A few weeks ago ESPN debuted ESPN Forecast, a panel of 216 ESPN employed or affiliated basketball experts. ESPN Forecast is an attempt by the worldwide leader to institutionalize the wildly popular NBA Summer Forecast and #NBArank features. Instead of having discrete times when ESPN calls upon a crowd for predictions, ESPN Forecast formalizes a panel at the ready to predict and react to any and all basketball news that may come up.
The justification for ESPN Forecast lays out its case: over the past five years, the Summer Forecast series has done a more accurate job of predicting win totals than any other system ESPN has measured, including Las Vegas. Despite the incomplete explanation—and lest you think ESPN is trying to pull one over on your by declaring 216 people “experts”—the statistical theory behind aggregating the wisdom of the crowd is sound. The basic idea is that aggregating the collective opinions of a large group is much more accurate than a single expert’s prediction, and has been proven true many times in a diverse group of settings.
Now, branding ESPN Forecast as a strict wisdom of the crowd model isn’t quite accurate, as ESPN is using it more broadly. It is best to utilize the wisdom of the crowd when predicting a single value on a continuum, such as a team’s exact win total between 0 and 82. It’s not nearly as accurate when predicting a discrete yes/no vote, and there is no way to ever tell if #NBArank rankings are “accurate” as they are completely subjective. Creating a panel to continuously pump out predictions has a bit of “manufacturing argument” to it, but let’s not look a gift horse in the mouth.
Back in January, when examining the future of ESPN’s TrueHoop network, I wrote: “The TrueHoop Network has the advantage of being comprised of highly-specialized experts on every NBA team. These experts could be encouraged to collaborate on longform pieces, crafting more knowledgeable pieces than individual writers alone can.” While I was specifying longform writing, the point stood across the entire TrueHoop network. With ~200 writers, wasn’t there a way for the network to better leverage its collective expertise?
ESPN Forecast is an ESPN.com initiative, but it has TrueHoop’s fragrance all over it. TrueHoop Network contributors make up 103 of the 216 panel members with nearly every blog being represented by at least one panelist, and TrueHoop editors Henry Abbott and Kevin Arnovitz are named as having “provided direction”. This is exactly the sort of thing the TrueHoop network should be involved in. They have literally hundreds of knowledgeable basketball writers who don’t collaborate often enough. What I didn’t expect is for them to be integrated with paid ESPN employees.
I’m not optimistic enough to believe that ESPN Forecast represents a significant lowering of the moneyed barrier between ESPN.com and TrueHoop, but also not pessimistic enough to say ESPN Forecast doesn’t matter. It is an interesting idea that I have no doubt will be successful. Hopefully it presages even more meaningful collaborations yet to come.