He'll live in baseball's Hall of Merit
He got there blow by blow
Our kids will tell their kids his name
Joltin' Joe DiMaggio

I'm thinking of maybe starting him off on my prelim at #1 and then giving it some more thought. What do you guys think?

Vince was sooo much better than Joe.

; )

Bref: DiMaggio's most similar player is Larry Walker.

I'll never speak of unadjusted similarity scores again.

155 OPS+ in a ballpark that wasn't suited to his handedness is pretty darn good. No way he's not #1 on my ballot.

Bref: DiMaggio's most similar player is Larry Walker.

Just imagine a CF putting up Walker's Colorado stats in Dodger Stadium. That's Joe DiMaggio.

How highly did the sportswriters of the day think of Joe DiMaggio?

He retired in 1951, and was elected in 1955, just four years after retiring.

This tidbit comes from one of the great moments on the late Neyer Board -- apparently, prior to the late '50s writers were told not to vote for a player until after he'd been retired for 5 years, but if they did, the votes were counted anyway. The result: the people that retired around then - Ted Lyons, or Red Ruffing or Al Simmons - would get 1 vote one year, 3 votes another year, and then have a deluge of votes another year. This is the real reason why there were no "first ballot" HoFers from 1936-62. I assume his election was what caused them to tighten up on their rule enforcement.

But with DiMaggio it wasn't 1 or 2 votes. He got 117 in '53 (good for 8th place on the ballot), was on 69.4% of the ballots the next year, and then got elected in '55.

He is, as far as I know, the only player elected to the HoF before he was eligible for induction.

I believe the last player to get an illegal vote was, of all people, Phil Rizzuto in 1956. He hadn't even retired yet.

Depends what you mean by 'eligible,' but Roberto Clemente comes to mind.

They waived the rule for Clemente, if I remember (I waas following the game by then) as they had for Gehrig.

JoeD's OPS+ was so high because his career was so short, and he wasn't hitting against the Yankee pitching. Somewhat overrated, though he's still #1, just over Beckley. But where's the Simon and Garfunkel song about Beckley, eh?

Also: "Almost as good as his brother Joe,
Dominic DiMaggio!"

karl,

>JoeD's OPS+ was so high because his career was so short,

Of course you're right, a normal decline would bring it down. It was 117 in his final year after four years of 154-64-78-52 and he was reputadly playing hurt (the 117). So give him three more years at 117 and his career OPS+ probably comes down, what?, about 9 points?

But what if he plays 3 years in '43-'44-'45 in between a string of OPS+ years of 176-86-48 and 142-54-64, and if he hit the same OPS+ (on average), or about 161.5. Of course now you're only adding about a point.

So I figure if you adjust on both ends you come up with 148ish.That drops him down from the company (at 156) of Aaron, Speaker, Greenberg, Mays, Mize, Musial and Ott (from 155-157) to Heilmann, Kiner, McCovey, Schmidt and Stargell (147-48).

Yes, it's a demotion. But add all 6 hypothetical years and he's over 10,000 PAs to Heilmann's 8600, forget Kiner, McCovey's 9600, Stargell's 9000 and Schmidt at about 9950 despite longer seasons. And other than Schmidt, none of them fielded much.

Still as you think about CFers, DiMaggio occupies a unique spot in the middle of the HoM/HoF type players. A fer piece below Mays, Cobb, Speaker and Mantle, and a fer piece ahead of everybody else. In order of James' ratings:

Mays OPS+ 157 in 12,400 PAs (could get a boost for 1952-53)
Cobb 167 in 12,800
Mantle 173 in 9900
Speaker 156 in 11,700

DiMaggio 156 in 7650
DiMaggio + WWII 157 in 9400
DiMaggio + WWII + decline 148 in 10,000++

Snider 138 in 8150
Griffey (through 2003) 145 in 8100
Puckett 123 in 7750
Hamilton 139 in 7550 (but season-adjust to 8500 or so)
Wynn 129 in 7900
Doby 137 in 6250 (adjust both upward but how much?)

I don't think you can add in a decline for DiMaggio and not consider that 3 gaping year hole in the middle. But with or without any of those adjustments, he's clearly #5, not #4 and not #6.

and he wasn't hitting against the Yankee pitching

This argument comes up periodically. I thought that OPS+ used the BPF which factored out the effects of not batting against your own pitchers.

