Stop Synergy Abuse

I also wish Synergy would track refs' calls, just to give fans something new to be mad about that isn't just "FT differential," which is the laziest, whiniest, sore-losing into which fans can engage. Step up, video statisticians! :D

@Ignarus: Very interesting. I didn't know anyone had tried to collect those data--kudos. I'd love it if Synergy would systematically collect that stat--it's just as observable as many of the other things they're painstakingly counting. I'd be a lot more comfortable with efficiency stats if they picked up things like this, double-teams (their effects), and a few other things, all of which are REALLY important in actual games.

Patrick J -- You make a good point about bail out shots. It would be useful if shots were catalogued with respect to time left in the shot clock and some indication of whether or not the shot was contested. I don't remember the numbers off hand, but I was trying to look over Lebron's 3pt shot selection when he got to Miami by basically watching all the clips on Synergy and I was surprised at how often he ended up taking spot-up or isolation threes at the end of the shot clock (i defined this before looking at it as shots w/less than 6 seconds left). I don't know how this applies to Melo, but in Bron's case, I very rarely saw the heat check threes that so often happened when he was in Cleveland. He also almost never put up the end of quarter halfcourt heaves that I'd come to expect to go in since the 08-09 season. I guess my point is that I totally agree that a lot of the time, top players get saddled with inefficient contested bail out isolation plays when all else failed. It can still be accounted for, but it'd take some real work.

Well put, Ethan. Good team defense play is still an area that individual stats don't quantify well for us. Synergy stats are a huge step up from the standard "these guys only allow X points per game" idiocy, but if you can't cover a guy's help defense tendencies, you're looking at a fairly incomplete picture. I'd like to add that on offense, I've never seen someone statistically account for a guy who spontaneously sets good screens, clearing lanes to the rim on broken plays and to stop a defensive rotation from getting to a corner three. Synergy's stats certainly add to the discussion, but you're absolutely right to ask folks to tone it down in citing good synergy stats as proof positive of a good defender.

On Rodman (easy to cherry pick the two most obvious examples, BTW - James and Rodman), are screen important to offensive sets? Does the fact Rodman was one of the best offensive rebounders (http://ow.ly/8GdS4) ever mean that opponents would be more likely hedging on his positioning? Perhaps resulting in fewer double teams, cheats, etc? Overall, does setting screens, making proper cuts, passes, rotating the ball out of double teams not significant impact the offence. Yes, taking LeBron James out of an offense hurts - wasn't questioning that... So to turn it around again: if Turkoglu doesn't participate in defense, he has Mr. Howard to compensate. Is that example not as valid? Regardless, I do 100% agree with your conclusion: "Outcomes are an important part of a complex story, but they should not be mistaken for the story itself." Like anything, one shouldn't rely on a sole data point to make definitive conclusions.

Very good article about how hard it is to analyze defense, but I wish this were written about the +/- stat. I feel that is the most overused stat to either denigrate or praise someone's defense.

I don't think this on Deron overhelping off Andre Miller. Can you even overhelp off Andre Miller? Looks like Shawne and Shelden got lost in leaving and rotating to Gallinari. They are probably the ones extending themselves too far.

@Ethan Sherwood Strauss: Great article. It's a nice "applied" example of one of the most important laws of science--that causality can't be determined without some way of observing the "counterfactual." In this case, it's denying the ball and therefore preventing points that isn't observed, meaning that an important part of great defense, which could change Synergy's ratings, isn't getting picked up in their estimates. I'm not an advanced stats fanboy, so I might not be aware of all the latest techniques being used to measure offense. But logically it seems that offensive stats must have the same non-causal interpretation, which their hard-core users never seem to mention. As far as I know, there's no counterfactual for shot selection. Some guys get labeled "volume shooters"; others are labeled "efficient." The latter is almost always considered a virtue and the former a vice. By watching games you can sometimes see that this is true, but far from always. Melo is always made to be the villain in this trope, and while I don't deny that sometimes he takes very difficult shots, it's often because the Knicks offense isn't producing better shot opportunities, so Melo ends up having to try to bail out teammates by shooting tough shots with the shot clock winding down. This is Mike Beasley's role on the Timberwolves too. In these cases, the teammates' efficiency stats never suffer, only the guys talented enough to *get* a shot in these situations, albeit a lower-percentage shot than anyone would like. Sound science would involve comparing guys like Melo's and Beasley's "efficiency" while accounting for these factors, which they don't, because they can't compare the shots that Landry Fields and Kevin Love can't create/don't take in those situations since these shots aren't observable. Btw: Loved the knights example. Felt like I was LARPing and reading a basketball statistics article at the same time. You rarely get that these days.

@Liston Cleveland once had the best offense. His name was LeBron James. Look, offense requires team work and coordination, but here is the difference: If someone doesn't participate on offense a la Dennis Rodman, the offense keeps humming. If someone doesn't participate on defense (they stand still when a rotation is required), it's death. Defense is more about collective competence and offense is more about individual brilliance.

"Offense is often created by one man, bursting through a defense en route to a tangible reward. This act is easily recorded and credited to the athlete." I couldn't disagree more. I would also argue: OFFENSE is about five people, working as one collective organism."