Here's the other thing. This election is a sample size of one. Everyone is trying to fit these very binary long-term narratives around it.

"Betting markets are bad at pricing elections."
"You can't look at polls, they're useless."
"Nate Silver is an idiot and/or the devil."

1/7

I am not a math guy. I will send out the bat signal to @PlusEVAnalytics for the Bayesian perspective of how this election should affect our priors about elections in general.

But qualitatively you can't use a sample size of one to make these sweeping generalizations!

2/7

I used to trade options for a living. Path dependency is very non-intuitive and hard to get used to. You would trade these options that started with a 67% chance of finishing in the money and halfway through the cycle they would be 10% to finish in the money and then...

3/7

...you come in one day and they're suddenly 99% to finish in the money because some event happened in the world (cc: @strassa2). And you're sitting there trading these things, trying to hedge, and pulling your hair out. But the point is: *it really was 67% at the start*

4/7

And that's my takeaway from all of this. I believe Joe Biden was around a 2/1 favorite to win at the outset, and he's going to win. The path we took to get there was unexpected to some, but that doesn't imply that the tools we had at the outset are flawed.

5/7

Betting markets are not perfect, but they are the best predictor we have of the outcome of an election like this.

If you disagree, bet into them and get rich (if that option is available to you legally).

If you disagree, build a better mousetrap. We're all eager to see it!

6/7

But these black-and-white, sweeping generalizations drawn from a sample size of one that criticize betting markets are really misguided.

Ok, fin.

7/7