The guy was a right-handed power hitter in Yankee Stadium long before the renovation. Anyone have home-road splits on Joe D?

IIRC, he out-hit Ted Williams on the road (using batting average here) by a few points, but I don't have any numbers with me to back it up.

JoeD's OPS+ was so high because his career was so short, and he wasn't hitting against the Yankee pitching.

The park factors at bbref and in Total Baseball do adjust for not playing against your own team. OPS+ is adjusted that way.

Note that Win Shares does NOT make that adjustment, so Win Shares will tend to overrate players for dynastic winning teams and underrate those for long term losers.

****

IIRC, he out-hit Ted Williams on the road (using batting average here) by a few points, but I don't have any numbers with me to back it up.

Also note that road-only stats will underrate players who are favored by their home park, and overrate those adversely impacted by their home park. It overcorrects for park.

On Williams vs DiMaggio, say, this means that 1/7th of their home stats should be added to their away stats. (DiMaggio played 1/7th of his road games at Fenway, Williams should get the same number of Fenway games, too, just as Dimaggio should play some games in Yankee Stadium when doing the heads-up comparison.) To finish the comparison, you then have to adjust hitters for any difference in their team's average pitching staff quality; pitchers for not having to face their own offenses.

We can discuss park factors and home and road stats, fine-tuning what we know, but there are two things lurking behind all that:

1. It's very hard to imagine not having DiMaggio #1 on this ballot.
2. In #10, sunnyday2 is basically right and DiMaggio's place on all-time CF lists isn't particularly sensitive to the fine-tuning.

"Note that Win Shares does NOT make that adjustment, so Win Shares will tend to overrate players for dynastic winning teams and underrate those for long term losers."

Not really. I don't think. Because in WS you are really only compared to the other hitters on your team (who also didn't face those pitchers).

Since WS splits the credit for offense and defense up by team, I don't think there's a problem there. I could be wrong here, but I think this is indirectly taken care of by the system, right Jim? Am I missing something here?

What I'm trying to say is that if a team has really good pitchers, the offense will have fewer WS to split up, right?

All right, if people want to get into a "CF Rankings" debate, how would you compare DiMaggio to the Negro League centerfielders? I think it's pretty uncontroversial to say Charleston would be in the tier above, and Hill in the tier below. How about Stearnes and Torriente?

Not really. I don't think. Because in WS you are really only compared to the other hitters on your team (who also didn't face those pitchers).


When I first saw this I thought Joe's statement must be true, because it is certainly true that Win Shares does not favor players on good teams over players on bad teams, at least not generally. Now I'm not sure that the statement is true.

It's late, and I may not be thinking clearly, but if Joe's statement is true, then why have park factors at all? If you are only comparing to other hitters on your team, and they all played in the same park, why adjust for park at all?

But the park factors do matter, so if the park factors are off by a little bit, so will the Win Shares be.

I think it is because the Win Shares system does not apply the park factors to the individual players' runs created, which would apply the adjustment equally to the players and keep everything proportional.

Instead, I think the park factors apply when determining by how much a player's runs created exceed those of an average player...the so-called background runs, from which the offensive claim points (really marginal runs) are calculated. In other words, the park factors are applied to the average player, not the player for whom we are doing the calculating. Then you subtract the average player "background" runs from the player's runs created.

Similarly, the park factors are in play in determining how many offensive win shares a team will get as a whole, because the offensive win shares are derived from marginal runs, which involve subtracting estimated league average runs from the team runs scored.

The subtraction is a key concept here, because that's how the team offensive win shares change (and the weights among players change). Subtraction is not proportional like multiplication would be.

Here are some players from a very very good team, which had a park factor of 104. The columns show how many WS each player had at a park factor of 104 and 108.

Player   @104   @108
#1       17.2   16.7 
#2        3.4    3.2
#3        1.3    1.2
#4        1.3    1.3
#5       12.6   12.1

The team (Boston 1875) won 71 games and lost only 8, and thus had 213 win shares to go around. At a park factor of 104, they had 119 offensive win shares and at 108 they had 115 offensive win shares. So the team got 4 more offensive Win Shares when the park factor went up.

