In 2010, the Captain’s Blog was a veritable crystal ball when it came to predicting the standings, at least when compared to a universe of experts and projection systems. How’d I do this season? Well, as John Sterling might say, you can’t predict baseball.
Note: Positive number indicates forecast was above actual total. Negative number indicates forecast was below actual total.
One season after averaging a 6.4-win divergence per team, my predictions deviated from the actual by 8.4 wins. In addition, the combination of exact predictions and “near misses” (defined as being within five wins of the actual total) plummeted all the way from 20 to 12. Meanwhile, projections that were “off the mark” (defined as a prediction 10 games or more from the actual total) increased from eight to 12. What’s more, the magnitude of those significant misses also increased. Whereas in 2010 my worst prediction was a 20-win divergence from the Padres’ final total, this year, two teams digressed from my forecast by at least 24 wins. Maybe I shouldn’t be so hard on ESPN?
Dead On refers to exact predictions, or where a simulator was done, those with an absolute value rounding to 0.
Near Misses refer to a prediction within five games of the actual total.
Off the mark refers to a prediction 10 games or more from the actual total.
In an attempt to save face, it should be noted that 2011 predictions were less accurate across the board. Of the 10 prognosticators considered, the leading average number was only 7.4, and all but three missed the cumulative mark by at least 8 wins. Also, my 12 near misses or exact predictions trailed only Yahoo!’s Jeff Passan for the highest total among those considered, and six individual team projections, including the Yankees, ranked as the most accurate. Still, nothing can excuse missing six of eight playoff teams, not to mention picking the Cubs to win the NL Central.
On a team basis, the Mets proved to be the most transparent team. All 10 prognosticators came within 3 games of pegging the Mets win total, and the average absolute value divergence was only 1.9 games. In addition, the average win forecast for the Amazin’s was dead on at 77. The Braves, Dodgers and Mariners were also pegged to their exact win total.
The Diamondbacks exceeded the aggregate expectations by a whopping 24 wins, by far the largest total of all teams. Baseball Prospectus’ PECOTA had the most generous forecast for Arizona, but still fell 18 wins short of the NL West champion’s final total of 94. On the flip side, the team that most disappointed was the Twins, which fell 23 wins below expectations. CAIRO was the most pessimistic about Minnesota, but still overestimated the team’s prospects by 19 wins.