Support NeoGAF

[brianmcdoogle]

So, we're at the end of all things, and I thought I would give my final thoughts on this thread.

I didn't create it to chicken little "why is Nate not giving me the numbers I want for Hillary", but rather start a discussion on what was up with his erratic model this entire season, and to illuminate to people who don't follow these things on what the other experts in the field thought of Silver's model (which is to say, skeptical this time about). And, of course, to highlight his absolutely stupid and terrible "punditry".

The thing is, even if Silver is dead on tomorrow and Wednesday as final numbers come out, it doesn't discredit nor confirm the big issues we and other people had with is model. If he's completely wrong, it will still be worth asking what throughout the campaign caused the final numbers to be so far off. And, if it is 100% correct in the final prediction, that doesn't excuse the problems his model has, the commentary experts have had on it, and the wild fluctuations and supposed dogma Silver prescribes to build his model.

To use my crude analogy I posted in PoliGAF: If a flight leaves on time and arrives on time with the plane and passengers in one piece, but along the way the pilot gets drunk and does barrel rolls, it's probably fair to ask questions about the flight, especially if other flights and pilots get the plane their without all the drama.

And, as just a last bit of snark to Punditry Nate: he tends to think pollsters are herding their results to one another at the end, yet when his model moves towards all the other model... ?

PEC's explanation on their model, their prediction percentage, and a bit last thoughts on Silver's model:

http://election.princeton.edu/2016/11/06/is-99-a-reasonable-probability/

Now think about the FiveThirtyEight approach. I don’t want to get into too much detail. Although they discuss their model a lot, to my knowledge they have not revealed the dozens of parameters that go into the model, nor have they released their code. Even if they did, it is easy to make errors in evaluating someone else’s model. Recall Nate Silver’s errors in his attempted critique of PEC in 2014. So let me just make a few general comments. I am open to correction.

Their roots are in detail-oriented activities such as fantasy baseball. They score individual pollsters, and they want to predict things like individual-state vote shares. Achieving these goals requires building a model with lots of parameters, and running regressions and other statistical procedures to estimate those parameters. However, every parameter has an uncertainty attached to it. When all those parameters get put together to estimate the overall outcome, the resulting total is highly uncertain.

For this reason, the Huffington Post claim that FiveThirtyEight is biased toward Trump is probably wrong. It’s not that they like Trump – it’s that they are biased away from the frontrunner, whoever that is at any given moment. And this year, the frontrunner happens to be Hillary Clinton.

And then there is the question of why the FiveThirtyEight forecast has been so volatile. This may have to do with their use of national polls to compensate for the slowness of state polls to arrive. Because state opinion only correlates partially with national opinion, there is a risk of overcorrection. Think of it as oversteering a boat or a car.

Lots more at the link.

 

[soco]

Unlike Monday-morning QBing a term in office, where it is impossible to say who would have handled an event best, we get to see on Wednesday morning, if not Tuesday night, whose forecasts were best.

The problem is that they're forecasting with probabilities, so it's not like they're saying "this is definitely going to happen", and thus can't easily be "wrong". They're just saying it's more or less likely for X and Y to happen.

A single point won't prove any of them are better. The only way to compare these models is to compare after many elections, and to compare the generalized versions (as they'll likely tweak the specific models).

 

[Rentahamster]

To use my crude analogy I posted in PoliGAF: If a flight leaves on time and arrives on time with the plane and passengers in one piece, but along the way the pilot gets drunk and does barrel rolls, it's probably fair to ask questions about the flight, especially if other flights and pilots get the plane their without all the drama.

Maybe the pilot was fine, and it's the uninformed passengers who get scared out of their minds and imagine that a little turbulence was barrel rolls.