Earlier this summer I had put together an enormous database (9.51MB) with the last two years stats. I have always felt that it should be possible to set-up a range where we could make a reasonable assumption that the score should land. I attempted to calculate this earlier based on a combination of setting up ranges for OE, DE and Pace this somewhat worked but revealed very few games that would meet my requirements. However my latest attempt where I calculated the ranges using a different method has been enormously successful using the 2007 stats. I ran a real time simulation from Jan 25 to the season conclusion and revealed the following results for 1498 games. Here are the results for those games that where outside of my calculated range:
#/Rng W L Pct.
>0.0 287 145 .664
>0.5 244 114 .682
>1.0 212 94 .693
>1.5 180 71 .717
>2.0 148 52 .740
>2.5 114 41 .735
>3.0 90 30 .750
>3.5 69 20 .775
>4.0 56 14 .800
>4.5 43 7 .860
>5.0 35 6 .854
>5.5 27 4 .871
>6.0 22 2 .917
I’m going to try it using the 2006 schedule and see how this holds up next.