Horsehide Hindsight

So we're rapidly approaching the start of a new baseball season, and I'm beginning to get that itch again, as I do every time this year. And I will go out on that proverbial limb again and make my predictions by the first week of the season. But first, I think it would be appropriate to review how I did last year as a prognosticator.

My results were decidedly mixed in 2010. I knew going in that my process of refining my formula for evaluating teams had gone awry, that flaws had been introduced into the mechanics of the thing. Too many teams were predicted to finish with too many wins, probably because I gave too much weight to minor team characteristics; i.e., giving too much credit for having a good manager or bullpen closer, or for having a solid hitting lineup. In my experience, these are mostly minor details; they help, but only for a few games a season. The real meat of the matter lies in starting pitching; having four or five good starters does more for a team's chances than any amount of sluggers or managerial wisdom.

The flawed formula led me to make some preposterous predictions about win totals. I somehow concluded that four teams would finish in excess of 100 wins (something that I'm sure has never happened before), and I had another team pegged for 99 wins. As it turned out, none of those teams won 100 games; two of them actually made the playoffs (NY Yankees and Philadelphia), but only one--the Phillies--even won a division title. Boston did finish with a respectable 89 wins, but failed to make the postseason. The real disasters were the Los Angeles teams. The Angels and Dodgers, picked to win 106 and 99 respectively, both flamed out and sank in the standings--the Angels finishing a staggering –26 below their predicted win total.

There were other disasters, too. Seattle unexpectedly stank up the joint, finishing –25 games from my prediction. The other wild swings and misses dramatically improved on my predictions; the Reds beat the rap by winning +19 over my figure, while the Rangers rode a +16 in wins all the way to the World Series.

Nevertheless, there were some triumphs, too. Most notably, I predicted the order of finish in the NL East perfectly, and all five teams finished within ±4 games of my win-total forecasts. In all, I nailed the exact record for three NL teams: the NY Mets, Pittsburgh, and Colorado all finished exactly where I predicted with exactly the records I predicted. The AL East almost finished in order of my predictions; only Tampa Bay, sneaking up from third place to take the division crown, ruined that bit of divination. Eleven of the 30 teams finished in exactly the place in the standings that I predicted.

Overall, out of 30 teams, 14 teams finished within a ±4 range of their predicted wins. That's some pretty tight predicting for almost half the teams in the majors; given that we're talking about a 162 game schedule, missing the final number of wins by no more than 4 games is damn good. (It comes to 97.6% accuracy for those 14 teams.) Eight of the remaining 16 teams fell within a ±10 game margin of error; only the eight remaining teams were wildly off the mark--and two of those teams, Boston and NYY, did finish with high win totals and in the upper half of their division, just not in the positions nor with the win totals I predicted. In all, seven teams were predicted to be winners and finished losers, or vice versa. Most teams meekly followed orders and spent the season as winners or losers as predicted.

So all in all, the mojo could use a tune up, but even so the results were decent. I expect this season will provide my greatest triumphs yet, when I post my preview predictions in about three weeks. So keep watching this space and place your bets early.