July 20, 2015

Rebuilding from the War – Part One

The inaugural article for both the “What If?” feature and the first ever “Hall of Fame Week” we’re going to look at the careers of six all time greats and see what their career numbers would have looked like if they hadn’t lost seasons to serving in WWII. This is a two part article, looking at three players in each half – look for part two later this week.

Ted Williams swearing into the Navy 1942

Baseball and the War

In 1942, with the world at war and the country beginning to ration resources, there was a serious chance that Major League baseball would be disrupted. Six weeks after the bombing of Pearl Harbor, MLB commissioner Kennesaw Mountain Landis petitioned the president himself to keep baseball running.

Roosevelt letter to Landis.jpg

FDR famously replied: “I honestly feel that it would be best for the country to keep baseball going. There will be fewer people unemployed and everybody will work longer hours and harder than ever before. And that means that they ought to have a chance for recreation and for taking their minds off their work even more than before.”

Even though Major League baseball did not stop operations, many of the game’s biggest stars served in the military during the war, often in physical education instructor positions or in special service positions in which they played baseball to boost the morale of their fellow soldiers. A few players, including two we’ll examine, saw combat theater missions during the war (and a third did during the Korean War), while several more requested combat posts but were kept from the front lines by military brass.

The players we’re examining aren’t the only players who lost seasons to the war, nor are they the only players who served – for example, both Yogi Berra and Jackie Robinson saw combat prior to their MLB careers (during his service, Robinson fought for desegregation of the Army, something that did not happen until 1948 – after the desegregation of baseball).

They are however six of the all time greats (or at least five all time greats and one player who would probably considered such if he’d had his wartime seasons back) and the biggest name players to lose prime seasons to war.

A note before we get into the numbers: several times in this article we’ll use phrases like “lost seasons” or “his service cost him X home runs” when discussing our projections. At no point do we want to diminish the significance of the player’s military service or the importance of military service in general. At the end of the day, baseball is just a game and its importance pales in comparison to even the smallest contributions made by the servicemen and servicewomen who have served their country in uniform.

Projecting the Lost Years

In order to “fill in the blanks” of these players’ lost seasons, we’re going to be using a modified version of Tom Tango’s Marcel projection system. Named after Ross’s pet monkey from the early seasons of Friends, the system is very simple (“So simple a monkey could use it”) but it’s predicative reliability is actually rather high (higher than several more complicated systems).

The way Marcel normally works is by taking a weighted measure of a player’s previous three seasons, giving more weight to the most recent seasons to determine the rate at which the player gets hits, walks, home runs, etc. per plate appearance. It then adjusts those rates based on age (placing the player’s peak at age 29, with stats climbing to that peak then declining afterwards).

Normally, Marcel then adjusts the prediction by “regressing to the mean” (an important term in statistics). The basic concept is that the more unusual something is (when compared to “average” or “mean” performance), the more likely it is that outcome is an outlier rather than the beginning of a new trend.

A great example is Brady Anderson’s 50 home run season in 1996. Anderson had never hit more than 21 homers in a season before (his three highest previous totals were 21, 16, and 13) and both the raw total and the gigantic leap in his total were far from what an average player would do (especially at age 32).

Regression to the Mean therefore suggests that his following season is far more likely to reflect his career prior to that point (and the league average) than it would his 1996 season. As it turns out, Anderson’s next seasons totaled 18, 18, 24, 19, 8, and 1 before he retired at age 38. Weren’t the late 90s and early 2000s great?

Because we’re not predicting an unknown future here but trying to fill in the gaps in a player’s performance, we’re far less concerned with regression to the mean. In a sense, we “know the future” and the player’s standard of production for the years following the wartime gap.

Because of that, rather than adjusting for regression to the mean, we’re going to extend Marcel’s equation for determining the player’s basic rate stats. Whereas Marcel normally looks at the three prior seasons, we’re going to look at the three seasons prior to the wartime gap and the two years after the gap (weighing them the same way). By doing this, we’ll try to follow the player’s existing “career arc.”

Hank Greenberg

The first American League player to enlist for the peacetime draft, Hank Greenberg was one of the few players to have served during 1941, prior to US entry into the war. Greenberg missed all but 19 games of the 1941 season, serving as an anti-tank gunner at Fort Custer in Battle Creek, MI. Shortly after his promotion to Sergeant he was discharged, two days before the bombing of Pearl Harbor.