If I understand Jim's point, a good team not having to face its own pitchers (assuming the pitchers were good) ought to have a higher park factor when judging that good team. The park appears to be 4% easier to hit in than average, but that's only for visitors who had to face the excellent home team's pitchers. If we are trying to measure what an average hitter would have done in the park if he was in the home team's shoes, then we need something higher than 104, because 104 assumes the average hitter faced the home team's pitching.

If that's the case, then it appears that players with a significant number of Win Shares get a boost from the lower park factor, because the average player is deemed to have scored fewer runs -- therefore less to subtract from the home team player's runs created (and from the team's runs scored in calculating overall offensive win shares).

If you increase the park factor to eliminate the home team's good pitching, the average player would have scored more runs, which means the home team player will have fewer marginal runs after the subtraction (and the home team will have fewer marginal runs and fewer offensive win shares).

The boost isn't too large, though. In the chart above, player #1 gets 1/2 a win share more when the park factor is 104 instead of 108. Going from 104 to 108 is a significant difference -- I don't have any data suggesting how much of a difference good pitching would make on the park factor, but I'd be surprised if it were as large 4 points. Over a 15 year career that would be 7.5 Win Shares. A superstar would gain more, of course.

Good teams would gain less if there was a smaller difference in park factor. At park factors of 105 or 106, there might be a 4 to 5 total Win Share boost to an average player on a good team over a 15 year career, and 8 to 10 Win Share boost for a superstar. That, of course, assumes the player plays with good teams with good pitching for his whole 15 year career -- and there's a finite number of those players. Presumably it works in reverse for bad teams, though I did not test it.

It sounds complicated to do park factors for every team, and subtract the effect of their own pitching. It would be useful for accuracy's sake, but I wonder if the cost of doing the extra work outweighs the benefit of a few Win Shares here and there for players who constantly played on either a good team or a bad team. I suspect 99% of the players in MLB would be essentially unaffected, and the other 1% would gain or lose maybe 10 Win Shares over a 15 or 20 year career.

Many many people (maybe even Joe Morgan or Tim McCarver) know more about Win Shares than me, so I don't intend this to be gospel. It was just turning around in my head, so I had to get it out.

1. It's very hard to imagine not having DiMaggio #1 on this ballot.

With WWII credit, I don't even think anyone is close. As far as I'm concerned, he should be unanimous.

I would rate the CFers:

1. Charleston though there is some uncertainty about that
2. Cobb
3. Mays
4. Speaker
5. Mantle
6. DiMaggio
7. Stearnes
8. Torriente
9. Snider
10. Hamilton

There isn't quite the gap after DiMaggio that you get just among MLers. Stearnes was probably pretty comparable to DiMaggio, and maybe Torriente to Snider (if that's not pushing it with the analogy).

Jeff did a nice job explaining the effects of the "park" factors on the direct WS calculations. (Thanks, Jeff.)

There is also a second effect which comes into play here. Good teams get more Win Shares to start with, because they play an easier schedule.

Take the 1932 Yankees (107-47), 1932 Red Sox (43-111), and a hypothetical .500 team (77-77). Let's complete the rest of the balanced schedule by assuming that each team also played against itself, going 11-11. Yanks (118-58), Red Sox (54-122), .500 (88-88). Now, rescale those records back to 154 games: Yanks (103-51), Red Sox (47-107), .500 (77-77).

As you can see, the Yankees effectively won 4 extra games because they didn't have to play against a team as good as themselves. The Red Sox lost 4 extra games because they didn't get the benefit of playing a team as bad as they were. 4 extra wins is 12 extra Win Shares, or 1 for each starting player on the team. It's within James' stated range of acceptable error for the seasonal shares, so it's no big deal for any individual season.

But it can add up over a career if the player played mostly for good teams, or mostly for bad teams. A player who compiles a .600 team winning percentage (such as Charlie Keller .618) over his career gets about .5 extra win shares each 154 game season. A player who compiles a .400 team winning percentage (such as Indian Bob Johnson .414) over his career loses about that amount.

This is another effect that can systematically add or subtract up to about 10 Win Shares over a career.

Here's a quote from the man who asked, "where have you gone?"

