Every spring Baseball Prospectus dusts off its algorithm, the formula that guides the computer that crunches the numbers and makes Baseball Prospectus devastating competition for Hot Type's Golden BAT.
Hot Type has nothing against algorithms —in their place. They do useful work when applied in moderation and under expert supervision. Hot Type even believes they have a place in baseball—though our national pastime did fine without algorithms for decades, coming no closer than Al Griffin, who led the American League in triples while playing for Toronto in 1980. Though some might point to a rousing speech given in New York City in 1993: "These days, there are people that attack baseball," said the vice president of the United States. "'Where have you gone, Joe DiMaggio?' asks the cover of last week's Sports Illustrated. Ridiculous. That was a good song but a dumb question. Joe is right here." It was an insightful Al Gore-ism.
But I digress. When the choice is between an algorithm on the one hand and what Hot Type generously chooses to call human wisdom on the other—well it's no choice. And that's why I'm pleased to announce that Ted Cox is this year's winner of the coveted Golden BAT. Cox made his picks in last year's pennant races the old-fashioned way. He played his hunches.
The Golden BAT, as you know, has been a Hot Type fixture since 1981—the same year a certain chewing gum company sold the Cubs to a certain media empire. Hot Type was under more coldhearted management back then, and the idea was less to hail the sportswriter who the year before had submitted the most accurate pennant picks than to demonstrate that by and large the bards of the press box had no more idea what they were doing than blindfolded monkeys at a dartboard.
That case was made and remade, and eventually ceased to be interesting. Over time, the present custodian of this space found himself increasingly moved and deepened by the sheer humanity displayed by the winning scriveners, whose delight at winning is so intense that it's redefined the BAT—which I see now as the one consolation a sportswriter can aspire to for that daily parade of wet towels flicked by steroid-addled noncommunicants in reeking locker rooms that constitutes his professional career.
"Really? Oh my god! That's something to crow about," said my friend and colleague Ted Cox when I delivered the good news. Cox is, in a way, a perfect representative of the modern newspaperman. He won his first BAT back in 1985 as a paid Reader sports columnist; he wins his second as an unpaid Reader sports blogger. Fortunately for the Cox family, sportswriting has always been just a sideline: he holds down a paying job as a feature writer at the Daily Herald.
Last year the Baseball Prospectus algorithm—which has no mouths to feed, nor any reason to lie awake at night fearing layoffs—swept to victory in the BAT. PECOTA, to call the soulless enemy by name—it's short for Player Empirical Comparison and Optimization Test Algorithm—had correctly predicted six of the eight teams that reached the playoff in 2007, while a seventh team it picked wound up a game out of first. It exactly predicted the White Sox' 72-90 season and overshot the Cubs' 85-77 record by one win.
Endlessly refinable, PECOTA looked a year ago like the irresistible future of baseball prognostication. It may yet be; but this time around Cox fought it to a standstill, and in Hot Type's book the tie favors the mortal. BAT, after all, stands for Baseball Aptitude Test, not Bloodless Algorithmic Technology.
And to be honest, the algorithm didn't ace 2008 like it did 2007. Its performance wasn't disgraceful—it did foresee that the Brewers would end 2008 as the NL wild card team, and in a bit of showing off it accurately predicted that the Tampa Bay pitching staff would end up allowing the third fewest runs in the majors. But it named only five of the eventual playoff teams: Boston, the Los Angeles Angels, the Cubs, Milwaukee, and the LA Dodgers. And though it picked the Cubs to win the NL Central with 93 victories—they won it with 97—it figured the White Sox for only 76 wins and third place in the AL Central, which they led with 89 wins.
The father of PECOTA, Nate Silver, had a sensational year personally, but he had it over at fivethirtyeight.com, the political Web site where everyone I know checked in first thing each morning during the presidential campaign and last thing each night.
As for Ted Cox, he too named the Red Sox, Angels, and Cubs to play in the postseason. He added Philadelphia in the NL East; and he and he alone in the field of 16 picked the White Sox to triumph in the AL Central. Everyone else liked either Detroit or Cleveland. Says Cox, "I had a good feeling about the White Sox and I didn't like the Tigers or Indians—the chemistry they put together."
We don't do algorithms here at Hot Type—we're partial to clean, hard common sense. PECOTA is frequently tweaked, and we did some tweaking of our own to put Cox over the top. By our old scoring system, PECOTA would have edged out Cox for the BAT because it accurately gave Boston the AL Wild Card. Cox believed the Beantowners would make the playoffs by leading the AL East.
But here's the thing—the two teams everyone focused on in that division were the Red Sox and the Yankees. No one saw perennial also-ran Tampa Bay winding up ahead of them both. PECOTA lucked into getting the Red Sox right: it picked them to finish second behind the Yankees, with both teams reaching the postseason. Cox put the Bosox ahead of the Yanks. That was the shrewder prediction and I'm rewarding it.
This brings us to the touchy part of the ceremony. It's the Wiffle BAT, a prize for failure. Paul Sullivan of the Tribune and Gordon Wittenmyer of the Sun-Times can each lay claim to it—they named only two playoff teams apiece. But let's make it a collective Wiffle. The Phillies beat the Devil Rays in the 2008 World Series, and no one human or algorithm saw either of them winning a pennant.v
Care to comment? Find this column at chicagoreader.com. And for more on the media, see Michael Miner's blog, News Bites.