Today, Nate Silver confessed that he screwed up.
Well, sort of.
While Silver admitted an error, he also tried to sweep it under the rug as a once in a lifetime event.
Nate Silver says How I Acted Like A Pundit And Screwed Up On Donald Trump.
The title is perfect. Perhaps Silver should have stopped there. He goes on to say …
I also wonder if there’s been too much #datajournalist self-flagellation. Trump is one of the most astonishing stories in American political history. If you really expected the Republican front-runner to be bragging about the size of his anatomy in a debate, or to be spending his first week as the presumptive nominee feuding with the Republican speaker of the House and embroiled in a controversy over a tweet about a taco salad, then more power to you.
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.
Please stop the self-congratulations. Nate, please tell us how the hell you put Indiana in the “win” column.
Silver was wrong on Indiana every step up the way with poll after poll after poll after poll projecting Cruz would win.
On April 30, “Silver’s Secret Sauce” projected Cruz had a 65% chance of winning Indiana. On May 1, I reported Silver Flushes Secret Sauce Down Toilet, Now Projects Trump has 69% Chance.
Silver puts that monstrous last-minute flip-flop in the win column. I call that pathetic.
Admission
There’s a lot more to the story, so I’m going to proceed in five sections:
1. Our early forecasts of Trump’s nomination chances weren’t based on a statistical model, which may have been most of the problem.
2. Trump’s nomination is just one event, and that makes it hard to judge the accuracy of a probabilistic forecast.
3. The historical evidence clearly suggested that Trump was an underdog, but the sample size probably wasn’t large enough to assign him quite so low a probability of winning.
4. Trump’s nomination is potentially a point in favor of “polls-only” as opposed to “fundamentals” models.
5. There’s a danger in hindsight bias, and in overcorrecting after an unexpected event such as Trump’s nomination.
Silver’s admission on point 1 is quote revealing. In effect he admitted blatant bias, overruling his own model.
In points 2, 3, and 5 Silver attempted to sweep his bias under the rug.
Point 4 is also an interesting admission. That is something I hammered Silver on for months on end.
Silver’s “Polls Plus” forecasts have been little more than biased bullsh*t.
The entire rest of his lengthy post was noting more than an attempt to sweep the mess under the rug.
Here’s the question I ask: Did Silver learn a damn thing?
Here’s a few things I know.
- Changing social attitudes matter.
- It was clear in December that voters did not care what silly things Trump said.
- Silver favored historical precedents over massive attitude changes. That was a big mistake, and one I warned Silver on in December.
Flashback December 10: Attitudes, Attitudes: Dear Nate Silver, Regarding Donald Trump, You Are Missing Something Big!
In January I wrote about Nate Silver’s Continual Underestimation of Donald Trump’s Chances.
Reflections on Being Wrong
I am not purposely attempting to brag. I may have been wrong.
I have been wrong on many things, most notably the prolonged effect QE had on the market.
I was totally wrong.
I now think Trump will win the election. I may be wrong about that. And I may change my mind thirty days from now.
But I won’t change my mind one day before the election with a proclamation that I got it right.
Mike “Mish” Shedlock
MISH, the humility to admit that you can be wrong is what makes your blog readable. The Silvers of this world believe their own PR which makes them part of the infomercial crowd.
I think I’m going to start myself a blog that predicts stuff. Actually, I’ll take other people’s predictions and reword them as my own. I’ll tell people what they want to hear. I’ll say great things about the things people already ‘know’ the answer too because it serve’s their self interest and vanity to believe it. Same thing about the stuff people hate. I’ll just tell them they’re right. Maybe I’ll even start a paid financial newsletter. My ‘secret sauce’ will be to use the words ‘would’ ‘could’ ‘might’ ‘possible’ strategically.
Nate needs to find another line of work.
Someone should be able to create a lobbyist job for him.
He is a lobbyist…… of the astroturf kind.
NYTimes preaches propaganda to an isolated parochial audience. Nate Silver is a reporter able to lick a pencil and stick it behind his ear. Nate was preaching to his choir, and it is not fair to pick on the mentally retarded.
