Enough data has been published previously on Denver and Indianapolis to do a direct comparison, but what do they look like historically?

Denver’s data set looks like this:

Denver Broncos 2005-2013
Year Team W L T SRS OSRS DSRS MOV SOS
2005 DEN 13 3 0 10.79 6.30 4.49 8.56 2.23
2006 DEN 9 7 0 1.32 -0.72 2.04 0.88 0.44
2007 DEN 7 9 0 -3.95 -1.57 -2.38 -5.56 1.61
2008 DEN 8 8 0 -5.79 1.15 -6.94 -4.88 -0.91
2009 DEN 8 8 0 0.32 -1.09 1.41 0.12 0.20
2010 DEN 4 12 0 -8.91 -0.54 -8.37 -7.94 -0.97
2011 DEN 8 8 0 -5.30 -2.87 -2.43 -5.06 -0.23
2012 DEN 13 3 0 10.10 7.08 3.02 12.00 -1.90
2013 DEN 6 0 0 13.95 21.22 -7.26 17.83 -3.88

 

And Indianapolis’s data set looks like this:

Indianapolis Colts 2005-2013
Year Team W L T SRS OSRS DSRS MOV SOS
2005 IND 14 2 0 10.80 6.82 3.98 12.00 -1.20
2006 IND 12 4 0 5.88 7.31 -1.43 4.19 1.69
2007 IND 13 3 0 12.01 6.44 5.57 11.75 0.26
2008 IND 12 4 0 6.49 1.58 4.91 4.94 1.55
2009 IND 14 2 0 5.93 3.65 2.28 6.81 -0.88
2010 IND 10 6 0 2.88 5.15 -2.27 2.94 -0.06
2011 IND 2 14 0 -11.28 -6.99 -4.29 -11.69 0.40
2012 IND 11 5 0 -4.71 -2.39 -2.32 -1.88 -2.84
2013 IND 4 2 0 8.89 1.72 7.18 8.33 0.56

 

Using SRS, you would say that Denver has a slight advantage. Let’s look at three different predictive techniques and what they say about point spread and odds of winning the game. (1) These three, for the Denver-Indianapolis game, yield very different results.

Odds of Denver Winning and Predicted Point Spread
Pythagorean Expectation Simple Ranking System Median Analysis
Pct Points Pct Points Pct Points
0.48 -0.6 0.57 2.1 0.77 9

 

The two techniques I trust more, Pythagoreans and SRS, yield different results for the winner but both say the game will be decided by less than three points. With games this close, small factors – a single turnover, a great punt return – can decide the results. I add the median prediction largely as a comparison. I don’t trust it as much as the other two methods in terms of predicting results.

All three predictions include a home field advantage effect.

Notes:

1. For a simple relationship between point spreads and winning percentages, look here. A different approach is given in the book “Mathletics“, worth reading if you’re into betting football.