(By Michael McNamara and Jason Calmes)
Just yesterday, ESPN Insider ran a story that set out to answer one simple question- Is Austin Rivers Having the Worst Season Ever? For those without Insider access or who simply lack the desire to read another anti-Rivers column, the author (Kevin Pelton) uses a player metric that he created for the NBA called WARP (Wins Above Replacement Player) and according to his projection, Austin Rivers is on pace to have the lowest score in the thirty-four year history of this metric. In order to balance out the perspective and help the answer his own question, we have posed some questions ourselves.
Is the Season Over Already?
Pelton is projecting how Rivers is finishing the season, but he is using the numbers Rivers has just a third of a way through the season. For the projection to be accurate, Rivers’ numbers over the final two-thirds of the season would have to mirror what we have seen from the first third. Here’s the problem with that, though- his numbers already look drastically different over these past three weeks when you compare them to what he was producing earlier in the season. Since December 7th, his points per game has risen by 50%, his field goal percentage is up 12%, and he has increased his rebounds and three-pointers made per game, all while decreasing his turnovers.
Even if he just maintains the numbers that he has produced in this recent ten-game stretch over the rest of the year, his end of season numbers will be quite respectable as the horrific numbers he posted in the first 16 games will be diluted. But what if he continues to improve? We have seen major jumps in the last three weeks, and while it might be unrealistic to see the numbers spike that much again, is it beyond the realm of comprehension that a guy who works as hard as Austin does will get even better? We have seen his PER practically double over this stretch, and even if he just maintains this recent play for the rest of the year, we are looking at a PER close to 10. Another jump in production could put him in the very respectable 11-13 region should Rivers sustain that improved play throughout the remainder of the season.
Increased individual numbers combined with better play from the Hornets as a team (and more wins) will drastically increase Rivers WARP numbers. It would take actual regression from Rivers at this point to maintain Pelton’s projection, and it would be surprising to see Rivers or the team move backward. This regression would also have to maintained while netting the significant minutes he has played up to this point. This could be affected by Monty reducing his minutes per his decision, injury, and an altered role to due to the expected return of Eric Gordon.
What are the Limitations of WARP?
Any statistic is limited. The more complicated the statistic, the more important it is to understand the limitation. WARP is immensely complicated, relying on layers of computations. The reach of the statistic is part of its power and value, but this power must be viewed along with its inherent limitations. According to the creator of the statistics and the author of the recent Rivers article, there are natural limitations to WARP.
Like all rating systems based on box-score data, WARP cannot account for contributions that are not tracked in the box score, most notably on defense. It does no better than linear weights methods at evaluating players like Bruce Bowen. Also, it requires a number of assumptions – the value of assists, the trade-off between usage and efficiency, and replacement level.
Defense seems to be a key issue with this statistic. In fact,
Generally speaking, a team’s total WARP plus the 10 replacement-level wins will be similar to its final win total. These tend to differ more for teams that are very good or very bad on defense, because not all of the team’s defensive performance is credited to its individual players. Some is assumed to be the function of the coaching staff and some we simply cannot fairly apportion to the individual players using box-score stats. Team WARP totals also differ from its record in the case of teams that outperform their Pythagorean projections.
With the Hornets having the 29th best defense in the NBA (they out-defend Charlotte), it calls into question if Rivers’ input data is appropriate.
Besides the potential for WARP to not predict wins for teams with extreme defensive prowess (in either sense), extreme defense at a team level can affect a player’s WARP.
These are the purely individual categories. The other two reflect team defense, and are not individually counted in the statistics. The two general schools of thought on team defense are to assign all of the credit to individuals, or none of it, and rate every player at league average. I’ve found a happy medium between the two.
What I’ve done is create what I call a “team defense factor” (TDF), which is simply Min/TmMin. If a player was on the court every minute of every game, this ratio would be 20 percent (TmMin includes minutes for all five players on the court, which is why you repeatedly see it divided by five in previous calculations). Self-explanatorily, this determines how much of the team’s defense the player gets credit for. What we’re assuming is that each player is equally responsible for the team defense his team plays while he’s on the floor. That’s not true, of course, but it’s a reasonable compromise.
This lowers the WARP of other players’ on team by attributing the horrible defense to them individually, but the argument presented is not that Austin Rivers is horrible; the argument is that he is historically horrible. Preferentially piling on terrible team defense into a poorly playing rookie’s reckoning due to massive minutes he’s playing that a totally out-of-line with his poor production is suspect at best.
