This is the first in our new “Roundtable” series. “Roundtables” will feature discussions between writers, analysts, and the like. This series is sponsored by King Arthur and the Knights of the Round Table.
On Monday, ESPN.com introduced Real Plus-Minus. It was developed by Jeremias Engelmann and Steve Ilardi, two well-respected NBA analytics people. For an introduction, go read Steve Ilardi’s primer and explore the RPM leaderboard. ESPN Insider’s Kevin Pelton also shared his RPM All-Stars.
Jonathan Gordon and Jacob Rosen share their thoughts below. Jonathan is the founding editor of Sports Analytics Blog. Jacob is a contributor to Sports Analytics Blog and a long-time writer at WaitingForNextYear. Feel free to comment with your thoughts.
So what was your first reaction to this unveil?
JONATHAN: I loved seeing ESPN embrace the metric. I knew it is fairly similar to other metrics already out there in the blogosphere (such as APM and RAPM), so I wasn’t so much “impressed” by the metric. However, I was impressed that ESPN is fully devoted to bringing advanced stats to the public. Much of the general public will not have had stumbled upon APM or RAPM, so this is their first look at such a metric. It’s a big step in helping the general public better understand the game.
JACOB: Surprise. It’s April 7, why is ESPN suddenly unveiling a basketball stat with only a week or so left in the NBA regular season? It seemed odd at first. Perhaps it makes more sense in the context of MVP debates and 6th Man votes -- as Deadspin’s Kyle Wagner pointed out -- but that’s relatively lame. Then, as I started to read into the posts, I was a bit disappointed. It’s hard not to be disappointed when the explanation of a new statistic leaves a lot to the reader’s imagination.
What did you know about APM and RAPM before today?
JACOB: A decent amount, mostly thanks to Kevin Hetrick’s work at gotbuckets.com. Hetrick, who writes usually at Cavs: The Blog, started that website to help RAPM gain some popular notoriety. So yeah, I was at least conversational in RAPM language. It has its flaws, no doubt. But the point here is that this stat -- or at least the ancestors of ESPN’s 2014 product -- was not new by any means.
[Spoiler Alert: We’ll have another post tomorrow with thoughts from Kevin.]
JONATHAN: I had heard of the two and understood what they meant. I’ve seen rankings and I’ve seen writers use the stats in their pieces. However, I haven’t known much about the methods behind getting such stats. Not because the methods aren’t explained online (they are), but rather because I never looked too much into it.
The statistic’s proprietary nature came under fire quickly. Any additional thoughts on this?
JONATHAN: As a fan interested in advanced statistics, I would love to see the formula that produces RPM. However, I also understand the business side of it. At the end of the day, ESPN is a business. Keeping the statistic proprietary is a competitive advantage for ESPN. Releasing the formula and other details would be akin to to McDonald’s releasing the Big Mac recipe. Sure, I’d love it for ESPN to release more info on it, but I understand why they don’t (and won’t.)
JACOB: It’s sad for the industry, in the end. As I wrote in my posts about the Sloan Sports Analytics Conference and about the Cleveland Cavaliers analytics team, the analytics landscape has changed. Gone are the days of open-sourced data in forums ripe with conversation and intrigue. I get it, sure. But they’ll be feasting on the non-stats-inclined with this metric, while there still exists a lot of potential for engaging the really smart analytical people in necessary dialogue.
Why do you think there’s such a yearning for the Holy Grail of stats, one number to rule them all?
JACOB: Why do we like lists? It’s the same psychological theory. People love clinging to one number, one truth. Perhaps it’s also why ESPN invested so heavily in Total QBR. In basketball, PER already has been fairly well established. WAR is the thing in baseball. So perhaps ESPN saw an opportunity for Real Plus-Minus, again in basketball. I’m skeptical because many already knew about APM and RAPM and the necessary caveats are well-documented. But it assuredly can still be packaged in SEO- and profit-friendly ways, no doubt.
JONATHAN: Because it’s the easiest way to implement statistics into broadcasts, quick analysis, and the like. It’s easier for a broadcaster to mention a guy’s Real Plus-Minus or PER than it is to dive deep into rebounding percentage (although that shouldn’t be too hard, but I digress). Furthermore, I think the general public is mostly interested in questions like “Who is the best player?” and “Who’s better, Player X or Player Y?” Frankly, that’s ESPN’s audience - the general public.
Moving forward, where do you see this RPM metric and ESPN’s investment changing the game and our consumption of it, if at all?
JONATHAN: I see it as a middle ground between standard Plus-Minus and Player Efficiency Rating (PER). Raw plus-minus data was the most easily understood “advanced” metric (it wasn’t really THAT advanced). It could be simply explained to the average fan. PER, on the other hand, was much more advanced, yet much harder to understand. Sure, it accounts for a variety of factors, but it’s not easily explainable. I think Real Plus-Minus fits in the middle. At it’s core, RPM measures the same thing as raw plus-minus. However, it incorporates a deeper and better analysis, similar to PER. All that said, I don’t see it changing the game too much. I doubt teams were using standard plus-minus to make decisions. It might be used more in broadcasts and by the media. At the end of the day, it’s just another tool in the toolbox to evaluate players. A constantly growing toolbox.
JACOB: Basketball is a very difficult game to “solve” because there are so many constantly flowing pieces and parts. Baseball is so much different in that the game is mostly a series of independent activities. In basketball, we’re still getting better and better about how to evaluate defense and intangibles in an easily-understood fashion. Perhaps this helps; at least it’s ESPN putting its massive weight behind one of those tools in the toolbox.
What do you think of ESPN's new Real Plus-Minus stat?