FiveThirtyEight Insults NBATV. FiveThirtyEight Is Super Wrong
I have no idea what Walt Hickey of FiveThirtyEight has against NBATV, but he attacks and insults the network in his recent article, NBA Nerds Are Right — Most Matchups On NBA TV Aren’t So Hot (or as the article’s url puts it: spurs-warriors-was-a-fluke-most-matchups-on-nba-tv-are-garbage). In the article, the author not-so-subtley shares his belief that NBATV is terrible at selecting games to broadcast, and that nothing on NBATV is worthwhile. These statements coming from an NBA partner (FiveThirtyEight is owned by ESPN) are not just wildly irresponsible, but wildly incorrect.
The author seems to have a fundamental misunderstanding of the role NBA TV plays within the NBA Family of networks. Unlike the NFL Network, NBATV does not siphon away games from networks with more visibility so as to increase demand for their own network, but presents additional content in support of networks that are more visible. It understands that its product reaches a significantly smaller number of households, and it behaves as such. There is a direct, intentional, correlation between the Average ELO of a National TV Network’s Schedule and the number of homes it reaches. Additionally, for most of the years in question NBATV Tuesday Fan Night – in which fans vote on which game is broadcast – is removed from the article’s analysis. Fan Night inherently gives fans exactly what they ask for, and if one assumes fans are looking for the best matchups based on ELO (which the article does), it’s not a stretch to say the inclusion of Fan Night games in the analysis would significantly boost the average NBATV ELO.
The author also seems to have a fundamental misunderstanding of what basketball nerds consider interesting. The NBA utilizes its network to showcase teams and individuals that are usually passed over by the larger networks to attempt to familiarize them with audiences. It could be a young team on the rise (think the Timberwolves), a young Superstar on a bad team (think DeMarcus Cousins), or the always overlooked Atlanta Hawks and Memphis Grizzlies. These are games and teams that so called basketball nerds geek out over, and it gives exposure to teams that are rarely if ever are given a nationally televised audience.
Lastly, disregarding the shortcut the author takes in using combined ELO and not taking into account any ELO disparity signaling a competitive mismatch (which because of the frequent presence of ELO outliers such as the Golden State Warriors is larger on average in non-NBATV nationally televised games), he seems to have a fundamental misunderstanding of how to statistically quantify the subset of games from which NBATV has the option to broadcast. I don’t have the time or the resources to do the full analysis, but using the average ELO for each network as a placeholder, the number of games this season on each network as a similar placeholder, and given what was discussed earlier about the nature of games that end up on NBATV (they get the network scraps) the average combined ELO of the games NBATV has to choose from is 3009 (not the league average of 3022 that he uses). Additionally, NBATV rarely has access to extreme outliers such as Monday night’s Spurs-Warriors matchup, which tend to skew network data. Even when they do, the outliers have a smaller effect on overall ELO, since NBATV broadcasts nearly as many games as the other networks combined. All this is to say that amidst all the noise inherent in the nature of the games NBATV broadcasts, the network still manages to present the viewer with an above average ELO product (ELO of 3038).
Walt Hickey’s analysis of NBATV’s output is misguided, incorrect, and irresponsible coming from a respected news source, but it certainly makes sense within the context of ESPN’s coverage of the NBA. As a viewer, one of the most frustrating aspects of the current NBA network landscape is ESPN’s unwillingness to play along with Turner (which owns NBATV) – they almost never include non-Disney (Disney owns ABC which owns ESPN) games on their upcoming network schedule or publicize games that are on other networks. I do, however, wish NBATV had been able to properly capitalize on the perceived magnitude of Monday night’s game, using it to publicize and drive traffic to their own website through a livestream event. The biggest real issue basketball nerds have with NBATV broadcasts is the resultant NBA League Pass blackout of games on the network. Since other Turner networks give viewers the ability to livestream the network online, it isn’t farfetched to believe NBATV has the resources to produce such a stream, and clearly given Walt Hickey’s hackneyed article, there was a demand for it.
Let’s hope the veracity of this article is not a portend of things to come for FiveThirtyEight, because if it is, Nate Silver is going to have a long year.