Jan.25

electoral-vote.com

Dem 51

   

GOP 49

FiveThirtyEight Could Be in Trouble

Way back in the Obama years, ABC News realized that it had missed the boat when it came to modern polling analysis. And they solved that problem by backing up the Brinks Truck for Nate Silver, acquiring both his services and his FiveThirtyEight brand. It seemed like a win-win for both sides; ABC became an instant leader in the polling analysis biz and Silver gained access to money and other resources that would allow him to build a much more ambitious and robust site.

Ultimately, it hasn't worked out all that well for ABC. FiveThirtyEight has been an awkward fit, and has been shuffled around a bit, sometimes placed under the ESPN umbrella, sometimes under the ABC News umbrella. Further, the site has never turned a profit, and there are now enough others in the polling analysis game that conversion to a subscription-based model is probably not viable, given the amount of overhead the site has. ABC is also cutting costs right now, as most media outlets are. Oh, and the ABC brass is apparently unhappy with Silver's performance in the most recent election cycle, thinking that he did not do enough to eliminate/downplay obvious partisan pollsters from his data set.

Add it all up, and it is very possible that ABC will try to offload the site. Failing that, the company might just shake Silver's hand and wish him good luck when his contract expires later this summer. Given that FiveThirtyEight has been leaving key posts vacant recently, rather than fill them when the occupant departs, it looks to be considerably more likely than not that the ABC-FiveThirtyEight relationship is nearing its end.

From where we sit, the current iteration of FiveThirtyEight, on the whole, just isn't that great. Certainly, it hasn't lived up to the promise that the merger with ABC seemed to suggest. The most significant problems, in our view:

• It's trying to do too much. It's interesting, perhaps, to try to bring data analysis to sports and science and the arts and politics and so forth, but the site ended up as a hodgepodge of stuff that just isn't very coherent as a whole. It does not help that, for whatever reason, turnover there has been staggering, such that there's little in the way of a consistent editorial voice.
  
  
• Silver's data-driven approach works well for politics. It also works pretty well for baseball, which is where he got his start. But it works considerably less well for a lot of the other things that FiveThirtyEight has tried to tackle. For example, its models for other sports, like basketball and football, have often produced results that were downright embarrassing (like predicting the Golden State Warriors would miss the playoffs in a season where they won the NBA championship). And trying to bring the numbers-driven approach to the arts often produced twaddle like "The Five Types of Nicolas Cage Movies."
  
  
• Some people do actual research, which often takes weeks or months or years. Others, including us, comment on such research. Often, FiveThirtyEight is in between those two poles—asking interesting questions, but then trying to produce substantive, data-driven answers on a daily production schedule. It just doesn't work like that.

What we are saying here is that Silver and FiveThirtyEight might be much better off if the current model is abandoned, and Silver goes back to running a much leaner and meaner site that focuses entirely on politics and polling analysis.

Of course, it's also possible that Silver decides he's weary of the publishing game, and that he instead will become a hotshot consultant or the in-house polling guru for, say, the DNC. Still another possibility is that ABC actually wants to keep Silver, but that they are currently playing hardball in order to get him to drop his price and or to accede to some of their complaints about his models. We should know what's going to happen in the next few months. (Z)