Discussion for article #228698
Silver was able to call the 2008 and 2012 presidential elections with almost perfect precision, because there was a lot of good polling data from the swing states.
That’s not going to work in a congressional election. There is not a lot of good polling in the Alaska senate race. I’m getting five calls a day from campaigns and “surveys.”
Most of the time, I just hang up on them.
To a certain extent, Milbank and other critics of Silver are correct. The data that Silver and other statisticians create is not absolute. This does not mean that they are necessarily wrong, but they create a narrative that can be wrong or moderately wrong.
5 Likes
I don’t think Milbank is totally wrong. The truth is that Silver and Wang’s feats in 2008 and 2012 were less impressive than their boosters and they themselves would have it. Sure, 49/50 states called correctly sounds fantastic, but at least 40 of those were not in doubt. In fact, in both cases, only a few states were really in doubt (NC, IN and MO in 2008; FL in 2012).
As for Moneyball, very nice, but the Oakland As won 6 pennants from 1972-1990 and 0 in the Billy Bean era.
2 Likes
Um, why should anyone care what Dana Millbank thinks? Isn’t he just a typical beltway media pundit? Someone help me out here.
3 Likes
Milbank better watch himself. His boss is Jeff Bezos, who doesn’t make a trip to the bathroom before first consulting a statistical model. The Washington Post is, and likely will remain, the home to all manner of right-wing foolish opinions. But with the new boss there the quickest way to get fired there will be to question math.
I’m not kidding about Milbank needing to watch himself. If Milbank keeps arguing that the future can’t be predicted with math I can confidently make another prediction: he won’t last the year at the Post. I worked for a few years at Amazon where I had many meetings with Bezos. Those who don’t know their businesses down to the 100th decimal point don’t last long with him.
He’s projecting and he doesn’t even realize it.
It’s the “gut feeling” talking heads who make assertions with 100% certainty.
Nate Silver et. al are the first to warn that their models are imperfect. I’m actually feeling kind of embarrassed for Milbank, as he’s confirming that he has never read Silver’s thoughts on the matter.
3 Likes
It is not surprising that Dana is attacking “Big Data.” Nate Silver is a threat to Milbank and all political pundits who make a living pulling predictions out of their asses between Washington cocktail parties.
3 Likes
Big Data means something other than the traditional little data provided by limited poll sampling and subsequent application of different statistical tests to account for the inevitable levels of error/uncertainty.
1 Like
Pundit bobblehead rages at his creeping obsolescence. Story at 11.
3 Likes
Got a couple of calls from “JBER” so picked up on one thinking it was from the Joint Base Elemendorf/Richardson hospital where I had had a recent MRI. Turned out to be a comically loaded series of questions vilifying Senator Begich and building up Ohio/Maryland Dan Sullivan his opponent. I outright started laughing at the poor young man who had been hired( wouldn’t tell me who was paying him but I wished him more gainful employment in the future) to do the “poll”. Every answer I gave was of course the opposite of what they wanted to hear because if anything it turned me off to Sullivan even more. Blowback is a bitch.
3 Likes
The most accurate poll aggregator in the 2012 Senate elections was Sam Wong who called every Senate race correctly. Silver missed on two of the races. Silver who has access to the likely voter models yeaterday discovered that this year Fox News was using the 2010 election turnout in their likely voter polls. If you used 2008 as your likely voter model the Democrats would win back Congress.
3 Likes
A lot of it really is how you parse the data. The problem is that they cannot actually account for all of the variables, and that means that the data is, at best, only going to be accurate only most of the time rather than all of the time.
1 Like
Until Silver is wrong, I’m sticking with Nate. Just because his models predict things that I don’t like, doesn’t mean he’s wrong. It’s embarrassing to see democrats bitching about Nate Silver for delivering the message. It’s twice as embarrassing that I clicked on a article about Dana Milbank’s opinion.
4 Likes
Milbank is just displaying his ignorance of Measurements, Statistics and Probabilities. He assumes certainty and precision, but anyone having even the most basic training in sciences has learned that all measurements have uncertainties that affect the precision of the outcome (hence the plus/minus citations attached to the end results, for starters). For those further interested, start with looking up “significant digits.”
As OccamsRazor2 pointed out, the term Big Data is misused in the column.
Despite that, Milbank does have a point that many forecasters are using silly levels of precision.
This column is hardly a rant against Silver, Wang, et. al. To me, it reads like a column written by someone who recognizes the value of the “quant” models but desperately wants the BubbleHeads (like himself) to maintain some semblance of relevance.
1 Like
“In other words, I’m inclined to believe the models are right:”
Dana can “believe” whatever he likes but Nate is using a system model. When math is performed by the system you will get fractional amounts unless you choose to throw them away.
1 Like
But, Silver and Wang aren’t really doing rigorous statistics (Wang probably does in his real job as a scientist, but he doesn’t in his political predictions). Take the example of “a 60% chance of the Senate going Republican”. From a statistical standpoint, that is meaningless without confidence limits. If the 95% confidence limits are 55-65%, then OK, that would mean the GOP are slight favorites. But if they are 45-75% (which is probably closer to what they really are) then, statistically speaking, it’s a dead heat.
As I’m sure Wang (and probably Silver too) knows, if you have a clinical trial that shows a drug did 20% better than placebo, but the confidence limits are wide enough to include 0, the FDA will demand additional trials.
3 Likes
Considering Silver’s data represents the uncertainty in these races, not the certainty. I’d just default back to my normal position that Milbank is a moron.
1 Like
I liked it better when Silver was predicting an Obama win rather than the real possibility that the Republicans take back the Senate, but until he is proven wrong, Silver is (more than) entitled to the benefit of the doubt. Milbanks claiming he isn’t a political sabermetrics denier is a little like proclaiming you aren’t racist immediately before adding “but”.