Game probabilities for week 16 are up at the New York Times' Fifth Down.
This week I highlight one the least interesting match-ups on a schedule packed with interesting games.
Weekly Game Probabilities - Week 16
published on 12/20/2012
Subscribe to:
Post Comments (Atom)
Here's an update on how the probabilities are performing this season. I am tracking the accuracy of the model's predictions compared to the predictions derived from the averages of the computer ratings tracked by ThePredictionTracker (www.thepredictiontracker.com/blog). Previous results were posted in the comments of these links:
http://www.advancednflstats.com/2011/12/weekly-game-probabilities_29.html
http://www.advancednflstats.com/2012/11/weekly-game-probabilities_15.html
Over the past two seasons, the model is slightly worse at predicting game outcomes than an average computer model. The issue of home field adjustment, as noted previously, and as Brian has acknowledged, is even more glaring this season. It may be that one could use the generic win probabilities of the efficiency rankings and apply a better home field adjustment to get better than average predictions.
The games where Brian's model predicted a different outcome than ThePredictionTracker averages are listed below with Brian's model's picks marked with an asterisk.
Week 4: NE at Buf* (L), Min* at Det (W), Cin at Jac (L)*
Week 5: Phi* at Pit (L), Sea at Car* (L), Den* at NE (L)
Week 6: Pit at Ten* (W), Cin at Cle* (W), Den* at SD (W)
Week 7: GB at StL* (L), Dal at Car* (L)
Week 8: SD at Cle* (W), Sea at Det* (W), NE vs StL* at London (L), Mia* at NYJ (W), Atl at Phi* (L), NYG at Dal* (L)
Week 9: Bal at Cle* (L), Car* at Was (W), TB at Oak* (L), Phi* at NO (L)
Week 10: Hou* at Chi (W)
Week 11: GB at Det* (L), TB at Car* (L), NO at Oak* (L)
Week 12: NE at NYJ* (L), Bal at SD* (L), StL* at Ari (W), GB at NYG* (W)
Week 13: Sea* at Chi (W), Cle at Oak* (L), Cin* at SD (W), NYG at Was* (W)
Week 14: Atl at Car* (W)
Week 15: Car* at SD (W), SF* at NE (W)
2012: Home 7-15, Away 10-3, Neutral 0-1, Overall 17-19
Since 2011: Home 16-29, Away 16-6, Neutral 0-2, Overall 32-38
I was not familiar with thepredictiontracker. That's pretty awesome.
AES - Could you possibly update your spreadsheet to include NNLL? I've posted this before, but it's just log(probability of event) / log(.5). So when prediction tracker says Green Bay had a 42% chance to win last week, and they won, that's ln(.42)/ln(.5) or ~1.24. These numbers can be averaged across a time period (they're linear), with lower being better and 1 being "everyone always has a 50% chance to win". This let's you differentiate between being right/wrong on a 51-49 game and a 70-30 game, etc.
I may run these numbers myself later -- that site is really easy to pull data from.
Great post. I was checking continuously this blog and I'm
inspired! Very useful information specifically the
last part :) I handle such
information a lot. I was looking for this certain information
for a long time. Thanks and good luck.
Feel free to surf my blog ; darkyrom.com
thanks AES.
As a lion fan (ouch) I've noticed that the lions have been having pretty decent chances to win a lot of their games, similar to this weeks 50/50 against atlanta.
If I get around to it, it would be interesting to see how likely every teams season has been.