It’s now official: the mid-range jumper has become anathema in Canada.
Need proof?
C’est ici:
No wait, that’s the formula for “uphill downhill flux,” or the average change in potential shooting percentage per pass, where “p ij is the probability of the link between players i and j, xi and xj are their shooting percentages.”
Of course.
So, the death of the mid-range jumper formula must be:
where “deg(v) is the degree of the node v, deg(v*) is the highest degree node, V is the number of nodes.”
Huh?
Wait, my mistake, once again. That’s actually the formula for determining if an opponent has a dominant player, which if used in conjunction with the “clustering co-efficient,” will tell you whether or not double-teaming the star player will cause the opposingteam’s ball movement to essentially stall.
Welcome to the wild world of hoops analytics, which is taking Canadian Interuniversity Sport by storm this season, what with Ontario University Athletics men’s and women’s programs, along with Canada West men’s programs, having joined CanWest’s women’s programs (who jointly got a head start on the craze last year) by shelling out for a group subscription (roughly $2,000 per institution) to Synergy Sports Technology’s “web-based, on-demand video-supported basketball analytics.”
It’s even mobile. It’s probably hotter than Steph Curry. Okay, maybe not. But hoops data-crunching is so hot that there are now annual basketball analytics summits, eight-week on-line courses and even a new research journal devoted to the mathematics of sport. Coaches report that they absolutely love Synergy, along with other programs such as Krossover.In fact, they’re convinced hoops analytics is the greatest invention since, well, basketball itself.
But then, they’re the kind of folks who used to watch 44 hours of video tape per week, and now they have a computerized tool that whittles their hours down to say, 35 per week, freeing-up a whole lot of time to pore over performance metrics like “offensive efficiency, efficient offensive production, and defensive stops gained,” so the upshot is 64 hours per week. And really, who could ask for more happiness than that?
Or as Calgary women’s coach Damian Jennings notes, “I think, because we’re nerdy work-a-holics, it means that we probably enjoy those extra 20 hours.”
University of British Columbia men’s coach Kevin Hanson says the scope and specificity of the available information is utterly phenomenal. Every play of every game is dissected and charted and funnelled into a mind-boggling array of metrics. “It’s been one of the best advancements in basketball that I’ve seen. … It gives you breakdowns, the strengths and weaknesses of all players. It tells you how many times a guy goes to his left, to his right. There’s so much information in there, you can’t possibly use it all.”
“We use it because FIBA is so ball-screen dominated,” says men’scoach David DeAveiro of McGill, which along with other RSEQ schools is currently negotiating a group subscription for all the league’s teams. “Who’s involved in the ball screen? What side of the floor? What’s the success rate? How is it defended and what’s the success rate with each defence? That’s a big component for us. And then we like to know how other teams are scoring, in terms of location, how many threes they’re putting up, which side of the floor, that kind of stuff. But overall, we’ve used it more to see what an individual opponent does. Does he shoot it from one side of the floor more? Is he a mid-range guy? Is he a three-point shooter? Does he post more on one side of the block than the other side of the block?”
It’s exactly the kind of “book” that most coaches across the country have long kept on all players, only in the past, it wasn’t anywhere near as comprehensive.
“You can actually click on an individual player’s name and watch what he did on every one of his possessions,” says Victoria men’s coach Craig Beaucamp. “You can also look at every side-out, every end-out by opposing teams” and quickly find out, for example, how many times a team ran a specific play off the inbounds.
Opponent’s offences can be broken down to readily determine whether they score 60% or 20% of their points in the post, or in transition, or however else they might be putting the ball in the hole, “while you can look at your own team’s efficiency, both offensively and defensively, in terms of points per possession. Who’s most efficient against a man offence, a zone offence?” Beaucamp adds.
“You could probably go a little overboard about and become a little obsessed about all the analytics and all the video clips you could pull out, so you have to pick your spots,” he notes.
Beaucamp and his colleagues also stress that all the data has to be used with a grain of salt, because it doesn’t capture several intangible, but critical, facets of the game.
“There’s nothing like the eye test,” says DeAveiro. “There’s nothing like having a feel for the game and the flow of the game.”
“I don’t think it can measure things like how hard kids are playing or whether a kid will take a charge, diving for loose balls, talking to his teammates,” says Guelph men’s coach Chris O’Rourke. “There’s other parts of the game that are so integral to winning. I don’t know that you can throw a number on those things.”
“I always ask: did we get the open shot?” notes Laurentian men’scoach Shawn Swords. “Did we run the play correctly and miss the open shot? I haven’t seen the analytics yet that actually shows me that.”
“I guess I’m a mix of young and old when it comes to analytics,” Swords adds. “I use it but I still trust my eyes more than the stats.” Only by watching a player on film or in person can you “get a feel for his game. A lot of players, if you watch them you can tell, he doesn’t want the ball. He’s scared to have the ball in a clutch situation. So why are you closing out like crazy, when he doesn’t want to shoot it?”
