages of 25-28, then a relatively quick decline to average by age 34 and out of baseball at 36. Then I'll look at Beckley -- outstanding as a rookie at 20, very good at 36, and maddeningly inconsistent in between.

This helps me paint a picture of what kind of player that person was, but it's very time consuming.

Anyway, the point of this posting (which seems to have gotten away from me) is that if you do use single "final" numbers for people, then it does make more sense to ask if you should credit them for time they did not play but might have if things had gone differently. I personaly try to eschew those "final" numbers, so it doesn't make much sense to me to do so.

Regarding Edgar Martinez, things were much different in the 1920's as opposed to the 1980's.

Lefty Grove was not allowed to leave, though major league teams wanted him. He was under contract to Jack Dunn and Orioles and they could have kept him forever if they wanted too. It had nothing to do with major league teams not thinking he was ready.

The fact that he 'struggled' (10-12, ERA+ 98) upon arriving in the majors doesn't mean he wasn't major league ready before then. Just look at how he pitched for Baltimore (I don't have the stats handy, but I have seen them before).

If he had been sold when he was 22 or 23, he would have adjusted that season instead of 1925.

Edgar was a late-bloomer.

Martinez hit .258 with 3 HR (357) at age 22 in AA. Age age 23 in 1986 he hit .264 with 6 HR in AA repeating the league.

At age 24 in 1987 he broke out at AAA, .329 with 10 HR, but it was the PCL.

In 1988 he tore up the PCL, .363 in 95 games. He only played 14 big-leage games - did he get hurt?

At this point - 1988 - he was obviously a major league player.

But in 1989 (age 26) he slugged .304 with a .314 OBP, in a little less than 200 PA for Seattle, it's easy to see why they weren't just handing him a job.

I don't see anything in his record that warrants extra credit. I assume he got hurt during his first trial (1988) and failed in his second (1989). He was simply a late bloomer.

In my mind there is a very clear distinction in applying Credit to players who PLAYED and players who DIDN'T PLAY.

On the one hand, there are players who PLAYED. They include the Negro Leaguers and any Caucasians who, for whatever reason, could not get a job on a AL or NL team, so played in another league. I believe that it is necessary to look at that play (whether we have stats or not), and consider what level he would be playing at if he had been allowed to play for the Detroit Tigers or St. Louis Browns instead of Minneapolis Miners or Kansas City Monarchs. I don't minimize these numbers at all for uncertainty, since the uncertainty has nothing to do with the player's ability.

On the other hand are players who DIDN'T PLAY. I am willing to give some consideration -- especially for military service -- but here the uncertainty extends beyond knowing how to adjust the numbers to major league levels, but whether they would have had an "off year" or not had they actually played.

Pat Burrell had a 149 OPS+ (920 OPS) in 2002, and will be comparable if not better this year (975 OPS so far), but had a pathetic 89 OPS+ in 2003 (713 OPS).

It would be reasonable to interpolate numbers if Burrell had actually missed 2003 due to war, holdout, strike, lockout, gender reassignment surgery, etc., but is it necessary or useful to assign the hypothical Patricia Burrell at 140 OPS+ for that missing season based upon his/ her performance on the two years on either side? Burrell's 2003 is not the norm, but it certainly is not so rare that an All-Star calibre player turn in a 9 Win Share season.

Now, it certainly would have been bizarre for Ted Williams to turn in 3 9-Win Share seasons in 1943-45, but there is also no reason to assume that his OPS+ would have been over 200 for the 3 year period, so the more information there is, the more benefit of the doubt I'll give. When a player's career is cut short (like Kauff) or started late (like McGinnity), and there was no major OR minor league seasons played in the gap, I simply could not justify giving any more than token credit, and I do not give any credit at all.

For a single missed mid-career season I give minimal credit (treating it no better than an off year, say 9 Win Shares), two missed season -- like Charlie Jones -- I give a little more credit (say 12-15 win shares), and in the extreme case of Ted Williams (3 missed years), I add about 18 WS per season (as if he needed it).

Kauff gets nothing, as he didn't play again at all.

Cravath (and Grant and Foster) gets a lot, because for every year a major league team didn't give him a shot, there was a high minor league team who rode him into contention, and often to a pennant.

I agree with your analysis completely PhillyBooster...well, almost completely.

I don't think being a star in a high minor league gets a player into the HoM. It may get him some credit, but it certainly is watered-down credit.

Could someone post a list of players who should get WWI credit? I know about Jackson and Kauff, but who else is there?

Rick:

From my #102 on the '26 Discussion Thread:

Well, I have a list of military vets, from a military database at the SABR web site. Here is who served among new (1926) eligibles, for anyone who might want to give additional credit.

