The Cloudy Point-Differential Forecast

At some point over the past few years, point-differential started mattering. [Editor’s note: On a very basic level, i.e. deciding who won the game, point differential has always mattered. But we get what Jackson is saying. Also, here’s an advanced stats primer and a reference guide for all the advanced stat formulas, if you like those kinds of things.]

Coinciding with John Hollinger’s rise in public perception at ESPN, regular fans and statisticians alike now love to reference point-differential as a means for predicting anything in the NBA. You can find arguments that margin of victory is a better indicator of a team’s anticipated playoff success than win-loss record, that point-differential is the most central factor of a team’s future record, and even that the statistic is the best predictor of the eventual NBA champion.

As a fan of the 2011-2012 Utah Jazz, this is a decidedly bad thing. The Jazz are currently 20th in the league in point differential, scoring 95.6 points per game and giving up 97.1, for a difference of -1.5 points per game. [Editor’s note: Mercifully, the Jazz are nowhere near the abysmal performance of the Charlotte Bobcats, who are outscored by 14 points per game, on average.]

One glance at the NBA standings will show you that point-differential is something the Jazz would rather not talk about (though they would talk about it a hundred times before mentioning Monday’s shameful, awful, absurd loss to the short-handed, disowned, and hopeless New Orleans Hornets… not that I’m still bitter about it.) NBA statisticians have seen that number in the red for the majority of the season and have accordingly been prophesying the downfall of this season’s Jazz team for weeks now.

At this point, with the Jazz only 2-7 in February, I begrudgingly admit that the numbers so far show them to be validated.

Despite having an average strength of schedule, a 14-14 record, and boasting wins over six of the top ten other teams in the Western Conference, that little -1.5 in the red results in ESPN predicting that the Jazz only have a 15.6 percent chance of sneaking into the playoffs this year.

Thankfully, this isn’t college football, where nothing more substantial than style points [and probably shady back-room deals] provide the final say on whether a team is worthy of advancement or not. Still, it is ridiculous to say that the Jazz are doomed because they have averaged scoring a half point less than their opponents. Mind you, that is cumulatively 11 fewer points scored than their opponents. When you remove the first two games of this season—in which the Jazz lost by a cumulative 42 points–the Jazz are suddenly 31 points above their opponents, rocking a +1.24 margin of victory across the past 25 games (well, up until the Oklahoma City game on Tuesday).

From a statistical prognostication standpoint, the Jazz dug themselves a huge hole in those first two games of the season. If they had lost both of those games by just a point each instead, would they really be that different of a team at this point? Would they really be that much more likely to make the playoffs?

Maybe I’m just playing the apologist because the advanced basketball statistical stars are currently lined up against the Jazz. I’d like to think that if they don’t make the playoffs this year it isn’t because they got blown out in some early, fairly insignificant games. I want it to be because of basketball reasons, like they couldn’t crack a zone defense when they needed to or because they weren’t able to keep the opponent out of the paint in the fourth quarter. At the same time, if this season turns into a wild success, I don’t want it to be a victory for the predictive power of point-differentials–I want it to be because Alec Burks developed into an efficient, confident scorer from the bench, Gordon Hayward learned to get all of his shots within the flow of the offense, and because Earl Watson was always willing to splurge for a technical foul when he thought the team’s fire was getting a little low.

I’m not entirely sure what legitimate case there is to made against point-differential as an integral NBA stat; for all I know, the law of large numbers could be in play here and it could have a really strong correlation with team success. All I’m certain of is that this isn’t a team we can count out yet, despite the last couple of disastrous losses. I haven’t seen the Jazz put up many style points this season, but I have seen them play with a lot of heart.