Stats scan: Is Ricky Rubio really 'irreplaceable' based on this skill?
Get the 1500 ESPN SportsWire delivered to your inbox daily, and keep up with all the news in Twin Cities Sports
I think Ricky Rubio is a tremendous basketball player. He has his strengths and he has one obvious flaw--his shooting.*
I've often wondered if the nature of his skill set makes it difficult to appreciate his ability. The easiest thing to observe is his individual scoring ability, where he doesn't stand out in a good way. His greatest assets - his passing and his individual defense - aren't measured in a standard box score and thus are harder to appreciate. But they matter.
The Wolves probably want Rubio to shore up his weaknesses and to be a more complete player. Fans certainly have grown frustrated with the point guard, but part of that is probably a larger-picture frustration with the way the season has gone for the Wolves.
Rubio isn't underperforming, though, he's exceeding expectations, writes Benjamin Morris at FiveThirtyEight.com, a website edited by Nate Silver. (The scope of this column isn't to dissect his process, but if you're interested in that stuff I'd recommend checking out the link.)
He says what many Wolves fans could agree with, that Rubio is elite in two areas: steals and not scoring.
Which is weighted more heavily?
Morris ran a regression that concluded steals are the most irreplaceable stat in an NBA box score. Even more surprisingly, they're more than 9 times more predictive than a point scored. This runs counter to a stat like John Hollinger's Player Efficiency Rating, which gives a steal about the same weight as a two-point basket.
What does that mean? Can steals be nine times more valuable than points?
Steals are unique because they end an opponent's possession literally every time and often start a favorable possession (fast break, numbers advantage, broken floor) for the thief's team.
Are they 9 times more valuable? No, I don't think that's what Morris is arguing. He's saying that if you take away a player from his team, you're effectively losing his stats. In areas like points and rebounds, his replacements will pick up some or most of the slack. Not the case with steals, and to a staggering degree.
Morris' actual argument is this: "A marginal steal is weighted nine times more heavily when predicting a player's impact than a marginal point."
A steal, Morris contends, is one of the most "informative" stats in the game, not that it's the most important. Scoreboards still count points and the team with the most wins.
Steals matter so much to Morris, though, that he says it's the most important stat to pull from a standard box score to tell you "as much as possible about whether a player helps or hurts his team."
He compared how teams perform with and without a certain player to try to figure out which statistics most accurately predict the differences. He concluded that the Wolves, for example, are nearly seven points per game worse when Rubio is not in the lineup.
I'd caution that we're dealing with somewhat of a limited sample, given the games Rubio and Kevin Love have missed over the past two seasons, and also given that before that, the roster composition was different.
Here's the final takeaway from Morris:
"Taken alone, this comparison [of Wolves teams with or without Rubio] doesn't answer the question of Rubio's value, and it doesn't prove that steals are as valuable as I think they are. But it's powerfully consistent with that claim. More important, it's a perfect example of how, even in a storm of complex, causally dynamic, massively intertwined data and information, sometimes odd little things that are known to be reliable and predictable are the most valuable."
*Anyone who criticizes Rubio for his poor shooting should also point out that he shoots 35% from on 3-pointers and 82% from the free throw line, both respectable percentages. He struggles to finish at the rim, although not as much as he did at the beginning of the season. His shooting needs to get better, but it's not holistically awful, as is often suggested.