"In the fifties and sixties, it was fashionable to refer to baseball as a metaphor for America, and DiMaggio represented the values of that America: excellence and fulfillment of duty (he often played in pain), combined with a grace that implied a purity of spirit, an off-the-field dignity, and a jealously guarded private life. It was said that he still grieved for his former wife, Marilyn Monroe, and sent fresh flowers to her grave every week. Yet as a man who married one of America's most famous and famously neurotic women, he never spoke of her in public or in print. He understood the power of silence." -- Paul Simon, New York Times, March 9, 1999.

Well said, jimd.

I like Win Shares fine as an extra tool, but sometimes I see a guy who seems to rank significantly higher in our voting just due to WS. That's a little dangerous, I think, if that's what's happening.
WS can be a good reality check against OPS+ as well as 'more raw' stats. But the WS alone shouldn't get it done.

Very few players nowadays are on consistent winners, between them changing teams and more volatile standings finishes of franchises (much as Mr. Selig likes to pretend otherwise).
But in voting for players from the 1920s thru the 1950s, for instance, it seems as if we need to keep some of these other things in mind.

From baseballlibrary.com:

December 19, 1934: The Yankees send five players to San Francisco as part of the payment for Joe DiMaggio. He will play another season in the Pacific Coast League and will report at the end of 1935.

I understand this is fairly common knowledge around here and Joe is such a slam dunk candidate that any extra credit is unnecessary and moot -- but this is the ultimate case of a player who deserves PCL credit.

I've decided to calculate a major league equivalent record for Joe DiMaggio's PCL seasons. This one is done just for fun.

First, here are his actual PCL batting statistics with the San Francisco Seals:

Year Age   G  AB   R   H 2B 3B HR RBI SB  AVG  SLG SH
1933  18 187 762 129 259 45 13 28 169 10 .340 .543  5
1934  19 101 375  58 128 18  6 12  69  8 .341 .517  5
1935  20 172 679 173 270 48 18 34 154 24 .398 .672  7

Joe signed with the Seals toward the end of the 1932 season (referred by brother Vince, who was already playing for them) and played 3 games at shortstop, batting .222. (I don’t have Joe’s complete statistical line for 1932 – if someone has it, could you please post it, so we can show his complete PCL record.)

In 1933, of course, he achieved nation-wide attention by hitting in 61 consecutive games. He also led the league in RBIs and placed 8th in the league in batting average (minimum of 100 games). He also led the league in outfield assists.

In 1934 he injured his knee while getting out of a taxi. I wonder if this early injury may have contributed to later injuries that ultimately shortened his career. When he was able to play, however, he played well—ranking 7th in the league in average. As we will see, the slight decline in his slugging percentage is deceptive, since offense in the league as a whole, and especially in Seals games, was down sharply from the prior year.

At the end of 1934 Joe was traded to the Yankees for $25,000 and 5 players. One of the conditions of the contract was that Joe would first spend another full season with San Francisco. It’s interesting to read about this transaction on various Web sites; most authors, reflecting modern-day biases, look at the transaction solely from the point of view of the Yankees, saying that they wanted him to get another year of seasoning in the minors. I think this is misleading—I believe that the Seals insisted on the provision to keep Joe for one more season because they wanted to make a run at the PCL championship. (Their last championship had been in 1931.) And they would be successful, as Joe finally led the team to a pennant in 1935.

1935 was a great season – Joe was named PCL MVP and led the league in runs scored, RBIs, triples, and outfield assists. His .398 average was barely behind the .399 average of league leader Ox Eckhardt. Joe also hit 34 home runs (2nd to Gene Lillard’s 56) and stole 24 bases. His MLEs will indicate that his season was comparable to those of the top 10 players in the majors.

To derive MLEs, I had to estimate the run environment (using methods that I’ve described on the Buzz Arlett thread). Since Joe would go on to play in the American League, I decided to convert his record to an AL-environment. As mentioned, I found that relative to the AL, the PCL was somewhat batter-friendly in 1933 and 1935, but pitcher friendly in 1934. The dip seems to have been even more pronounced in San Francisco. My estimates of San Francisco’s run environment are 5.77 runs/game in 1933, 4.12 in 1934, and 5.33 in 1935. These environments are 15.4 percent above the AL environment in 1933, 19.7 percent below in 1934, and 4.7 percent above in 1935.