1. Benny Kauff - Only played 67 games in 1918, doesn't say what branch.

2. Ray Chapman - Played in all of Cleveland's games in 1917 and 1918, probably served part time at home.

3. Buck Herzog - Says he served in WWII, but not WWI.

4. Lefty Williams - Only pitched in 15 games in 1918.

5. Happy Felsch - Only played in 53 games in 1918.

6. Joe Jackson - Only played in 17 games in 1918. Claims he served in the Army.

7. Swede Risberg - Only played in 82 games in 1918.


I'm trying to keep track on a year-by-year basis. The military database can be found at the SABR web site.

Ron,

Thanks for reposting the list. I must have missed it the first time.

The Neft+Cohen Baseball Encylopedia (updated annually?) notates MS (Military Service) and WW (War Work) next to a number of players in 1918 and 1919. I don't know how complete their annotations are (particularly for "War Work" which seems to cover only the 1917/19 World Series participants; maybe their source only covered those teams). Also, some players (eg Eddie Collins) apparently missed only a month, while others (eg Eppa Rixey) missed all of 1918 and some of 1919. (Italicized are players with at least 9 year eligibility; some will merit our attention.)

1918
AL - MS
Bos - Stuffy McInnis, Fred Thomas, Wally Mayer, Dick Hoblitzell, Jack Barry, Jimmy Cooney, Del Gainer, Hal Janvrin, Duffy Lewis, Mike McNally, Chick Shorten, Jimmy Walsh, Dutch Leonard, Herb Pennock, Paul Musser
Cle - Joe Evans, Josh Billings, Hank DeBerry, Lou Guisto, Joe Harris, Elmer Smith, Guy Morton, Otis Lambeth, Ed Klepfer, Red Torkelson
Was - Eddie Ainsmith, Val Picinich, Sam Rice, Patsy Gharrity, Joe Leonard, Mike Menosky, Horace Milan, Earl Yingling, Molly Craft
NYA - Wally Pipp, Bill Lamar, Aaron Ward, Muddy Ruel, Sammy Vick, Howie Camp, Bob Shawkey, Bob McGraw, Alex Ferguson, Ed Monroe, Neal Brady, Ray Fisher, Ernie Shore, Walt Smallwood
StL - Wally Gerber, Hank Severeid, Ken Williams, BabyDoll Jacobson, Bill Rumler, Ted Sloan, Urban Shocker, Ernie Koob
Chi - Eddie Collins, Swede Risberg, Ted Jourdan, Pat Hardgrove, Joe Jenkins, Jim Scott, Tom McGuire
Det - Harry Heilmann, Ben Dyer, Bert Ellison, Joe Cobb, Ira Flagstead, Fred Nicholson, Bill James #1, Eric Erickson, Johnny Couch, Howard Ehmke
Phi - Jimmy Dykes, Manny Kopp, Gene Bailey, Ray Bates, Pat Haley, Otis Lawry, Whitey Witt, Elmer Myers, Bob Geary, Tom Zachary, Walter Anderson, Jing Johnson, Dave Keefe, Rollie Naylor, Win Noyes, Socks Seibold

NL - MS
Chi - Pete Kilduff, Rowdy Elliott, Paddy Driscoll, Bill Marriott, Morrie Schick, Pete Alexander, Harry Weaver, Vic Aldridge
NYN - Benny Kauff, Eddie Sicking, Al Baird, George Kelly, Jesse Barnes, Fred Anderson, Rube Benton
Cin - Pat Duncan, Morrie Rath, Rube Bressler, Dutch Ruether
Pit - Buster Caton, Casey Stengel, Lee King, Ben Shaw, Fred Blackwell, Tony Boeckel, Adam Debus, Charlie Jackson, Ray Miller, Hooks Warner, Billy Webb, Earl Hamilton, Hal Carlson, Bill Evans, Elmer Ponder
Bro - Jim Hickman, Ray Schmandt, Ernie Krueger, Red Sheridan, Chuck Ward, Lew Malone, Leon Cadore, Jeff Pfeffer, Al Mamaux, Harry Heitmann, Clarence Mitchell, Jack Russell, Johnny Miljus, Sherry Smith, Duster Mails
Phi - Patsy McGaffigan, Pickles Dilhoefer, Cladue Cooper, Ben Tincup, Dixie Davis, Frank Woodward, Eppa Rixey
Bos - Ray Powell, Joe Kelly, Wally Rehg, Rip Conway, Zeb Terry, Rabbit Maranville, Walt Tragresser, Fred Bailey, Stan Covington, Hank Gowdy, Art Rico, Hank Schreiber, Ed Fitzpatrick, Dana Fillingim, Hugh Canavan, Bill James #2
StL - Doug Baird, Walton Cruise, Jack Smith, Frank Snyder, Tony Brottem, Dots Miller, Jakie May, Oscar Horstmann, Lou North, Bruce Hilt

AL - WW
Chi - Happy Felsch, Joe Jackson, Lefty Williams
NL - WW
NYN - Walter Holke, Dave Robertson, Jeff Tesreau
Cin - Larry Kopf

Jackson also was noted as a salary hold-out at the beginning of the season. Anyone know how much time was actually lost to War-Work as opposed to the Hold-Out? Usually, no credit is given for the latter (but everybody is different about those things).

