Dan Lewis

Barone v. Silver

Political analyst Michael Barone has a column out predicting that Governor Mitt Romney will defeat President Barack Obama rather handily. Meanwhile, Nate Silver at FiveThirtyEight, as of this writing, has Obama as better than a 4:1 favorite. Only one of them can be correct, of course, and that’s nothing new. But Barone calls the winner of 15 individual states — most analysts don’t bother to do that.

Let’s turn Barone’s picks into wagers, instead, with the odds set by Silver. Or, for sake of clarity, what if Silver sold $1,000 payout wagers on the outcomes of each of the 15 states Barone listed, with the price of each such wager being $10 per percentage point that Silver’s model predicts? For example, let’s say Barone wanted to bet that Romney would win Indiana. (That’s one of his 15 picks.) Silver says that Romney is a 99.8% chance to win it, so Barone pays $998. If Romney wins Indiana, he’d get $1,000; otherwise, Silver would keep Barone’s money.

Barone made the following picks. In the parentheses after each pick is Silver’s odds, at $1 per .1%, that the person Barone chose will win the state. That’s how much it’d cost Barone to make that $1,000 payout wager.

Barone’s picks, Romney to win:

* Indiana ($998)

* North Carolina ($796)

* Florida ($548)

* Ohio ($162)

* Virginia ($330)

* Colorado ($321)

* Iowa ($193)

* New Hampshire ($196)

* Pennsylvania ($34)

* Wisconsin ($56)

In total, Barone is paying $3,634 for a combined chance at winning $10,000. Most of that upside comes from longshot bets (according to Silver, at least) in Wisconsin, Pennsylvania, Ohio, Iowa, and New Hampshire. But that’s how odds work.

Barone’s picks, Obama to win:

* Minnesota ($990)

* Oregon ($987)

* New Mexico ($989)

* New Jersey ($998)

* Michigan ($986)

Not a lot of upside here, but also, not a lot of risk. (Again, that’s how odds should work.) Barone’s risking $4,950 to win $5,000, which means he can make $50 here. Consider it a hedge against his Wisconsin wager, I guess.

I’ll revisit this post on Tuesday/Wednesday to see how well Barone did against Silver’s odds. It may not be worth it to bother with his Obama-to-win bets (or, for that matter, Indiana) as there’s no meaningful disagreement between the two, and a miss by both would skew the numbers dramatically.