The other uncertainty is walks, which were not recorded in the PCL statistics. Adding to the uncertainty, I note that Joe drew only 24 walks during his first season with the Yankees, compared to 64 his second and 59 his third season. Was his 1936 walk rate an aberration as he adjusted to a new league? Or did he actually learn to take walks in 1937? For his PCL MLEs I’ve decided to split the difference, showing walk rates higher than his rate for ’36, but lower than his rates for ’37 and ’38.

Here are Joe DiMaggio’s PCL MLEs:

Year    G   PA   AB   R   H 2B 3B HR RBI BB  AVG  OBA  SLG OPS+ 
1933  154  652  618  79 185 33  7 17 104 34 .299 .336 .458  108 
1934   83  342  322  51 109 16  4 11  61 20 .339 .377 .516  128 
1935  151  636  598 126 218 39 11 25 112 38 .365 .403 .592  154 
Total 388 1630 1538 256 512 88 22 53 277 92 .333 .371 .522  130 

A minor note on the OPS+ calculation--I didn’t have access to David Foss's pitcher-removed contexts for the 1930s AL, so I used an earlier approximation suggested by Chris Cobb. I looked for teams on bbref with BPFs near 100 and used the lgOPS and lgSLG for a player from that team. It should be pretty close.

Here are estimates of runs created and of batting, fielding, and total win shares:

Year   RC  Bws  Fws WS
1933   97 16.2  5.5 22
1934   62 11.6  3.0 15
1935  135 27.8  5.4 33
Total 294 55.6 14.0 70

Thanks Brent!

If he were anywhere close to the in/out line, I would guess that Dimaggio certainly shoudl get credit for 1935.

When did his injury happen in 1934? Was it at the beginning or end of the season? the offseason? I only ask because it might change whether or not he should get 1934 credit as well.

I don't have a date for DiMaggio's 1934 knee injury, but I'm pretty sure it occurred in the first half of the season.

I don't own any of the biographies of DiMaggio, so I haven't checked them, but I've looked at various Web sites. They each seem to have the details a little different. Some say he was getting into a taxi, others say he was getting out; most of them place the injury at his sister's house; at least one says it was on a late Sunday afternoon after a doubleheader. A few sites say he hit a home run the next day but had to walk around the bases -- now that version sounds aprocryphal to me! But they all tend to agree that the injury involved a taxi, occurred during the season, and accounted for his missed playing time during the season.

Several Web sites suggest that the Yankees were taking a big risk on his knee when they purchased him - I guess that would depend on whether his knee had healed and he was back to playing before the end of the season.

For the sake of good accounting and making the correct precedents I agree we should give full credit for 1935, because he'd already been bought by the Yankees, but not 33 (when he was a marginal ML player anyway) or 34.

Adding 35 and some kind of war credit separates him from the herd considerably -- I was surprised Ott was so close, but a group of Ott/Foxx/DiMaggio is pretty exalted company, just not AS exalted as every 75 year old New Yorker is convinced he should be in. He wasn't Ted!

Joltin' Joe, in my WARP, with 1943-5 filled in using my war credit equations:



3-year peak: 24.9
7-year prime: 50.9
Career: 94.2
Salary: $287,917,820--24th among post-1893 MLB position players, below Foxx and Vaughan, above Mize, Matthews, and Bench.

DiMaggio definitely deserves minor league credit for his 1935 PCL season. I'm a little dicey on using Brent's MLE line posted year, since he didn't have info on the run environment. Do any of the group's contributors--Gary A? KJOK? Dr. Chaleeko? Chris Cobb? Brent himself now?--have the information to do a proper MLE for that year? If it was really anything close to a 154 OPS+ season, as Brent has claimed, that's probably something like an extra 7 WARP and $22.5M.

I'm not so sure he does deserve MLE credit for 1935. After all, 15 ML clubs weren't willing to take a chance on him due to his knee injury.

