Podcast Episode 3: Chase Stuart

Chase Stuart, founder of Football Perspective and a regular contributor to the New York Times, calls in to the show from his home in New York City. Dave and Chase begin by discussing two point conversions, and how expected value and variance come into play when coaches make the call to go for two. Chase goes on to discuss his own metric for looking at in-game point differential, called Game Scripts. He explains how Game Scripts can be used in conjunction with run/pass ratio to contextualize a team's general tendencies over the course of a season. The episode ends with a check-in on season win projections for the five remaining undefeated teams, along with some well-deserved gushing about Peyton Manning's unbelievable season.

Make sure to subscribe on iTunes so you don't miss an episode!

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

9 Responses to “Podcast Episode 3: Chase Stuart”

  1. NateTG says:

    The notion of using 'game script' as a control in analysis of play calling seems quite powerful. I'm quite dubious about how good it is at describing relative team strength.

    > ... if you score three touchdowns early, and then the other
    > teams scores three touchdowns in the third quarter, it's 21-21
    > entering the fourth quarter, a lot of people say that that's an even
    > game, but the game scripts...

    That seems to say more about the game being random than about one team being better than the other. If two evenly matched teams trade touchdowns, the one that scores first will have an average lead of +3.5, throw in an early pick six, and that average goes to 10.5.






  2. Brian Burke says:

    It provides a much better run/pass ratio that accounts for score difference. It reminds me of research Carson did on the same concept a couple years ago here. Instead of game script, we used WP-adjusted run/pass ratio.

  3. Chase says:

    Yes, I agree that while it is highly correlated with team strength, I'm not all that comfortable using it to measure team strength.

  4. Anonymous says:

    This was a great podcast.
    I love the Game Script concept and can't wait to see where it goes.

    Thank you Chase and Dave and Brian.

  5. Unknown says:

    I agree the podcast was great and the Game Script idea is very cool. I like the idea of being able to tell, in a general but still very "real" way, which team dominated the majority of a game, by looking at a single number (the Game Script) and then comparing that number to the final score. As Chase pointed out in the podcast, it's a fairly simple idea, but as with a lot of math and statistics, there is an eloquence in simple ideas. Great job guys!

  6. NateTG says:

    > Yes, I agree that while it is highly correlated with team strength, I'm not > all that comfortable using it to measure team strength.

    The average lead has to converge on team strength over time. It's really a question of how fast that happens.

    It's a little circular and retrospective, but if I look at teams per-season Game Script, it's comparable to the Vegas spread as a predictor of margin of victory.

  7. Joshua Perry says:

    I have a request, that would hopefully help grow your podcast audience. Add a podcast section to the top menu bar of your website. It would help the guests of the show with their self-promotion which they seem to be lacking. Also, I'd love to hear from the Pro Football Focus guys(Sam Monson @pff_sam ). It seems their analysis is filmed-based and I'd like to get some insight into their grading system.

  8. Brian Burke says:

    Joshua-Already in work!

  9. Anonymous says:

    The mp3 link is:

    http://traffic.libsyn.com/advancednflstats/003_AdvNFLStats_ChaseStuart.mp3

    Is there any way of making it easily accessible?

Leave a Reply

Note: Only a member of this blog may post a comment.