I tend to think this is a bit overrated. Obama didn't do very well in either Arizona or Georgia this year, for instance, which were two of the better theoretical candidates to shift toward the center. Also, Democrats have ~350 electoral votes in play already between the competitive states + the blue ones, and don't…

Well, there's always the chance that you've specified the model incorrectly. But I think about these things really, really carefully, and probably err on the side of making somewhat conservative choices about the degree of uncertainty that we allow for.

I like the show called The National Football League. 95% of my TV viewing is sports, or HBO.

Only a wizard could have predicted that Jeffrey Loria was a scam artist.

I think it's important for reporters to be better versed in math, statistics, logic, economics, etc. A lot of this stuff isn't all that complicated, so a little bit might go a long way. If I were running a gigantic news organization, I'd mandate that my reports take a few math and stats courses.

Bayes, duh.

Gallup has something of a point, in that if everybody just aggregates everybody else, you no longer have as much original content. And you may get groupthink and herding. You can draw analogies here between polling and the media landscape more broadly, I suppose.

In the book, I talk to Bob Voulgaris, who is maybe the best sports bettor in the world. Super sharp. Works his butt off. He gets 57% of his NBA bets right, not more than that. The luck component is pretty large. I think we add value mostly by comparison because the pundits are the equivalent of the 1899 Cleveland…