I may have been unclear about what we do and don't know about the PCL run environment. In 1935 the PCL scored 7,182 runs in 1,395 games or 5.15 R/G. (The AL scored 5.09 R/G, so about the same run environment.) The PCL's batting average (including pitchers) was .294 and its slugging average was .403. What I didn't know was the park factor for San Francisco. Judging by the team's actual runs scored and my estimate of runs allowed, the ballpark appears to have been roughly neutral during the early to mid-1930s, swinging from above league average some years to below league average others.

Let me try a simplified method for calculating DiMaggio's MLEs for 1935. DiMaggio's actual PCL BA and SLG were .398 and .672. If we adjust for quality of competition by multiplying his BA by .906 and his ISO by .82 (ratios from the Bill James 1985 study), his BA is reduced to .360 and his SLG to .585. If we then adjust to an AL context by multiplying his BA by (.280/.294) [the ratio of the AL BA (including pitchers) to the PCL BA] and multiplying his ISO by (.122/.109) [the ratio of AL ISO to PCL ISO], the resulting "MLEs" are .343 and .594. Note that the BA is lower than the one I calculated three years ago, but the SLG is about the same. (The methods I used then were more complicated, but followed the Bill James methodology more closely.)

The last step is we need to come up an estimate of his walk rate. Unfortunately, the PCL didn't record batter walks, so we literally have to make up an estimate. To complicate matters, his major league walk rate in 1936 was much lower than his rate in subsequent seasons--his rates (AB/(AB+BB)) for 1936 through 1940 were .036, .093, .090, .101, and .107. For 1935, I decided to assign him 38 walks and 598 AB for a walk rate of .060, but you are welcome to plug in your own best guess. With this assumption for walks and the simplified MLEs for batting and slugging, his line will be .343/.382/.594 and his relative BA/OBP/SLG/OPS+ are 119/107/143/150. So it's a little lower than what I came up with 3 years ago. You can play with these assumptions and see what you come up with, but I suspect that you'll find a plausible range for his OPS+ running from about 145 to 155.

Cblau, it says earlier on this thread that the Yanks wanted him for '35, but the Seals refused to give him up.

Ah Brent, that *is* a lot more conclusive than I thought. Well in that case, I'd say '36 looks like his "adjustment" year in the majors, and then in '37 he was right where we'd expect a guy who had a 150 MLE OPS+ at age 20 to be. (Jesus, he was good). Now the question becomes, do we think he would have underperformed his MLE OPS+ due to the adjustment had he played in the majors in '35? And if so, would he then have done better than he actually did in '36? Thoughts?

OK, well, using Brent's MLE's (with a slightly lowered walk rate) which now seem reasonable to me, and extrapolating his fielding from later years, the '35 comes out to 6.0 WARP2 (4.1 BWAA, .2 BRWAA, .7 FWAA, -.9 Replc). That increases his career total to 104.2 WARP2 and his salary to $305,795,835, which moves him ahead of Vaughan and just nudges him past Foxx, comfortably within the inner circle.

The contemporary reports of the DiMaggio deal suggest that the Yankees wanted to make sure he was 100%, which is why he was left in SF for 1935. They also suggest that the deal allowed the Yankees to call for him during 1935 if they so chose.

DiMag's 1934 injury occurred on May 20. He was out for about a month.

Interestingly, DiMaggio held out prior to the 1935 season; he didn't report until mid-March.

-- MWE

Just for the historical record, DiMaggio's 1934 injury almost certainly occurred on June 10th. According to Dimaggio's autobiography it was after a doubleheader at Seal Park, which in those days were morning-afternoon affairs. The Seals had three of those doubleheaders in June that year, on the 3rd, 10th & 24th, but since there's a newspaper reference to the injury on the 18th, that would eliminate the 24th. And since it took place on the last day of a home stand, that would rule out the 3rd, which was in the middle of one. In Lucky To Be A Yankee, he gives the following description:

After a double header at Seal Park in June, I went over to the home of one of my married sisters for dinner. It had been a long day at the ball park and when dinner was over, I decided to take a jitney cab home. I was riding in close quarters, cramped, and my left foot must have fallen asleep from the awkward position in which I was sitting. Getting out of the cab, I put that leg on the pavement first, with all my weight on it. Down I went, as though I had been shot.

There was no twisting, just four sharp cracks at the knee, and I couldn't straighten out the leg. The pain was terrific, like a whole set of aching teeth in my knee, and I don't know yet why I didn't pass out.