1919
AL - MS
NYA - Bob McGraw, Walt Smallwood
Bos - Dick Hoblitzell, Paul Musser
Was - Tom Zachary
Phi - Ray Haley, Otis Lawrey

NL - MS
Cin - Pat Duncan
Pit - Adam Debus, Elmer Ponder
Bos - Hank Gowdy
Phi - Eppa Rixey

Any typos or misreadings of the small print in these books are entirely my fault, natch.

Philly Booster - why assume the worst?

Williams could have had a career year in 1943, 44 or 45. Burrell could have just as easily had a 180 OPS+ last year. You have to project based on the best available evidence.

IMO it's a much bigger error - a monumental one - to give Williams 9 WS for each of 1943-45 than it is to give him 50. I'd just give him a simple average of the 2 years before and after, if I were being conservative, since considering his age, it's likely those would have been years where he was improving not declining.

Obviously for Williams it doesn't matter, but for someone like Rizzuto it makes all the difference in the world.

The Burrell example is an extremely loaded one. You could just as easily have used Cecil Fielder 1990 to prove the opposite point. The best you can do is to make an educated guess. But to assume the absolute worst doesn't make any sense at all to me.

"there is also no reason to assume that his OPS+ would have been over 200 for the 3 year period,"

Sure there is reason. There's plenty of reason to assume it, as his OPS+ were 235, 217; 215, 20 in the surrounding seasons, and he was 24-26 in the missing seasons. I think there's more than enough reason to assume his OPS+ would have been over 200 those years. There's infinitely more reason to assume it would have been 240 than there is to assume it would have been 89 - considering his career low was 113 when he was 40 and his second lowest was 160 when he was 20.

should be, "as his OPS+ were 235, 217; 215, 205"

Stupid keyboard - listen to me damnit!

"IMO it's a much bigger error - a monumental one - to give Williams 9 WS for each of 1943-45 than it is to give him 50. I'd just give him a simple average of the 2 years before and after, if I were being conservative, since considering his age, it's likely those would have been years where he was improving not declining."

I reread your comment - I'll change mine to, "IMO it's a much bigger error - a monumental one - to give Williams 18 WS for each of 1943-45 than it is to give him 50."

Doesn't change my point at all - but I wanted to fairly characterize your point in my rebuttal . . .

But to assume the absolute worst doesn't make any sense at all to me.

I agree with this. There should be a discount factor that takes into account the risk that a season is not as good as the surrounding two, but only if there's a significance to that risk. Here's what I mean:

Seems like we would be concerned only with significantly bad seasons sandwiched between pretty good seasons (or possibly the reverse). We aren't concerned about estimating that someone had a 29 WS season squeezed between a 26 and a 32 when the number really should have been 22. Who's gonna quibble over 7 WS? Instead, we would only be concerned about the random occurrences like Burrell, where there is a significant drop and then a return to performance equal to the first "bookend" performance.

It seems like someone could see how many players there have been with 20+ WS on either side of a season where there was a dropoff of at least 12 WS (or the reverse, where someone had fewer than 15 WS on either side a season where the WS increased at least 12). I vaguely remember seeing a Bill James article on this.

Anyway, if there are a statistically significant number of seasons like that in major league history, then we could consider factoring in some sort of random chance discount that took this risk into account. I suspect, however, that such seasons are very rare.

If that's true, then Joe is certainly right that there's no point in assuming the worst...and for simplicity's sake, averaging the seasons probably is about the best way to go about it.

Here's a technical point though: Assuming Williams played during the war years, in theory his team would win more games. But how many more? If the Sox would have won 6 more games with Williams in the lineup during a particular season, but he gets 45 WS, that means that some WS will have to be subtracted from his teammates, because the team WS can never exceed three times team wins. In my example, Williams would clearly get 18 WS for the 6 wins, but if he got 27 more than that, some other players would lose WS. You could start by subtracting them from whichever player played in Williams' spot. If there are any left over (and there certainly should be), you could subtract them pro rata from the teammates. I dunno, probably not worth the trouble, but fun to think about.

I agree - David - players that did play during the war should be discounted - guys like Stan Musial and Marty Marion, for example.

Off topic sort of, but I always thought it would be cool to make Diamond-Mind projection disks for 1943-45 with all of the players in military service added with reasonable projections (as well as discounting the players that played somewhat). I think it'd be cool to see which teams (Boston) likely got 'screwed' the worst . . .