Nate Silver sounds like a real schmuck
“While Silver admitted an error, he also tried to sweep it under the rug as a once in a lifetime event”.
Trying to protect his cottage industry as THE expert.
During the NCAA basketball season they (I think ESPN) kept using 538’s odds for various teams to win.
http://fivethirtyeight.com/sports/
Another mark against Silver was his bashing of Gallup following the 2012 election. Let your work do the talking … and others do the bashing.
Who really cares about Nate ? Why are you even giving him any attention?
Silver is a symptom of the much bigger issue of progressive states (obviously with decent success) indoctrinating their subjects, to believe in the existence of “experts”, that they should unquestioningly listen to.
“studies prove…”, “surveys show”, “Buffet says…”, “Goldman Sachs economists says…”, “The CEO of Google indicated…”, “statistically…… says experts.”
It’s all (withing comfortable margin of statistical uncertainty 🙂 ) just nonsense. Or, at least, no more nor less reliable means of discerning the accuracy of something than “Nostradamus said…”
And this is becoming ever more true, now that information flows freer and in greater volume. Back in the day, you could possibly beat the average punter at horse race prediction by having the phone number of some “guys” working around the stables. Now, those “guys” have blogs, so the odds are evened out again.
Similarly, if you ran a polling organization when noone else did, that could possibly give you an edge in election prediction. But once everyone has access to the same information, it’s all back to being random again. Whether one is election handicapping or stock picking or making predictions about centuries into the future.
Hey, how ’bout that Lacker, and multiple interest rate hikes in 2016? Or is that pertinent to the subject matter on a political campaigning blog?
It’s nice not having a dog in this fight, between blue political hack and red political hack,,,,but from a purely financial perspective, my money is on the most cunning, lying, connected and corrupt of the lot. And if this was an economic blog, I might suggest a little snippet or two such as, the last time Billery was in office, they left with gold in the mid to upper $200 range.
Then there was Clinton whitehouse OMB budget director Franklin Raines, who was handed the keys to FNMA. We saw where that led. Oh, not to forget who the Yellen fed is cheerleading for.
So, on the ‘odd’ chance that Vegas’ bookmakers favored pony wins, what might be the economic implications,,,,if we can take the time out from electioneering, that is?
There will be another hike this year, maybe as early as June according to the minutes, and that would qualify for multiple hikes.
The rise of “independents” in the 2016 and 2018 elections was forecasted by Armstrong in the late 80’s. Govt largess and debt turns the economy down, which causes the people to rebel against the establishment. All of the knee jerk and self-serving actions by govt only hurts the economy more. It’s happening all over the world for similar reasons. Why is it so difficult to see? I guess it’s the same reason people missed the impact of QE on markets, and will miss the rise in rates, the dollar, and stocks when the economy is so weak. It’s called being US-centric, instead of having a global view of capital flows.
“…US-centric, instead of having a global view of capital flows…”
This is the root of MANY analytical errors, here on this blog & elsewhere.
Americans will be the LAST group to notice/cause the destruction of the $USD – the rest of the world will impose that inevitable fate on the dollar long before 99.9% of americans have any idea what is happening. Having an exclusively US/Dollar-centric POV (to the exclusion of obviously global/multi-lateral economic mechanisms) is a serious handicap.
Amusingly you a “hate” Silver for being a stats nerd…when that’s what he’s 100% about. Sorry Mish, a stopped clock is right twice a day. Looking at Silver’s argument — it become almost impossible to get a statistically significant marker that would make Trump a sure thing. Again, his entire site is about statistical analysis.
His mea culpa was well reasoned, and generally stats have few obvious bias. If your opinion of Silver’s work is that bad, well don’t read the bastard!
Silver just got lucky in 2012 with Obama. Romney would have won if not for unions cheating.
I think Silver knows the “mistake” he made, but it is one he can’t admit since it will undo all his marketing to date- he was trying to shape the outcome. While not a pollster himself, he is just a kind of a push pollster. We have seen this before in his work- sudden revelations at the last moment to get it “right” before the hammer falls on a previously wrong prediction. He always explains this as a sudden, previously unnoticed shift in the electorate that he just happened to catch in time. Secret Sauce will remain secret.
