Dugout Central Challenge Updated Update
Before the season began, Dugout Central staff and readers were challenged to predict win totals for all thirty MLB teams. A total of five entries were received. For details about the contest rules. [Author note: I misplaced John Bowen's and forgot to include him in the original article, as John very politely pointed out to me. He has now been added, although he may wish that he'd left well enough alone.]
Here are the predictions, by division, in alphabetical order. The Over/Under win totals from the Las Vegas Hilton are also shown.
|Jon Ellis (Seven)||90||80||79||95||86|
In the NL East the Phillies are favored by everyone except for Bob, who chose Washington. The Mets are expected to be mediocre to bad by everyone.
|Jon Ellis (Seven)||70||85||69||85||67||85|
We have the NL Central as a three-horse race, with St. Louis, Cincinnati, and Milwaukee all getting first-place votes. Houston is expected to trail the field, although by varying degrees.
|Jon Ellis (Seven)||85||72||83||69||87|
The Giants got three first-place votes, the Dodgers two and the Dbacks one. San Diego is the clear favorite to claim the cellar. John and I are the only ones to have the Dodgers under .500 and the Rockies over .500 (which at the moment is not looking so good).
Wild card votes were all over the map, with Arizona, Atlanta, Cincinnati, Colorado, Miami, Milwaukee, Philadelphia, and St. Louis all getting a vote or part of a vote — there were some wild card ties, and the contestants were not asked explicitly to choose a wild card team.
Jon Ellis had the most extreme NL picks, predicting the Rockies to win only 72 games, both 7 games below any other prediction. We were mostly in agreement on the Brewers, as our picks differed by at most 3 games for Milwaukee.
As a group we predicted the Cardinals to win 1.8 games more than the Vegas line, our most optimistic choice. (Note that the Vegas line averaged 81.7 wins, so apparently bettors in general are slightly optimistic.] We were most pessimistic on the Giants, 3.3 games below the Vegas line. Overall, our average predictions were fairly close to the Over/Under number.
For the AL we have:
|Jon Ellis (Seven)||68||86||90||94||83|
We favor the Yankees and Rays to win the AL East. The Orioles are expected to be mediocre to bad by everyone.
|Jon Ellis (Seven)||79||72||95||74||78|
We are unanimous on the Tigers to win the AL Central, the only division where we all agree on a division winner. As a testament to the weakness of this division, the White Sox, Indians, Royals, and Twins all got votes to finish in last or tie for last.
|Jon Ellis (Seven)||96||67||66||95|
As might be expected, the Angels and Rangers were our choices for division winner, and the A’s and Mariners are predicted to trail by a wide margin in the AL West.
We were almost unanimous on which five teams will make the playoffs in the AL – Anaheim, Detroit, New York, Tampa Bay, and Texas. John and I were the only exceptions, predicting the Red Sox to beat out the Rays for the second wild card spot.
Bob Owens had the most extreme AL picks, predicting the Mariners to win 77 games, both 7 games above any other prediction. We were mostly in agreement on Anaheim, Boston, and Toronto, as our picks differed by at most 6 games.
As a group we predicted the Rays to win 3.7 games more than the Vegas line, our most optimistic choice. We were most pessimistic on the A’s, 5.7 games below the Vegas line. Overall, our average predictions in the AL were fairly close to the Over/Under number, although not quite as much as in the NL.
So, how are our predictions doing so far? There are two different contests – one uses average win difference (AWD) and the other root mean square difference (RMSD). The difference scores are projected to the end of the season in two ways – extrapolate the current team wins to 162 games, the other to prorate our 162-game predictions to the current number of games. One method overstates the error, while the other understates it. By taking the geometric average (square root of the product) of the two methods those effects tend to cancel out. The standings below use this combined score. In practice, the two projection methods give very similar results for the ordering, even though numerically they are far off from each other and the final scores.
The projected scores for the two contests are shown in the following table, ordered by the AWD score. All games through May 16 are included.
Also shown are the average picks for the five of us, the Over/Under line, predictions using last year’s win totals and last year’s Pythagorean wins, and the simple-minded choice of 81 wins for all teams.
Bob leads in the both contests. Jon is second in AWD, while Patrick is second in RMSD. John trails the field in both contests in the early going.
Updates will be provided periodically throughout the year using winning percentages, so there’s nowhere to hide, and may the best prognosticator win!