I could have conducted this experiment, identifying the Greatest Hitter in Baseball at Any Given Point, not on an annual basis, as I have done, but on a daily basis, as Bill James is doing now with his ongoing Greatest Pitcher in Baseball project. Well, I couldn’t have done that, but someone with more resources and more brains than I could have done it. It would have yielded a much faster turnover rate, a new champ every few weeks, perhaps, rather than five champions (Ruth, Foxx, Gehrig, Mize, Williams) in fifteen years, but in addition to the logistical nightmare that I’ve spared myself (and you) in setting it up the way I have, this is also the more satisfying philosophical way to approach the question.
It’s a conservative approach: you cannot be The Best simply by having a 4-HR day or a 35-RBI month, though they’re both indications that you’re Pretty, Pretty Good if not The Best. As my provisional, partial, pretty good study has shown so far, there are points in history where the title falls into dispute—among several other points, it’s a close call in the winter of 1935 who the best hitter in baseball is, Foxx or Gehrig. As I explicate above, there are good reasons for choosing either one of them, and I settle the issue in my typically conservative fashion by choosing Foxx, mainly because he had held the title going into the year and Le Cheval de Fer hadn’t quite run him over. Gehrig had simply forced the judges to make a decision as to who had won the fight. My way of doing this study is to favor the sitting champ when a close call must be made, sort of a "tie goes to the runner" rule.
And actually, if you’ll allow me a digression within a digression for a moment, that rule isn’t quite as arbitrary as I’d thought for years it was: when I was reviewing Doug Harvey’s obnoxious autobiography a few years ago, he made the sound point that having a rule was more important than serving some idealized form of justice, in that following a principle, however ill-founded or illogical, led to a consistency that following justice didn’t lead to. One judge will find justice where another will see injustice, and the same judge might find justice in an unjust decision if he’s had an especially satisfying breakfast that morning. Harvey’s principle, in other words, was that a good umpire should have the eyesight and the sense to make the right call on a close play at first base 99% of the time, and other 1% should be decided the same way every time: in the runner’s favor, because in Harvey’s view, the throw had to beat the runner, not the other way around. If it was unclear that the throw beat him to the base, then the runner got the call. It was better, in Harvey’s view, to make the same, consistent call (instantly) if the decision was hazy in the ump’s mind. (And, of course, to be adamant in defending that call to any outraged manager who flew out of the dugout to challenge the call.) Subjectivity, in other words, led to inconsistency which led to managers thinking that the call was made according to the ump’s feelings about the players, the teams, the game situation rather than his vision of the play. Better to have the managers thinking "Harvey always gives the runner a break on close plays" than "Harvey doesn’t like us and he screwed us that time."
In that conservative sense, I don’t want to be making judgment calls here. Wherever it’s possible to avoid them, I want to. And that means adopting a principle of "Barring clear-cut decisions, the crown stays on the champ’s head."
I’m unsure if the single metric I’m using here, that of OPS+, makes for more turnover or less turnover in title-holders. At first blush, I’d assume that using multiple metrics—Win Shares, or WAR, or MVP shares or Total Bases—makes it easier to dislodge a sitting champion. After all, multiple metrics might conflict with each other and present complicated competing cases to be made for certain runners-up. Hank Greenberg, for example, won the AL MVP award outright in 1935, which must improve his case for Best Hitter considerably. But on reflection, maybe multiple metrics make it harder to dislodge a sitting champ. The path to the title, remember, is to achieve a solid victory over the reigning champ, and more metrics just introduce more ambiguity into the equation. The overarching principle here is "No titles change hands without a clearly demonstrated superiority to the current title-holder."
Which means rare changes in title-holders. But here, in 1942, I have a very subjective call to make: with Williams off to war, is he still the champ in 1943 or is he not? You can imagine the good arguments either way for such a call, so I won’t go into them in great detail. The real problem isn’t the title-holder in 1943, 1944, or 1945—I could just continue my process throughout the war, and identify somebody (probably Stan Musial) as the title-holder for those years. The real problem is what to do in 1946.
Does the crown revert to Williams? Does the holder as of 1945 (who probably isn’t Musial anymore, since Musial served in the military that year) keep it until Williams, or someone else, earns the three-year title back in 1948? To answer that, I have to go back to the question I’m seeking to answer here: Who, as far as it is knowable, is the single best hitter in the world at any point in time?
Is Ted Williams, in flight gear at a Navy base in northern Florida, still the best hitter in the world?
Maybe. Probably. Possibly.
But not definitely. In retrospect it is much easier now to call it "Williams" than it was at the time. In retrospect, we know that Williams returned from the war (returned from two wars, actually) as capable as he was the day he joined the Navy.
