Independent writer, editor, consultant. Former senior director of research quality at Mackinac Center, analyst NBC/WSJ survey, editorial board at Detroit News.
@lindsay_rhodes@SumerSports This is great, but the Purdy equation probably has to include the Shanahan-McCaffrey variable. It has a sizable negative coefficient.
Just kidding. This is great.
@lindsay_rhodes So that was helpful. It seems EPA per play functions somewhat like slugging percentage and success rate functions somewhat like batting average (ignoring the difference between team stats and individual stats). I’ve heard you define both, but the contrast didn’t quite sink in.
@lindsay_rhodes@SumerSports Is this arguably a coaching stat as well? And has Sumer worked up coaching stats? Aside from play-calling (which might be at work here), one could also imagine coaching stats on penalties and players doing the “right thing” (getting out-of-bounds, eating up the clock, etc.).
@lindsay_rhodes@SumerSports Great thread; great podcast. But is the finding counterintuitive? There’s a robust debate near the playoffs over “play them”/“rest them” and the value of a first-round bye. Our gut leads us both ways. And isn’t the rest between practice and game time about the same for each team?
@lindsay_rhodes O-line, D-line, because—with apologies to George Mallory—they’re there. But that hardly matters. Thank you for this exchange:
“I'm going to game theory this a little bit and take Ashton Jeanty as my flex.”
“That's messed up, dude. That's messed up.”
😆😆😆. Evil genius.