I don't know enough about the specifics of his methodologies to comment intelligently on their relative complexity. I would say, however, that, assuming the same data used in Silver's models are available to any similarly quantitatively inclined election/political junky (certainly a non-trivial assumption), it's telling how consistently well Silver's models perform compared to his competitors. Who else does as well as he does (I'm not asking rhetorically; I don't know the pop pundit/modeler landscape well enough to know)?

I definitely agree with your last statement.

My point is, to act like your somehow more plugged into the election cycle because you can regurgitate Silver ad nauseam is really stupid.

Every election junky I know has been following Silver since his 2008 predictions.

What are the best ways to be plugged into any given election cycle?

The Burtryenolds acount

This just doesn't seem like the right was of looking at it. "Lean" and "Likely" mean it could go the other way, which is what happened. They don't make predictions, and they repeat daily that nothing is certain. It's just not a right/wrong thing.



Because his work is good.

i'm gonna crack open this post's west-most window and let its cool breeze dance across my collarbone dimple

I never really said it wasn't. I just place less value on it than some of people on RM.

It can't be denied that he looks like shit.

And to sum up my problems with his analysis in 2016. He (and to be fair, many/most people) kept saying "History shows us! History shows us!" when it was clear Trump's candidacy had been defying history for months.

I agree with you 100% regarding his 2016 primary "analysis" which was objectively bad; I disagree that he did that in the general.

Edit: or to the extent that he did so in the general it didn't negatively influence his model, which I think held up pretty well all things considered in such a unique election cycle.

Hey dude, I just wanted to take a moment to say I really like your style.

We all love a Florida recount, right?


Edit: you’d probably love for me to put this in the right thread, too, but disappointments abound.

Some weird stuff going on; in Broward county, it appears that 25,000 people voted for someone for Senate but not for governor.

Has there ever been a recount that flipped the result on a margin that large before?

Sorry, that should be the other way around. Looks like 25,000 people voted for governor but not for Senate.

It's about 30,000 votes now. Nelson campaign wants them hand-counted in case there was a machine error. Others suggesting it may have been poor ballot design.

This should probably be in the midterms or the 2000 election threads.

Fuck that Kemp guy in Georgia. If they can whittle his lead down just a TINY bit more, he'll have to have a run off.