But at the time, it was an open question. Plenty of players did, in fact, lose more than just the war years—they lost some of their skills during a two-or three-year layoff and never made it back. We couldn’t even know if Williams would literally make it back, of course, or make it back in one piece, and we wouldn’t know until after the war if he had suffered some injury serving in the Navy—wrenched his back permanently squeezing his lanky body into the tiny cockpit of a fighter plane or suffered a concussion making a hard emergency landing. We wouldn’t know. We couldn’t know. We didn’t know.
So to answer my original question, I think we have to accept that, during the war, that question will be answered better with a series of names that are undeserving of a championship, perhaps, but actually belong to the best hitters playing the actual game at that point in time. If this leads to chaos, or uncertainty, or a rapid turnover rate, well, war is hell.
(If you want to award the title to Williams beginning in 1942 and ending who-knows when? 1948? 1952?, then by all means feel free. It may be the right choice. I can’t see, however, how you can easily decide that Musial, say, or someone else, is NOT the title-holder after a strong season that Williams has missed. There’s no sure way to be right or wrong, and it’s sorta fun figuring out the mid-1940s rankings, invalid though they may be.)
This may be the right place for me to voice an unpopular opinion about players serving in the military (or otherwise missing games for non-baseball-related reasons such as an illness or a work-stoppage): sometimes, they may benefit from the games they missed. The presumption, with which I agree mostly, is that Williams was done great harm by serving five of his prime seasons in the military, but in the back of my mind I also ask: what if military service (or illness, etc.) actually prevented a baseball-related injury?
Flying fighter planes in combat may not be the best example of safe military service, though some players did serve in relatively safe positions during World War II. (DiMaggio, for example, spent much of his service in the hospital being treated for ulcers or in Special Services.) But to sidestep the argument about war-time dangers here, let’s use someone like Red Schoendienst as our example. Schoendienst (or as I’ll call him, "Red," to save repetitive stress on my typing fingers) missed almost the entire 1959 season because he came down with tuberculosis. Now, while I wouldn’t recommend contracting TB as a wise health choice, Red might have used his recovery time to heal in other ways. If he were suffering from one of any number of the health issues that 35-year-old second basemen are prone to—like worn-out tendons, or sore backs, or stressed-out knees, ankles, and wrists—then maybe the season he spent recovering from TB gave him time to heal those body parts and actually extended his career.
I’m not arguing that this was the case with Red, or with any other injured player specifically. I’m just trying to argue that there can be an unseen upside to missing games, that of involuntary rest or relief to stressed-out body parts. Bill made this argument, effectively, in regard to pitchers like Warren Spahn or Nolan Ryan missing a chunk of games early in their careers that—who knows?—may have gotten them safely past some strain on their growing arms that allowed them to have incredibly long and fruitful careers.
Another way to look at it is that the baseball diamond isn’t exactly safe—we can’t look at it as "Playing MLB is safe but missing time in MLB is very dangerous" because of the career-threatening (or -ending) injuries that take place during MLB games. If you think this point is specious, just ask Tony Conigliaro or Herb Score or Ray Chapman or Dickie Thon or Justin Morneau or Dizzy Dean. You probably won’t get much of an answer, but time spent off the diamond isn’t any more dangerous than time spent on it is safe.
So here’s 1942:
Oops. Just oops.
Not OPS, oops. I seem to have forgotten that Williams played all of 1942, so take my remarks above about 1942 and delay them one season. Obviously, Williams has a clear claim to the undisputed heavyweight crown for 1942—both runnersup, Mize and DiMaggio fall a little further behind Terrible Ted. I’m not even going to bother computing the blue three-year chart for 1940-42 this time—it’s obvious who heads it, and who’s way behind on it. Let’s go straight to 1943, where the rankings should get lively:
OK, basically everyone on the 1943 ranking is a newbie. (Eric Tipton joins the ranks of "players I’ve never heard of before today.") With everyone you’ve heard of off to war, except for Musial, the three-year chart for 1941-43 is all new, too, and because this is only Musial’s second season, he doesn’t lead the three-year chart.
So far, we’ve seen all the three-year charts being topped by at least two hitters with over 1000 OPS+ totals, but this year there are none. The best hitter on the planet in the winter of 1943 (if you don’t count men who didn’t play in 1943 like Williams, who still scores 667, based on his 1942 and 1941, which is amazing, and Joe D.) is King Kong Keller.
|
1941-43 rank
|
|
OPS+
|
|
1.
|
Keller
|
954
|
|
2.
|
Nicholson
|
939
|
|
3.
|
Heath
|
883
|
|
4.
|
Ott
|
882
|
|
5.
|
Musial
|
833
|
Charlie Keller is the Greatest Hitter in the World? There must be something screwy with my system.