Greenberg being sworn in

With the beginning of the war, Greenberg re-enlisted in February of 1942, volunteering for service in the Army Air Corps (the precursor to the Army Air Forces and the modern Air Force). He ended up serving 47 months, the longest service of any major leaguer, reaching the rank of captain while serving in the 58th bomber wing in the China-Burma-India Theater of Operations. In June of 1945, he returned to baseball, the first major leaguer to do so.

Because Greenberg missed the most time of any MLB player, and because he played partial seasons in 1941 and 1945, we calculated his stats a bit differently than the others. His 1941 season was only 19 games, far too small a sample for useful analysis, so we treated 1938-1940 as the the three prior seasons for purposes of the calculation.

He played the second half of the 1945 season, appearing in 78 games, a big enough sample to use in our calculations. However, because he went straight from his military service to playing – coming into MLB play “cold” – we’re not weighing that as his first season back. In the calculations we’re weighing 1946 more heavily than 1945, essentially treating 1945 as his “second season” back in the calculations.

Both of the partial seasons are shown on the table below, but are not included in the recalculated career stats (being replaced by our full year projections).


One thing to note is that our projected stats are going to look very similar year to year. That’s because we’re using the average rates per plate appearance (and average plate appearances per game) to fill in career totals rather than try to predict specific seasons.

Also, even though his raw numbers are on par with the surrounding seasons, his stats relative to the league (wOBAWeighted On Base Percentage: A comprehensive measure of offensive production put on a scale roughly equal to “normal” on base percentage. .320 is average, .340 above average, .370 great, and so on. and wRC+Weighted Runs Created Adjusted – a comprehensive stat measuring offensive production relative to the league average) are all quite high; in fact, they’re some of the highest of his career. That’s because offense was way down during the war years and thus Greenberg’s (and the other hitters we’re looking at) stats would have been that much better relatively to league average.

In essence, while the raw numbers are projecting what might have happened if MLB as a whole avoided disruption, the stats weighted to league average show what might have happened if the just that specific player had avoided military service and the rest of the season played out as it did historically.

Hank Greenberg 1937 cropped

For Greenberg, the change in his career numbers is significant. His lengthy service likely cost him upwards of one quarter of his potential career and replacing those seasons boosts many of his totals by almost 50%.

Perhaps the most obvious change is his career home run total. During his career, Greenberg was one of the few players until the PED era to make a serious run at Ruth’s 60 HR record, hitting 58 in 1938. However, because of his lost seasons, he missed out on the 500 Home Run club by a wide margin.

With those seasons replaced, Greenberg projects to have hit just over 500 home runs for his career. At the time, this would have made him just the fourth player in MLB history to have done so, just behind Mel Ott (who also retired after the 1947 season). Even if he had hit a bit fewer during those seasons, nearing the 500 home run mark likely would have inspired him to hang on until the mark was reached (as he was still a productive hitter in his final season, if well down from his prime).

His career RBI totals also receive a major boost, placing him well within the top 25 of all time. At the time of his retirement, a mark of 1820 would have placed him 8th all time, again, behind Mel Ott (who had 1860).

Even though his career numbers took a hit from the time lost to military service, Greenberg still held the reputation as one of the game’s premier power hitters (and possessor of one of the all time great nicknames, “The Hebrew Hammer”). With his career numbers filled in with our projections, it’s clear just how deserved his reputation was.

Joe DiMaggio

Yankee great Joe DiMaggio enlisted in February of 1943. Like Greenberg, he enlisted in the US Army Air Forces, and like Greenberg, he served as a physical education instructor. Unlike Greenberg, DiMaggio never saw action in a combat theater. DiMaggio requested combat duty, but his request was rejected and he ended up serving in Hawaii, California, and New Jersey.

Army Air Force Sergeant Joe DiMaggio and Navy Chief Specialist Harold

During the war, DiMaggio’s parents, Giuseppe and Rosalia – both immigrants from Sicily, were classified as “enemy aliens” by the government. Required to carry identification papers at all times and barred from traveling more than five miles from their home without a permit, the restrictions prevented Giuseppe from practicing his trade as a fisherman during the war (he could not travel into San Francisco Bay and his boat was confiscated).


DiMaggio’s lost seasons were his age-28, 29, and 30 seasons – generally considered the prime years for a players statistical production. It’s also often said that DiMaggio’s post-war career lost just a bit of the magic that his pre-war years possessed.

