Just as a minor note, your vote also has very little value if the great majority in your state disagree with you. When I lived in California I could be sure its electoral votes were going Democratic; my voting Republican had no chance of changing the outcome.

This is great! Is there a way to estimate probability that a vote is pivotal by state from the relative voter power index? Nate co-authored this paper that estimated the likelihood that given vote would flip the election by state, and found that some states had a > 1 in 10 million chance. http://www.stat.columbia.edu/~gelman/research/published/probdecisive2.pdf

Under your $25 billion value of winning assumption, that places the value of a vote at $2500 in those states. I think that is still probably an underestimate in swing states, as I think the difference in outcomes is probably worth more than that most of the time.

> The big, obvious factor in determining the value of a vote — more specifically, the chance of a vote determining an election — is how close the election is in a state. But another big factor is the correlations between how different states tend to vote, as well as discreet number combinations.

Isn't the population of that state also highly relevant? The fewer votes there are, the higher the chance that yours is the deciding one.

> It shows that, for example, Republicans are spending 72% more in North Carolina, compared with their other spending, than Silver’s model would advise.

Just as a minor note, your vote also has very little value if the great majority in your state disagree with you. When I lived in California I could be sure its electoral votes were going Democratic; my voting Republican had no chance of changing the outcome.

Yes, good point — either agree or disagree.

This is great! Is there a way to estimate probability that a vote is pivotal by state from the relative voter power index? Nate co-authored this paper that estimated the likelihood that given vote would flip the election by state, and found that some states had a > 1 in 10 million chance. http://www.stat.columbia.edu/~gelman/research/published/probdecisive2.pdf

Under your $25 billion value of winning assumption, that places the value of a vote at $2500 in those states. I think that is still probably an underestimate in swing states, as I think the difference in outcomes is probably worth more than that most of the time.

Interesting! Agreed. I will look into whether it’s possible to derive absolute probabilities from that.

edited Aug 18> The big, obvious factor in determining the value of a vote — more specifically, the chance of a vote determining an election — is how close the election is in a state. But another big factor is the correlations between how different states tend to vote, as well as discreet number combinations.

Isn't the population of that state also highly relevant? The fewer votes there are, the higher the chance that yours is the deciding one.

Yes, but it's basically proportional to electoral votes, so population essentially cancels out with that.

A great takedown of a very sad situation.

> It shows that, for example, Republicans are spending 72% more in North Carolina, compared with their other spending, than Silver’s model would advise.

Judging by table, should be Georgia!

Yes, my mistake. Now fixed!

Yep. Profits from division and unrest.

Who ends up with all the money spent?

TV and radio stations, websites, campaign operators