While few of us were satisfied with a single-digit win over a UMass-Lowell team that only beat Boston College by two, Rodney Purvis earned universal acclaim with an outstanding performance. Purvis is the only UConn player to score double-digit points in all ten games this season, a particularly impressive distinction given that the most common criticism of Purvis’ first season at UConn was a lack of consistency. So has Rodney really gotten more consistent?
To determine that, we first need to figure out what consistent performance looks like in a box score. And to do that, we’re gonna math and we’re gonna math hard.
One way to measure consistency is by calculating the standard deviation between a group of numbers. The issue with this method is that it doesn’t account for the relative value of numbers. In other words, the standard deviation for a players scoring is going to be higher than for steals just because he’s going to score a lot more points than he’s going to get steals. So we have to go one step further, and find the coefficient of variation, which really just means dividing the standard deviation of a group of numbers by the average of those numbers. If you’re a mathematician, you’ll know this is a slightly quick-and-dirty way to do this, but it will give us a pretty good picture of how consistent Purvis has been on a game-to-game basis.
We can do this with all of a player’s individual stats (points, rebounds, etc.), but to get the whole picture, we’re also going to use Game Score (GmSc), a total measure of a player’s performance in a game based on all statistical contributions.
Last season, Purvis averaged 11.6 points on .429 shooting, with an average game score of 4.15. That’s a respectable figure, if not a particularly noteworthy one. This year, those averages have increased to 15.1 points on .500 shooting, with an average game score of 7.95, the highest on the team.
Now here’s where things get really interesting* – Purvis’ scoring average of 11.6 last year came with a standard deviation of 6.97, which gives us a coefficient of variation (CV) of .599, meaning that a single standard deviation for Purvis represented about 60% of his average scoring. In other words, a “typical” game for Purvis last year meant anywhere from 4.6 points to 18.6 points.
So far this season, Purvis is showing a smaller standard deviation (5.3) on his scoring, this despite his average going up to 15.1 points. That’s a CV of about 35%, meaning that Rodney’s “typical” game this season is between 10 and 20 points. A glance at his game logs backs this up, as prior to yesterday’s game, those numbers represented his season-low and season-high in scoring, respectively.
But what if we want to look at more than just scoring? Well, his average GmSc has gone from 4.15 to 7.95, but the standard deviation has actually dropped from 6.65 to 6.16. This might not seem like a huge difference, but when you calculate the respective CV percentages, you’re looking at a drop from 1.60 last year to .775 this season. To put it more bluntly, using game score, we can actually calculate that Purvis is twice as consistent as he was last year.
The results are very similar across Purvis’ entire statistical record. Purvis’ average field goal attempts have barely budged since last season, going from 10.2 per game last year to 11.0 this year. By the eye test, Purvis seemed to disappear for stretches last year, something that I’ve noticed less this season. This is borne out by the numbers, too, as his SD has dropped from 3.85 (CV of .378) to 1.73 (DV of .157). His season-low for field goal attempts is eight, and his high is 14 (yesterday’s game). He has ten, 11, or 12 attempts in six of ten games this season. As far as offensive roles go, that’s about as consistent as it gets.
For some comparison’s sake, Daniel Hamilton has a CV of .369 for his scoring and .395 for his shot attempts. His CV for field goals (.358) is virtually identical to Purvis’ (.359). His game scores have been more consistent (.557) largely because he continues to provide rebounding and assists when he’s not scoring, but in terms of pure production, Purvis is right there with him. None of the other starters is even close to these two in terms of consistency.
While the team hasn’t been quite as good as many of us had hoped, going into the season, Purvis has answered the bell on nearly all of the areas he needed to improve: his three-point shooting has gone from passable to elite, and he’s gone from being one of the team’s least consistent performers to one of its most consistent. The only thing left, from this blogger’s chair, is to improve that free throw shooting. Let’s just say that given the improvements he’s shown thus far, I’m not betting against it happening.