And there is. There are. Plenty is wrong, but I’ll stand by the statement that Keller is the world’s # 1 hitter in 1943. Things that are screwy:
1) Ted Williams is #1 if you want to look at it like that, which I choose not to but which I respect. Actually, Williams’ OPS+ for 1941-43 is damned good, without even considering that he is missing the three-times-counted 1943: he clocks in at 667, just by counting his 216 OPS+ (for 1942) twice and his 235 (for 1941) once. In other words, he averages a 111 OPS+, or 11% better than average while missing half the years we would normally count in this system.
2) Keller was a damned good hitter. He had a lifetime 152 OPS+ in 1138 games. But he was even better than that in his first five seasons, of which 1943 was the fifth: he had made three All-Star teams and had gotten MVP shares in four seasons by that point. He was on track to make the Hall of Fame, though the war (he joined the Merchant Marine before the 1944 season) and a bad back derailed him. As of 1943, he was a great hitter—we can’t put so much on his post-1943 career that we forget entirely about his standing at the time.
3) The war was a bitch. When I wrote before about players who "lose some of their skills during a two- or three-year layoff and never made it back," I was thinking of Cecil Travis, Bill’s favorite frostbitten example, but it also applies to players such as Keller, guys who came back from the service but for whatever reason just were no longer the same player they were. This happens all the time in baseball, of course, so we can’t pin it completely on "the war" as the sole reason, but the frequency of instable careers generally makes the mid-1940s screwy for identifying a stable hierarchy of great hitters. There’s a lot of turnover here, not only at the top of the rankings but all up and down the list.
4) Stan Musial, the MVP in 1943, has an 833 OPS+ for 1941-43 despite getting 0 OPS+ points for 1941, when he had a 179 OPS+ but in only 12 games. You might say "Musial was the game’s best hitter after the 1943 season," and you might be right—this is one spot where my system’s conservative values might well be a skosh too conservative.
5) Of course, you can also look at the title as being more like this: the champ, Teddy Ballgame, has given up his crown part-way through the season. Statistically he must have held the crown for quite some time into 1943, since he had a lock on it on Opening Day. And since Musial banged the hell out of the ball when he came up in 1941 (a 179 OPS+ in his twelve games) we might extend his championship quite a bit further back in time, and call the Keller championship the result of Teddy B. failing to defend his crown, rather than being knocked out in the ring. (When Muhammad Ali was stripped of his title in real life, you will recall, there was a period in which no one really knew who the World’s Champ was. The Heavyweight Boxing Champion of the World was, for a period in the late 1960s, "Hoothe Hellnose.") The chaos becomes clearer in 1944:
For the first time, the 1944 rankings are headed by someone who I’m unfamiliar with. I know that this guy is "Indian Bob" Johnson and that his brother Roy played MLB and that he was a power hitter, but when I see "Bob Johnson" I first think "The dude who freakishly batted .348 for the Mets in 1967?" and then I think "The other dude the Mets traded to Kansas City for Joe Foy in 1969?" (there was another throw-in in that deal) but this Bob Johnson was a stellar slugger, not least considering that he broke into MLB at the age of 27 but still managed to put in 13 years, never getting an OPS+ lower than 125. 1944 was hardly a career year for Indian Bob, though, despite his MLB-leading OPS+: looking at his Total Bases, he had five other years where he got 20 more Total Bases than he got in 1944. He was a helluva hitter. His career-high here, 174 OPS+, must be marginally better than Stan Musial’s 174, somewhere around the fourth decimal place, getting him the lead for 1944, but the championship goes to Musial, who (finally) has put in three full years:
|
1942-44 rank
|
|
OPS+
|
|
1.
|
Stan Musial
|
1027
|
|
2.
|
Bill Nicholson
|
973
|
|
3.
|
Mel Ott
|
949
|
|
4.
|
Bob Johnson
|
911
|
In addition to Johnson showing up here, there are other wart ime oddities all over the place. The man who finished right behind The Man in 1944, Bill Nicholson, is perhaps the archetypal wart imes tar, a guy who led the NL in HRs and RBIs two years running (1943 and 1944) but never quite matched those levels again after the war (or before it either, though he had some other years—1940, 1941, and 1947—he hit well in). We don’t think that highly of Nicholson these days, but he finished not too far behind Stan Musial in 1944 OPS+-wise, and they were facing the same exact pitching. Mel Ott also had a good year in 1944. He has to win some sort of award for longevity—he’s been showing up on these rankings since the late 1920s, often very near the top, and he’s still not completely done in 1944.
