Keith Goldner, chief analyst at Numberfire and a regular contributor to Advanced NFL Stats, joins Dave to discuss his recent research. Keith begins by outlining his Markov model for football, explaining how to use historical data to calculate the probability of an offensive drive ending in a particular "end state". The two then discuss the concept of "net expected points", and compare and contrast its value as a tool in combination with other efficiency stats. The show ends with a detailed analysis of two fourth down decisions from last week's games, highlighting the frustrating difference between good process and good results.
For more of Keith's work, and to check out some of the visualizations and data sets discussed in the episode, visit his site: Drive By Football
Make sure to never miss an episode of the Advanced NFL Stats Podcast by subscribing on itunes.
Podcast Episode 2: Keith Goldner
published on 9/26/2013
Subscribe to:
Post Comments (Atom)
Great, I am really enjoying these podcasts. Thanks for all the good work, keep it up.
Cheers,
J