[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

cross-check of (nearly) independent models

Posted By: Chuck Bower
Date: Thursday, 10 February 2011, at 12:35 a.m.

In Response To: let's make a deal -- OT (Chuck Bower)

It's nice to have an independent cross-check. Analogous to backgammon, there is a football backstep model similar to the construction of bearoff databases:

http://www.footballcommentary.com/twomindrillprobs.htm

Our method is a rollout. We have very little in common with the Krasker model, except some commonality in NFL data (i.e. historical results) which go into the front ends of the models.

Krasker says that (assuming equal, generic NFL opponents) with 2:00 remaining and 1 TO for the offense, a team is 7% worse off with the ball on their own 10-yardline (90 yards from the goalline), trailing by 3 than they are if on the 30-yardline trailing by 6.

Our model agrees with Krasker's in absolute GWC at 0.1% (10-yardline case) and within 0.25% GWC for the 30-yardline case.

Bottom line is that most of the equity for GB going for the TD isn't from their chances to score the TD but rather from leaving PIT with 90 yards to go when the TD attempt fails.

Messages In This Thread

 

Post Response

Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:

 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.