How I Acted Like A Pundit And Screwed Up On Donald Trump (Nate Silver)

[oberliner]

Since Donald Trump effectively wrapped up the Republican nomination this month, I’ve seen a lot of critical self-assessments from empirically minded journalists — FiveThirtyEight included, twice over — about what they got wrong on Trump. This instinct to be accountable for one’s predictions is good since the conceit of “data journalism,” at least as I see it, is to apply the scientific method to the news. That means observing the world, formulating hypotheses about it, and making those hypotheses falsifiable. (Falsifiability is one of the big reasons we make predictions.1) When those hypotheses fail, you should re-evaluate the evidence before moving on to the next subject. The distinguishing feature of the scientific method is not that it always gets the answer right, but that it fails forward by learning from its mistakes.

http://fivethirtyeight.com/features/how-i-acted-like-a-pundit-and-screwed-up-on-donald-trump/?ex_cid=538twitter

Really detailed self-analysis of Five Thirty Eight and Trump by Nate Silver.

[CobaltBlue]

1. He’s no longer a wonder…

…2016 has been a screw-up year for Nate Silver.

[sharp_stick]

4. He's still the best there is

Just because he's not predicting the candidate you may want as well as you'd like isn't enough to call it a screw up year.

The FiveThirtyEight “polls-only” model has correctly predicted the winner in 52 of 57 (91 percent) primaries and caucuses so far in 2016, and our related “polls-plus” model has gone 51-for-57 (89 percent). Furthermore, the forecasts have been well-calibrated, meaning that upsets have occurred about as often as they’re supposed to but not more often.

If there is anyone that's got a better record than this I sure haven't seen it.

[fasttense]

2. I use to be one of his biggest fans

But I don't trust him anymore. I think he plays pundit a whole lot more than he lets on.

[oberliner]

5. At least he is willing to do some reflection

Not every writer is willing to be so thoughtful about such matters.

[Agnosticsherbet]

3. Excellent article.

And I understood most of it.
Good writer.

[joshcryer]

6. Kudos to Nate.

I would expect nothing less. Even the conservative forecasters admit it when they botch it.

edit: and nice of him to admit the arbitrary polls plus was less accurate than his old model.

[Angel Martin]

7. Interesting article

my interest in modelling and forecasting comes from financial markets.

One of the conclusions I have come to is that wrt black swans (improbable but consequential events) the experts are basically useless.

They are using past data and trying to find a model that minimizes forecast error.

one of the things you quickly see in financial markets forecasting is that models that are even capable of forecasting extreme events like black swans, are useless in forecasting "normal" times. I'm sure the same is true for election models.

A good example from financial markets is current interest rates. Interest rates are at a 5000 year low. Negative interest rates are now widespread in europe and japan.

No-one forecast negative rates, of course. And economic forecasters are simply ignoring negative rates, since they have no precedent, so no way to quantify their impact.

I don't know what kind of economy follows from negative rates.

But I bet that when we find out, we are not going to like the answer.

[frylock]

8. cough*hack*cough