Batter's Box Record Prediction Contest Results

Sunday, December 13 2009 @ 08:35 PM EST

Contributed by: Dave Rutt

Well, the regular season ended over 2 months ago, but it's never too late to re-visit the crazy predictions we made way back in March. These are the results of the Batter's Box Record Prediction Contest. Click through to find out who will be taking home the inaugural title of Da Box Resident Seer (and can someone come up with a better title?).

A reminder of how the contest worked. Contestants were asked to predict the regular-season record of every team, and their score is calculated by the average deviation in terms of wins. For example, if you predicted the Jays to win 98 games, that's a deviation of 23 since they only won 75 (sorry to rub it in). The total score is the average of the deviation for all teams. First, some data (sorry about the ugly chart):

Team  Actual Wins
Average Box Prediction
Bang-on Guesses
New York Yankees
103 91
none
Boston Red Sox
95 94
AWeb, Ron, seeyou, nathanelias, esteBENloaiza
Tampa Bay Rays
84
89
none
Toronto Blue Jays
75 82
parrot11, Ron
Baltimore Orioles
64
72
electric carrot




Minnesota Twins
87 83
esteBENloaiza
Detroit Tigers
86
81
parrot11, 92-93
Chicago White Sox
79
78
Mudie, 92-93
Cleveland Indians
65 85
none
Kansas City Royals
65 75
none




Los Angeles Angels
97 88
none
Texas Rangers
87 76
none
Seattle Mariners
85 72
none
Oakland Athletics
75 85
none




Philadelphia Phillies
93 88
esteBENloaiza
Florida Marlins
87
78
none
Atlanta Braves
86 84
PECOTA
New York Mets
70 90
none
Washington Nationals
59 66
none




St. Louis Cardinals
91 83
none
Chicago Cubs
83 92
none
Milwaukee Brewers
80 83
Mudie
Cincinnati Reds
78 77
Mike Green, electric carrot
Houston Astros
74 72
none
Pittsburgh Pirates
62 67
Brian




Los Angeles Dodgers
95 90
none
Colorado Rockies
92
76
none
San Francisco Giants
88 77
Ron
San Diego Padres
75 70
John Northey
Arizona Diamondbacks
70 86
none


By the way, if you want to re-visit your predictions, check out the original contest thread.

Some Analysis

AL East - Nothing terribly out of the ordinary happened in the AL East this year. The Yankees were 12 games better than we thought they would be, but at least we all knew they'd be a very good team. PECOTA actually saw this coming, pegging the Yanks for 99 wins, while the most optimistic (or pessimistic) Bauxite said 95. Props to Mudie and esteBENloaiza who predicted the AL East with the most accuracy - an average of 5 wins off.

AL Central - The top three teams were pretty easy to peg, but we all thought the Indians would be right there with the Twins, Tigers and White Sox, and boy were we wrong. The Indians won 65 games, 20 fewer than Da Box' collective prediction of 85. The closest Bauxite was seeyou, who was still 11 games too optimistic. seeyou did well for the whole division, projecting the Central with an accuracy of 6 wins/team, compared to a Box average of 8.4.

AL West - Boy, was this a wild one. On average, Bauxites missed AL West teams by almost 11 games, and there weren't even any individual teams that we came close on. (In fact, not a single Bauxite correctly predicted any single team in this division.) We were off by 9 games on the Angels, 11 on Texas, 7 on Seattle and 10 on Oakland. electric carrot had the best showing, as he was only a game off on his Seattle prediction, and 3 off on Oakland for a division average of 7.5.

NL East - Another division that was tough to predict, though it was mostly because of the Mets. Bauxites thought the Mets would take the division with 90 wins; they won 70, finishing 4th. Other than that we didn't do too badly. Props to esteBENloaiza for nailing the Phillies.

NL Central - the easiest division to predict in 2009; on average, we mis-fired by 5.6 wins. For the Cubs and Cards, the actual results were just about flipped from our predictions, and other than those 2 teams we were very close. Mad props to the roster's own John Northey, who foresaw this division with more precision than any Bauxite foresaw any division. John was wrong by no more than 6 wins for any team for an average deviation of 3.8 in the division.

NL West - we all knew the Dodgers would win the division, and we all knew the Padres would be bad, but other than that the West was a struggle (we weren't even particularly close on the Dodgers and Pads). Props to Ron for pegging the Giants when most Bauxites were 10+ games off.

And now, the results! The explanation of the scoring is at the top - the lower the better.

Name Score
electric carrot 7.33
nathanelias 7.47
Mudie 7.53
92-93 7.67
seeyou 7.90
parrot11 8.23
ANationalAcrobat 8.27
Mike Green 8.30
brent 8.43
Dave Rutt 8.53
John Northey 8.60
Brian 8.70
AWeb 8.73
Ron 8.73
mathesond 8.80
Nick Holmes 8.90
esteBENloaiza 8.93
PECOTA 9.03
VBF 9.47

Congratulations to electric carrot for taking home the crown! nathanelias takes the silver and Mudie rounds out the top 3. Further congratulations to Ron and esteBENloaiza, who each nailed three separate teams despite finishing in the lower half of the overall standings. Also: congrats to everyone except VBF for besting BP's fancy-schmancy projection system PECOTA! I guess PECOTA just had an off-year. Or Bauxites are just brilliant. Yeah, I like that better.

By the way, in case you were wondering (you weren't, but I was): Predicting 81 wins for all 30 teams would have resulted in a score of 9.53. You just squeaked in there, VBF!

Some more data for your consuming pleasure: each year, Vegaswatch looks back at predictions from projection systems and notable baseball writers. (Here is the write-up for the 2009 edition.) They use a different criteria: RMSE (root mean squared error). I believe RMSE rewards consistency a bit more, so making a wildly inaccurate prediction would have more of a negative impact on one's final score. I'll let the math majors debate the merits of each evaluation method, but for now let's just look at what the standings would like if we combined Da Box' contest with the predictions evaluated by Vegaswatch (which are bolded):

Name Score
nathanelias 9.04
seeyou 9.06
CAIRO 9.09
electric carrot 9.41
RLYW 9.44
Marcel 9.57
CHONE 9.63
Mike Green 9.64
92-93 9.64
O/U 9.65
Mudie 9.66
MGL 9.66
brent 9.75
THT 9.77
ANationalAcrobat 9.85
Brown 10.03
Dave Rutt 10.08
Sheehan 10.11
SI 10.15
Henson 10.21
ZiPS 10.25
Nick Holmes 10.26
Law 10.26
Brian 10.29
AWeb 10.44
mathesond 10.45
parrot11 10.47
Passan 10.50
Edes 10.83
John Northey 10.88
esteBENloaiza 11.18
PECOTA 11.42
VBF 11.49
Ron 11.91

Man, I gotta say, nice work guys. We did just as well as a bunch of "experts" and "projection systems". I'm sure ESPN will be coming knocking any day. They could use someone to fill the Gammons void, anyway - why not an entire online community of brilliant Jays fans???

If you're wondering why the scores/rank order are different, it's because this is calculated using the aforementioned RMSE.

9 comments



https://www.battersbox.ca/article.php?story=20091118224707817