“he was trying to shape the outcome”
Bingo. Him & the rest of the poll-quoting media… and they all failed miserably.
Dear Mish, Thes pundits all have full tiem jobs
The backlash against the establishment due to growing govt debt, that always turns down the economy, was forecasted by Armstrong decades ago, and his unparalleled model continues to be right. How can anyone listen to anybody that does not have the historical data to back up their claims?
To avoid having to admit another mistake, I strongly encourage you to reconsider your stance on rates. Whether the Fed raises rates to belatedly try to save pensions, or a misreading of the tea leaves as the minutes would indicate, or simply as a last chance before it becomes obvious that the rising cost of Socialism is taking its toll on the consumer, which the relative strength of retail stocks has been foretelling for over 6 months; make no mistake – rates are going up. And when the flows drive up asset prices further, causing the Fed to raise rates further to avoid the label of serial bubble blower, the dollar, rates, and stocks will go higher as dollar-based debts blow up balance sheets across the globe. When sovereign debt starts defaulting and the govt debt bubble officially pops, where will global capital run in anticipation?
Well, you are in for a surprise.
Outside of another (ceremonial) move in June or July there will be no more rate increases for quite a while. Any moves past that point will be seen as political (helping Trump) heading into the Election. Anyways, Rising rates will hurt housing / refinancing (wall street revenue source) … so dead end there.
Furthermore, yuan pegged to $US and China has been quite vocal about not raising rates to keep yuan weak(er).
Look for more of the same. Jawboning about moving rates higher at some future meeting … but no action.
The Fed – as always – is in full “reactionary” mode.
Real rates & global capital flows could very well “call the Fed’s bluff” and FORCE further reactionary moves from the Fed, whether viewed as political/unnecessary or not.
Again, the point here is that the real world (global, not US-centric) is DICTATING conditions TO the Fed, not the other way around. The Fed can only ‘bluff’ for so long… until it can’t. And current global capital flow trends may well be indicating that the Fed’s bluff is not & will not work anymore.
Buckle up.
Like many who conduct and analyze polls, Nate Silver injects his own personal bias into the wording of questions and analysis of the data.
Weak scientific method and weak statistics are why most polls fail. Nate has just been lucky to date and needs to enhance his model and let it lead rather than follow his views.
It is a way to common an error in polling and it finally caught up to him.
I see that the latest “predictions” on the Presidential race in a few counties that neighbor each other in Ohio and Pennsylvania are based on how they voted for Obama and Mitt.
Well, Obama is an excellent lying politician and has been able to fool most of the people most of the time. At least enough to win. Hillary is just a fair liar and a terrible politician except among black women. Trump doesn’t take crap off anybody, especially the Clinton’s or the media.
And, it’s not 2012.
Indiana, OH & W.PA will all end up being very solid Trump territory. Trump’s entire platform is geared to appeal to the mid-west/rustbelt, not unlike Reagan in ’80.
Pundits who still think OH is “in-play”/battleground are kidding themselves.
“All your Rustbelt are belong to Trump.”
Once the Dem’s and the military industrial complex’s opposition research gets really done on The Donald, it will be all over for him.
It will not be long before they will be calling Trump the Teflon candidate because nothing he says or does will affect how people feel about the state of this nation. Trump is code for “I am mad as hell and am not going to take it anymore.”
Could it be that Silver’s $ecret Sauce is some insider’s measure of behind the scenes corruption (Colorado & Georgia) or perhaps a “Diebold Index”?
Silver made a mistake once: He thought he was wrong.
Pingback: Nate Silver, Jeff Bezos, And The Art Of Failing Well – Investor Maven
I just referred to Silver as “some guy who got lucky once” in an email to a friend…went back and found the best headline on 538 from Oct 2016. You have to love it: “The Cubs Have A Smaller Chance Of Winning Than Trump Does.”