This brings into question the use of a statistic like WARP to evaluate this Rivers’ play. After all, a WARP has two main components, according to the author: minutes played and per-minute-production. When per-minute-production is lower than replacement level, increased minutes played decreases the WARP statistic for the player. A certain number of minutes played are needed to ensure the per-minute-production is robustly estimated, but Rivers is playing more minutes than any rookie excepting Davis, Beal, Waiters, Lillard, and Sinlger (among those on pace to play at least 500 minutes for the season). The first four were chosen 1, 3, 4, 6, thus not available at 10, and Singler is a 24 year old drafted last season but who spent a season in Europe before joining the NBA. These players, excepting Singler, however, are performing better than replacement level and playing them more than Rivers is logical, or at least has the potential to be so depending on the alternatives.
The minutes played component of this statistic is highlighting his below-replacement level of play in a major way. This, however, is not to be placed on at the feet of Austin Rivers; rather, it is the choice of the Hornets’ organization, likely Monty. There is a huge difference between having a historically poor per-minute performance, which is how a historically bad player would perform, and hindering a team at a historic level, which is the choice of an organization whose goal differs from the fundamental assumption of the statistic in question:
Using wins gives a measure of value that is easy to understand and constant over time.
A rational coach who understands his players’ contributions will deploy a player with the lowest per-minute production on the team less than those with higher per-minute production when trying to win games at the highest possible rate. When the goal is to develop the player so that the maximum per-minute production is achieved sooner, then the statistic is not comparable to other similarly computer statistics for other players. After all, the team is not going after wins, at least wins during this season.
Additionally, Michael Beasely is producing at a rate lower then Austin Rivers according with WS and WS/48. This shores up the idea that the per-minute production being influenced by Rivers’ heavy minutes on a poor defensive team. Beasely plays fewer minutes on a team that is slightly better defensively, both of which affect Rivers’ WARP negatively compared to Beasely’s. Again, this shows that Rivers is merely a very bad player today, not a historically bad one.
Unlike Beasely, there is realistic hope that Rivers will improve. After all, some improvement has been noted.
To Play Poorly or Not to Play. Which is Worse? That is the question!
Finally, we have to address the question that this article aims to answer by using this statistic – Is Austin Rivers having the worst season ever? Even if we were to grant the author his claim that WARP is a superior measuring tool and also grant him the claim that Rivers numbers will not increase, thereby giving him the lowest WARP in history, does this support the claim that he is having the worst season ever? Which is worse, playing poorly or not being viewed by your coach as good enough to play?
When you look at it in relation to the value of a player to one’s team, is Austin Rivers or Terrance Jones having a better season? Rivers is putting up bad numbers for a Hornets team that is struggling, but has he not helped them win a game or two at the very least this season? His third quarter performance last night against Orlando helped them to jump out to a double digit lead, his 14 points and 6 assists with three key treys at Los Angeles helped them beat the Clippers, and so on. Meanwhile, Terrance Jones has only gotten off the Rockets bench in garbage time and is currently spending his time in the D-League. This can be said not only for Jones, but for the majority of players taken after Rivers in this year’s draft.
Some may say that Rivers hurts his team more than he helps because he is playing so much and that if you were to just insert a replacement player in his place, that the Hornets would be much better off. That is easier said then done however with this roster. The other guys who have gotten their turns at the shooting guard position, Darius Miller, Roger Mason, and Xavier Henry, are all below replacement level players so giving Rivers minutes to them likely would do nothing to improve the Hornets record.
What you have here is a player, who like most of his fellow rookies, is not ready to play heavy minutes in this league. He was forced to because of an injury to the team’s superstar and has struggled because he was given responsibilities that he was not ready for quite yet. While he started off at a historically poor rate, his numbers have jumped recently to the point where he is still poor when viewed as a starting shooting guard, but average when viewed through the spectrum of NBA player – 1 of 450. Upon Eric Gordon’s return, he will be moved into the role that he should have assumed all along and his numbers should shift quite drastically again, but that is not in Mr. Pelton’s projection. I am sure this article got plenty of page views and attention, and it is interesting to think about, to pontificate about the future of a young man just 18 months removed from his senior prom. But alas, it does not seem that the projection will come true. While we are likely to witness a season in which Rivers struggles, as most rookies do, it will by no means end up being historically bad.
Making the Transition is a Weekly piece that you can find every Thursday only on Hornets247. For past columns in this series, click here.