Thursday, October 8, 2015

WEEK FIVE NFL PREDICTIONS

Below please find the 2015 week five NFL predictions based on the model we built last year. This model uses a learning algorithm fed by data from just the last four weeks. To date, success rates this season are as follows: Week 2, 50%: Week 3, 62.5%: Week 4, 60%.

We have slightly altered the 2014 model in various ways to yield three sets of predictions. Model 1 (M1) is the procedure from last year and is based on normal averaging. So far, this has been the most accurate model. M2 is based on moving averages, and M3 is based on weighted percentages (thereby allowing games later in the season to count more than earlier games.)

Experience says to trust M1, but this week there is greater divergence between the models. So, it's tough to say. In any case, we will know soon enough.

So here goes - (The % indicated after each game represents the crowd-sourced prediction from nfl.com as of this afternoon. The two digit decimal numbers following team names should be interpreted relative to each other for comparison of the models and to determine, in part, how close the game will be.):

  • IND @ HOU - close game (Correct)
    • M1 - HOU (.46) over IND (.42) - Incorrect
    • M2 - IND (.48) over HOU (.43) - Correct
    • M3 - IND (.46) over HOU (.45) - Correct
      • Overall Prediction: IND over HOU
      • IND 27 over HOU 20 - Prediction Correct
      • Crowd Correct @ 69%
  • CHI @ KC - easy win for KC (Incorrect)
    • M1 - KC (.45) over CHI (.32) - Incorrect
    • M2 - KC (.42) over CHI (.35) - Incorrect
    • M3 - KC (.42) over CHI (.32) - Incorrect
      • Overall Prediction: KC over CHI
      • CHI 18 over KC 17 - Prediction incorrect
      • Crowd Incorrect @ 89%
  • SEA @ CIN - fierce game (Correct)
    • M1 CIN (.61) over SEA (.61) - Correct
    • M2 SEA (.64) over CIN (.61) - Incorrect
    • M3 SEA (.65) over CIN (.60) - Incorrect
      • Overall Prediction: SEA over CIN
      • CIN 27 over SEA 24 - Prediction incorrect
      • Crowd Correct @ 65%
  • WAS @ ATL - ???
    • M1 - ATL (.59) over WAS (.50) - Correct
    • M2 - ATL (.63) over WAS (.50) - Correct
    • M3 - ATL (.62) over WAS (.51) - Correct
      • Overall Prediction: ATL over WAS
      • ATL 25 over WAS 19 - Prediction Correct
      • Crowd Correct @ 96%
  • JAX @ TB - close game (Correct)
    • M1 - JAX (.39) over TB (.38) - Incorrect
    • M2 - JAX (.39) over TB (.37) - Incorrect
    • M3 - JAX (.39) over TB (.39) - Incorrect
      • Overall Prediction: JAX over TB
      • TB 38 over JAX 31 - Prediction Incorrect
      • Crowd Correct @ 55%
  • NO @ PHI - fierce game (Incorrect)
    • M1 - PHI (.47) over NO (.46) - Correct
    • M2 - NO (.50) over PHI (.48) - Incorrect
    • M3 - NO (.49) over PHI (.48) - Incorrect
      • Overall Prediction: NO over PHI
      • PHI 39 over NO 17 - Prediction Incorrect
      • Crowd Correct @ 66%
  • CLE @ BAL - close game (Correct)
    • M1 - BAL (.47) over CLE (.45) - Incorrect
    • M2 - BAL (.49) over CLE (.46) - Incorrect
    • M3 - BAL (.48) over CLE (.47) - Incorrect
      • Overall Prediction: BAL over CLE
      • CLE 33 over BAL 30 - Prediction Incorrect
      • Crowd Incorrect @ 86%
  • STL @ GB - easy win for GB (Correct)
    • M1 - GB (.65) over STL (.42) - Correct
    • M2 - GB (.72) over STL (.44) - Correct
    • M3 - GB (.69) over STL (.42) - Correct
      • Overall Prediction: GB over STL
      • GB 24 over STL 10 - Prediction Correct
      • Crowd Correct @ 95%
  • BUF @ TEN - close game (Correct)
    • M1 - BUF (.54) over TEN (.52) - Correct
    • M2 - TEN (.52) over BUF (.47) - Incorrect
    • M3 - BUF (.50) over TEN (.47) - Correct
      • Overall Prediction: BUF over TEN
      • BUF 14 over TEN 13 - Prediction Correct
      • Crowd Correct @ 83%
  • ARI @ DET - easy win for ARI (Correct)
    • M1 - ARI (.66) over DET (.40) - Correct
    • M2 - ARI (.62) over DET (.40) - Correct
    • M3 - ARI (.65) over DET (.39) - Correct
      • Overall Prediction: ARI over DET
      • ARI 42 over DET 17 - Prediction Correct
      • Crowd Correct @ 87%
  • NE @ DAL - easy win for NE (Correct)
    • M1 - NE (.63) over DAL (.52) - Correct
    • M2 - NE (.66) over DAL (.49) - Correct
    • M3 - NE (.64) over DAL (.50) - Correct
      • Overall Prediction: NE over DAL
      • NE 30 over DAL 6 - Prediction Correct
      • Crowd Correct @ 92%
  • DEN @ OAK - easy win for DEN (Incorrect)
    • M1 - DEN (.59) over OAK (.47) - Correct
    • M2 - DEN (.58) over OAK (.48) - Correct
    • M3 - DEN (.58) over OAK (.50) - Correct
      • Overall Prediction: DEN over OAK
      • DEN 16 over OAK 10 - Prediction Correct
      • Crowd Correct @ 88%
  • SF @ NYG - easy win for NYG (Incorrect)
    • M1 - NYG (.56) over SF (.36) - Correct
    • M2 - NYG (.62) over SF (.25) - Correct
    • M3 - NYG (.60) over SF (.24) - Correct
      • Overall Prediction: NYG over SF
      • NYG 30 over SF 27 - Prediction Correct
      • Crowd Correct @ 87%
  • PIT @ SD - ???
    • M1 - PIT (.57) over SD (.46) - Correct
    • M2 - PIT (.54) over SD (.47) - Correct
    • M3 - PIT (.57) over SD (.45) - Correct
      • Overall Prediction: PIT over SD
      • PIT 24 over SD 20 - Prediction Correct
      • Crowd Incorrect @ 61%

Games to watch: SEA @ CIN, NO @ PHI, and then IND @ HOU, JAX @ TB, CLE @ BAL, BUF @ TEN. The overall predictions pretty much match my intuitions at this point, though, as I noted above, limited experience says to trust M1. 

71% of the overall predictions agree with the crowd. In this light, PIT @ SD is also a game to watch. This should be an interesting weekend for NFL football.

No comments:

Post a Comment