Indeed, our projected numbers don’t quite reflect the brilliance of his years leading up to the war. Nothing resembles his early career power, his .381 average in 1939, the overall excellence of his 1941 season (capped, of course, by his 56-game hitting streak).

Of all the players we examined, DiMaggio is the one whose projections changed most if we looked only at the preceding years (as the Marcel system normally does), improving between 4-10%. This seems to bearing out the slight, post-war dimming of DiMaggio’s star.

Despite this, his projected numbers are still impressive and his career remains so. Unlike Greenberg, there are no new career milestones projected for DiMaggio. Instead, his new career numbers merely bear out what his actual numbers do – the well-rounded excellence of one of the game’s pantheon players.

Ted Williams

No other Major League player gave up more of his career to military service than Ted Williams. He served for three years in WWII and then another two during the Korean War. Because of this, the effect on his career numbers is even more significant than with Greenberg.

Williams initially had received a dependency deferment on the grounds that he was his mother’s sole means of support. When this resulted in public criticism and the loss of endorsements (notably Quaker Oats), Williams enlisted in the US Navy Reserve in May of 1942.

1944 Ted Williams Kokomo

Rather than play service baseball, Williams was as a Naval Aviator during the war. He served as a flight instructor and was preparing for combat deployment in the Pacific Theater when the war ended. Upon the end of his active duty he remained in the Marine Corps reserve.

Six years later in 1952, Williams was called up to serve in the Korean War. By this point a Captain in the Marine Corp Reserve, and despite not having flown an aircraft in eight years, Williams again refused to play service baseball as a way to avoid combat.

Over the course of his Korean War service, Williams flew 39 combat missions. On many of his missions, he flew with future astronaut John Glenn, serving as his wingman. Williams earned the Air Medal with two gold stars for his service and Glenn described him as one of the best pilots he had known.

As with Greenberg, Williams served partial seasons in both 1952 and 1953. The original stat lines are shown below, but they were not used in calculating the projections (being too small a sample) and are not part of the new career totals.


Obviously, five lost seasons had a massive effect on Williams’ (still impressive) career totals, but seeing just how much is rather staggering. Williams is often called the greatest hitter of all time, but his adjusted career numbers bear out just how strong an argument he has – his projected totals place him at or near the top of just about every offensive category.

Williams is currently 7th all-time with a .344 batting average – his adjusted total of .351 would place him 4th all time (behind Ty Cobb, Rogers Hornsby, and Shoeless Joe Jackson). In addition, he moves from 43rd all-time in doubles to 5th all-time (behind Cobb and ahead of 2015 Hall of Fame inductee Craig Biggio).

Currently 74th all-time in hits (and outside the 3,000 hit club), Williams leaps to 6th all-time, just ahead of Derek Jeter and just behind Tris Speaker. Also, Williams moves from 4th all-time in walks to 2nd all-time behind Barry Bonds. On top of all that, his already best all-time on-base percentage gets even higher, as does his 2nd all-time slugging percentage and 2nd all-time OPS.

Of this long list though, three stats in particular stand out. Williams now sits at 19th on the all-time runs list, well behind Rickey Henderson at 1st. With his five lost seasons added back in, Williams jumps to 1st all-time, almost 150 ahead of Henderson. Likewise, with his five lost seasons added back into his RBI totals, Williams jumps from 14th all-time to 1st, ahead of Hank Aaron (also ahead by almost 150).

Finally, Williams career HR total jumps from 521 to 685, placing him within a season or two of Babe Ruth’s total of 714. At his retirement, a total of 685 would have put Williams second on the all-time list behind Ruth. Would Williams have played another season to try and surpass the Bambino? It seems likely.

Ted Williams BBall Digest May 1949 raw

Williams was a notorious competitor (like most great athletes) and he still had the skills to perform at a high level. In his final season in 1960, despite being 41, Williams hit .316 with 29 home runs (in just 113 games), slugging .645. He didn’t play enough to qualify for the batting title, but if he had, he would have been tied for 6th in batting average in MLB, 1st in slugging (by 50 points), and 1st in wRC+ (by 16 points).

As it was, he was tied for 11th in MLB in home runs, despite having at least 125 fewer plate appearances than any man ahead of him. Suffice it to say, even at 41 Williams could still rake. Would he have stuck around to chase the record? Who knows. In a different world though, it could have been Williams’ record Hank Aaron broke instead of Ruth’s.

That’s the first three of our WWII-era greats, come back later this week for part two (view it here).

Comments (must be logged in to Facebook)