Beaucamp concurs. “It’s tough to get a feel for the players you’re facing and team you’re playing without actually watching the game itself because the clips are isolated clips, 10 seconds, 12 seconds in nature. … Obviously, the art is to try to pull out those nuggets that are useful.”
The question then becomes: How much, if any, of that veritable mountain of data is useful on the floor?
Pundits say that basketball analytics has actually caused a paradigm shift in the game. One of the central tenets of most analytics programs, for example, is that it puts a greater premium on shooting a trey, or putting the ball on the floor and attacking off the dribble, rather than attempting a mid-range jumper. In the former instance, a bucket is worth three points, rather than two, and in the latter, there are significantly increased odds of drawing a foul off a penetration-dribble than off a pull-up jumper. (To be precise, according to basketball analytics calculations, shots taken between one-metre from the basket to the three-point arc are, on average, worth 0.76-0.81 points, while those taken at the rim are worth 1.5 points, corner treys are worth 1.2 points, while treys taken from the top of the key and the wings are worth 1.05 points. Hence, the mid-range jumper is, statistically, a less-efficient or productive shot).
There’s some legitimacy in that argument, particularly with players who can’t hit those mid-range jumpers, O’Rourke says, adding that he has one player “whose percentages and scoring have gone through the roof this semester … once we were able to show him that if he got to the rim more, he’d draw more fouls.”
But the coaches also caution that it’s altogether too easy, if not pointless, to overwhelm their troops with reams of data generated from analytics about their opponents. Is it realistic, after all, for a coach to expect that defender ‘X’ will remember that opposing ballhandler ‘Y’ takes a runner 17.6% of the time when moving away from a pick? Or to recall that he goes 1.1 kilometres per hour faster when moving to his right than his left?
Or, as Hanson notes, “every team in the country would tell you, we have guys on our team that still don’t know what we do, let alone worrying about memorizing what another team does, and all its players do.”
“A lot of players will take in a lot,” Swords notes. “But you can’t expect them remember all of it.”
Still, the coaches appear generally agreed that analytics are incredibly useful during scouting and game preparation, by helping them to understand what foes are doing and then crafting the best strategies or tactics to counteract those strengths. Coaches can even get statistics on the likelihood that a foe will do something unpredictable.
DeAveiro’s a believer. “It’s definitely changed the game. Everybody is more prepared now.” Observers might argue that just means everybody is clogging the paint (a fundamental tenet of most hoops analytics programs). But DeAveiro says there’s a quantitative value beyond just general strategy, in that judicious use of data can result in an extra possession or two, which can alter a game’s outcome.
But Jennings cautions that an increased level of preparation is not necessarily, or entirely, a function of basketball analytics. It’s a new era in Canada, in which coaches are now full-time mentors and far more professional about their craft. “There can be a number of reasons why a team is more prepared than in the past. Perhaps one coach last summer went to watch a lot of Australian basketball and suddenly brings in a lot more as a 94-feet defensive, disruptive break-up ball screen, likes the flex screen, shuffle-continuity offence, which Australians seem to be infatuated with, and suddenly, you’ve got another ingredient. … We’ve got some super-bright people leading these basketball programs,” and that may be as much a factor as anything else. “The coaching IQ here is very good. What Synergy and these analytics programs may have done is accelerate what is a very good standard of coaching.”
“The artistry is in how we lead and manage. And how much of that information do we give them? More isn’t necessarily better,” Jennings adds. “The best coaches know what not to tell their players.”
That’s certainly a valid argument given that all the basketball analytics in the world doesn’t necessarily translate into an improvement in an individual player’s actual skill level, or a team’s performance on the floor, and that it’s so easy to simply overwhelm players with data.
Has analytics actually elevated the level of play or made CIS games more competitive because teams are better prepared? Or is all just statistical, if not evangelical, zeal?
“I don’t know, but I reckon it has helped,” Jennings says. “Remember you’re talking to a coach who has always used video analysis as a teaching tool, as opposed to a ‘bore them by opponent-scouting’ stuff only. So we have two film sessions per week and we purposefully do that because that outside perspective of review, and reflection, and personal analysis, is important to the teaching and learning process.”
Therein lies, perhaps the greatest value of basketball analytics. They are a marvelous teaching tool, the coaches say.
“It’s a great accountability piece for our players,” DeAveiro notes. “You can show them the data (and then ask) …does this warrant more playing time?”
It can be a real “eye-opener” for a player if he’s shown all of the turnovers he’s made over the course of a season, Swords notes.
“There’s certainly value in it with your own players,” O’Rourke says.
Do players sometimes become aggravated when shown the cold, hard truth about their player efficiency rating (which calculates per-minute productivity)?
O’Rourke wryly responds: “I think that, as coaches, we’d probably all agree that a lot of times, players are a little delusional about their own abilities.”