April 8, 2004

Two Thousand Years Of Willie Mays. That's the title of a short Bill James article I mentioned back on April 5. I said that it had appeared in one of his early Abstracts, but it turns out it was published much later, in The Baseball Book for 1992. (And there's a good reason why we need stats; because personal recollections can be way off the mark.) Anyway, the 2-page bit begins:

"What it is possible for a great player to do in a great year? Suppose that a great player, a Willie Mays or a Joe DiMaggio, hits in good luck for an entire season. What kind of stats could he compile? Suppose that he hits in bad luck for an entire year, what then? ... What if he played forever ... What are the limits here?"

It took James "about three days on my little home computer" to simulate 2,000 seasons of Willie Mays. Here are a few of them:

Year#   AB   R    H    AVG  HR  RBI
  29   583  145  201  .345  50  127
 226   562  137  203  .361  45  133
 826   559   88  132  .236  21   89
1203   553   90  141  .255  16   79
1800   580  132  198  .341  50  149
James concludes that it's possible for a player to hit .360 one year and .240 the next just by random chance. Mays's highest HR and RBI totals were 54 and 155, he never came close to hitting .400 and his longest batting streak was 34 games. ... James then ran 1,000 years of both DiMaggio and Wade Boggs. Here are two consecutive DiMaggio seasons:

Year#   AB   R    H    AVG  HR  RBI
  6    592  135  200  .338  38  150
  7    553   96  146  .264  18   79
James: "Suppose that a real player did this ... won the MVP Award if not the triple crown, and then the next year was an ordinary player. Every explanation under the sun would be offered for his decline, and ultimately one explanation would come to be the accepted one -- he got fat and lazy after a winter of celebration, or he was affected by the absence of X, who hit behind him in his big year ... or even that he began pressing after a slow start. Do you suppose that anybody would write that he was the same player, just hitting in tougher luck? That's my point; that when a player has an off year, even a terrible year, there may not be any reason for it. It may just be the breaks of the game."

In 2,000 years, the longest hitting streak Simulated DiMaggio compiled was 48 games. ... Simulated Boggs made repeated runs at .400. In Year 358, Sim Boggs batted .425, breaking Rogers Hornsby's record by .001. James writes: "Think about it: after trying 357 times to hit .400 but failing all 357 times, the simulated Boggs not only hit .400, but cleared it with fourteen hits to spare. It's a Bob Beamon feat."

James calls this whole thing a "silly exercise," but I've always thought it was pretty cool -- and a fantastic example of what is so instructive and entertaining about James's best work. It may be silly on some level -- but it also shows us that flucuations in performance should be expected. ... Jason Giambi hit only .250 last year, but was still in the top 5 in OBP; his low batting average was nothing more than a random fluke. ... Knowing stuff like this, you can then laugh at (or ignore) broadcasters and writers who go on about 6-game hitting streaks or what some batter did in 10 at-bats against a pitcher over 5 seasons. It's completely meaningless.

This little article has always been, for me, a great reason why James should have published a "Best of" of his Abstracts/Baseball Books. ... I also wish James would reprint his very early (late 70s) self-typed and stapled Abstracts in one volume. While a lot of the information would be quite dated by now, the costs of reprinting would be minimal and I know there are thousands of fans who would snap it up sight unseen.