There was a movie house nearby, the Milane, and a friend of mine, whom I knew only as Frank, was the manager. I hobbled into his office and he drove me to the Emergency Hospital. I can't say, even at this late date [1946], that the treatment was 100 per cent. The doctor was rather vague about the whole thing and told me I had sprained alot of tendons. My knowledge of medicine, or even first aid, was limited to the extent of taking aspirins for headaches, so I thought the doctor must know what he was talking about when he advised me to go home and apply hot towels and Epsom salt packs.

The next morning when I attempted to get out of bed, I hit the deck with a bang. My leg buckled under me. It happened that it was an open date and we were leaving for Los Angeles** that night. I called Caveney and told him I had injured my knee but that it wouldn't prevent me from making the trip.

Caveney kept me out of the first game and used me in the second as a pinch-hitter. I hit a home run but even walking around the bases was painful but I didn't let on to Ike exactly how painful. I explained I was saving my knee by walking instead of jogging around the bases.

Again the next day I was used as a pinch-hitter but was lucky enough not to get on base. In the third game, I again pinched-hit and this time doubled. Between hobbling and walking I managed to make second. Caveney realized at once that it was no ordinary game leg which was crippling me. Had I told Ike the truth, of cours, he wouldn't have allowed me to make the trip, let alone use me as a pinch-hitter, but I minimized my injury because I didn't want to lose my job as a regular.

Caveney shipped me back to San Francisco and a Dr. Bull, the club physician, packed my left leg, from ankle to thigh, in an aluminum splint. The splint and I kept seady company for three solid weeks. And for a guy wo had the reputation of being lazy, you never saw anybody as impatient as I was. I wanted to get back into action, to prove that I was as good as new.

It was while I was out of the lineup that I heard stories that the New York Yankees were after me. The had both their Pacific Coast scouts, Joe Devine and Bill Essick, looking me over before I threw my knee out of kilter. The Yankees were my ideal. That was the club I wanted to wind up with, but I thought my chances were gone now. No club was going to fork over big money for a rookie sensation who had broken down.

When the splint was removed I went back to the outfield again but was obviously favoring my trick knee. After a week or so, I slipped in the dugout and the knee popped out again. I was through for the season, playing only 101 out of a possible 188 games and hitting .341. The average was good enough, but the "wonder" tag was off me. I was labeled a "cripple" now.

** This would have been Sacramento, where the Seals played on June 12th. San Francisco only played in Los Angeles or Hollywood in May and mid/late July that year, in the traditional week-long PCL head-to-head series, which cut down excessive travel. The fact that the injury took place after a game against Los Angeles is likely what caused DiMaggio's memory lapse.

Reading the details of that injury, and knowing how much DiMaggio must have been favoring it and still seeing how many games he had been forced to miss, you can see why the Yankees had little or no competition for his services after the season.

DiMag's 1934 injury occurred on May 20. He was out for about a month.

Where did you read that, Mike? Not doubting you, since I was basing my June 10th date on his autobiography and the PCL schedule, and not on a specific contemporary newspaper citation. But since he wound up playing only 101 games, that does make that "month" you say that he missed a bit hard to explain. If he tried to tough it out through late June, and gave up at that point, it would make more sense that he missed 87 games than if it had happened in May.

OTOH the Seals did play a home doubleheader (against Hollywood) on May 20th, and then did play in Los Angeles on the 22nd, so it may be that he simply misremembered the month.

Where did you read that, Mike?

Fresno Bee, through Newspaper Archive.

-- MWE

DiMaggio was in and out of the lineup during the second half of the 1934 season, apparently.

-- MWE

Where did you read that, Mike?

Fresno Bee, through Newspaper Archive.

-- MWE

Thanks. I saw the June 18 Bee article cited by google for "Joe Dimaggio knee injury" but didn't want to pay the subscription fee for just one article, and no matter how I spelled "Dimaggio" or "De Maggio" that was all I could come up with---I got nothing from either the NY Times or TSN. What did the Fresno article actually say?

Appropriately for comment #39, DiMaggio struck out 39 times as a 21-year-old rookie in 1936 and never again exceeded that seasonal total. Such a feat would be inconceivable nowadays.