The question may be chafing at you by this time, if not well before, whether OPS+ is the best single measure to use in figuring out offensive prowess. Let me repeat: of course OPS+ isn’t sufficient. It is one of several metrics that will have to be used collectively to decide the correct title-holder. But using three or four other metrics doesn’t make the title more elusive. Take our starting point: I decided that at some point in the mid-1920s, Babe Ruth was the title-holder, which I don’t think is in serious dispute. Certainly this method, using three years’ worth of OBP+, measured in just this way, supports that assertion—but so would most other metrics. To topple Ruth, I learned that his OPS+, that towered over baseball for the first few years of the 1930s, gave way to Jimmie Foxx’s in 1933. I’m pretty sure that other metrics would yield the same conclusion, maybe not all at the same exact point in time but very close, and that Foxx’s OPS+ lead would yield in more or less the way and at the same rate if we threw in a bunch of other metrics. I won’t swear that every single result would be the same no matter how many ways we applied to looking at the qualifications for the title, but I’d bet a lot of money that mostly they would follow the OPS+ results.
Steve161 commented (in Part 2) that an OPS+ difference is not a significant one, and I’ll take a second to clear that up a bit. OPS+ is WAY different from just-plain OPS, in that it measures batters against the league; just-plain OPS measures their absolute statistics, which is why OPS is subject to the era and the year in which they’re made and OPS+ applies across the board. The 25-point difference that steve161 was dismissing is a HUGE difference, and is certainly significant.
OPS+ is just the easiest example I could find to embody total offensive contributions. I could make it even easier to use, and I’d almost think I should. Instead of looking as closely as I have at the exact three- or four-digit OPS+ for each year, I could have gotten similar results by sticking to the rankings (left-hand column) rather than the OPS+ numbers (right hand column). This is especially true if we have a bunch of metrics: if the guy who has led in 4 out of 5 measures for the past few years suddenly leads in none of them anymore while a newcomer takes the lead in 3 of the 5, then the newcomer becomes the champion. But for now, I wanted to show the process with precise stats, so I used the actual OPS+ numbers. You could probably approximate the same results with the rankings in multiple categories, with very few changes in the identity of the champion hitter.
It would be challenging, but not impossible, to include the whole of a player’s contributions, not just his offensive contributions, which is what OPS+ is measuring. Metrics like Win Shares and WAR (supposedly) account for defensive contributions, so if we added those to the mix, in some form or other, that would get us one baby step closer to "Best Player in MLB" and not just "Best Hitter." Again, I wanted to give the least complicated example first time around, which is why I went for "Best Hitter" but honestly I doubt that defense would change the title very often if at all. If Joe DiMaggio were hot on Ted Williams’ heels at some point, then, yes, Joe D.’s superior fielding would tilt the decision in his favor.
Except we haven’t seen a case like that so far. Ruth was pretty far in the OPS+ lead up until about 1932, so he wouldn’t have been toppled by a defensive star unless that star were OPS+ing as well as Foxx or Gehrig. The 1932-34 and 1933-35 results were, if you’ll recall, virtually tied in OPS+ between Foxx and Gehrig—maybe defense would have pushed Gehrig to #1 if he were Keith Hernandez to Foxx’s Dave Kingman. Was there a huge advantage for one huge first baseman over another in defense? I’d need to look at the defensive stats to make that call, but generally most of these sluggers were all the way at the right end of the defensive spectrum (Arky Vaughan and Hornsby are the exceptions) and none was a defensive wizard as far as I know, so it’s hard to see how a lot of ground could made up by comparing their fielding skills. (Their dWar numbers are both in the low negatives: -0.5 to -1.0 over the course of their careers. Bill’s defensive win shares book rates Foxx as an A and Gehrig as a B-, so I don’t see how that helps Gehrig’s case for breaking the tie in his favor.)
It gets pretty hairy in 1945—really, 1945 is what we’re talking about when we talk about wartime baseball. It’s the bottom of the barrel.
Only Mel Ott is left on the top ten OPS+ list as a legit HoF candidate, much less a legit HoFer, and his career as a perpetual bridesmaid finally pays off for him:
|
1943 -45 rank
|
|
OPS+
|
|
1.
|
Mel Ott
|
931
|
|
2.
|
Phil Cavarretta
|
906
|
|
3.
|
Tommy Holmes
|
888
|
Ott gets to claim one slightly dented war-time crown—again, a good case can be made for Stan Musial, and a better one for Ted Williams, both off to war, as the best hitter in MLB, except in 1945 they’re not really IN baseball, and Ott has been close to the top of the rankings for a long time, so this temporary title seems reasonable. (It’s Ott’s last year as a regular.) The other contenders’ OPS+es are weak sauce as these things go: Cavarretta’s three-year 906 and Holmes’ 888 would have been far down the rankings before the war. It’s sort of fitting that for the third war-time year in a row we have a different one-year title-holder.
Let’s push ahead into the post-war period.