538’s climate science data problem compounded by legal threats
Paul Krugman has been throwing elbows at 538-ESPN lately because of a measurable decline in the quality of content from what we’re used to seeing from Nate Silver when he worked at the New York Times, and before that, writing for himself on Blogspot.
The first response I saw to that criticism was supposed to be a sort of self-parody as an apology, I guess. To me it comes off as a petulant act that he’d never have gotten away with while at the Times, and an example of where he’s gone off the rail.
Going through the 300 some comments, a lot of Nate’s/538’s defenders are outright climate science deniers making statements like these:
My gut tells me that these folks are only defending 538 now because its most significant mistake thus far: hiring people with questionable views who are going to write stories advocating and defending those questionable views backed by equally questionable data and methodology. It can’t be any other way than that, one leads to the other. If you let your views lead the data then your methods are unsound and so are your views.
Nate Silver knows that and he preached it while working at the Times and does it again in his book about statistics. Yet he went on anyway to hire people who operate that way. That’s where 538 has made its bed now, in a home where views lead data in a neighborhood occupied by logic fallacies, lies, and other awful things that are quickly destroying the reputation Silver worked hard to build.
Again, returning to the public’s reaction to what’s happened because they can tell you more about what’s happened than just about anything else:
The exact criticism of 538-ESPN is that it’s not doing the same thing that Nate Silver was doing at the Times. Read any of his election coverage from 2012 and compare it to this blog post and you’ll see the difference immediately.
On Blogspot and at the Times, Nate went to great lengths to acknowledge unlikely outcomes, just like he does in his book. Simulations, public methodology, consistency, open data, highlighting dissent when it rarely deserved it, answering criticism with data, etc. They were essays on statistics and logical frameworks for prediction.
Do you see any of that here on the new 538?
It wasn’t fair or realistic or to expect every writer working on this new site to live up to the standards that Nate Silver has set. He’s wouldn’t have been an outlier in quality and approachable statistical analysis if there were a bunch more people just like him ready and waiting to gain exposure if given a platform for their ideas.
The disappointment I feel is that he’s done nothing about it and said nothing about it. That science-challenged article on climate science that people have been using for target practice is the kind that usually gets people fired. But has there even been a statement from management acknowledging its problems?
Nate was the guy that you could rely on to do that. If his stuff had a hole in it, he’d be all over it. That’s where the trust came from. Not always being right, but knowing that when he was wrong, he’d find out why and fix it.
I don’t blame him and I’m not looking down on him personally or professionally because of this site and its writers. At worst, I now know that while he’s an exceptional mind in statistics when applied to elections, he’s just not that good at building media companies. And maybe in the end, that’s a compliment.
When Nate posts something on the upcoming mid-terms, I’ll know that I can rely on it because of his reputation for creating content like that. But everything else on this site, as far as I’m concerned, is just more flotsam of the type that Weblogs Inc was famous for. As short a post as possible, as many times per day as possible, for as little money as possible. Quantity over quality because this all has to be paid for with ads. Nate’s content may keep things afloat during election season, but otherwise they bills still have to be paid. What we’re seeing so far from 538 is garbage that pays the bills.
About half of this post was written as a comment on a fivethirtyeight.com post yesterday and I shamelessly stole it from myself to expand into a post for this blog, which is my excuse for why at least one thing I said is no longer relevant.
In fact, there have been two significant developments just since yesterday morning. The first is that Nate Silver has responded to the heat he’s taking over the climate science debacle. Although it’s an odd one. 538 is going to commission a trusted climate science expert to respond to the one written by Professor Roger Pielke Jr. Silver is going out of his way to pay a climate science expert to rebut an article on climate science written by one of his own employees, who himself teaches environmental science.
I don’t know if that kind of response is appropriate or meaningful. If you have to go outside your company to pay an expert to debunk your own content, wouldn’t the more appropriate response be to retract the problematic content and sever ties with its writer, and then refine your standards so that content based on questionable data doesn’t get published on a data-driven specialty site?
And isn’t this a waste of time and money? A number of climate scientists have already published critiques of Roger’s story, including Michael Mann, whose piece was pretty devastating in my view.
It also seems like maybe Nate doesn’t truly understand the scope of the problem he’s facing. The Pielke Jr. story is the most talked-about junk story at the new 538 because climate change is (because of politics, not the science) a hot-button issue. It’s far from the worst content being published there. Here’s a 499 word post about how many calories you can burn having sex that’s nothing more than a blurb post about a study. Another post wonders if American teens are becoming more promiscuous. One post, which ran a grand total of 188 words and contained an info graphic that was literally longer than the article itself was, counted religious keywords used to tag films on IMDb.com.
The problem facing 538 is that with the exception of Nate Silver’s own content, everything else is the sort of crap you can find in a men’s health magazine or on The Huffington Post.
The second big development is that the aforementioned Professor Roger Pielke Jr. is threatening to sue his fellow climate scientists who are criticizing the science of his post on 538, threats that 538 had to apologize for in private.