Tag: advanced stats

Oklahoma City Thunder v Miami Heat - Game Four

Former missile tracking technology revolutionizing NBA stats


People who dismiss the NBA’s statistical revolution as a bunch of nerds in front of computers with Google-like algorithm don’t get the real picture.

The reality is that a stats-loving coach who incorporates them into formulating his game plans is about to win an NBA title in Miami. The reality is pretty much every team has an advanced stats guy in house.

And those stats are going to the next level now thanks to former missile tracking technology that can literally track a player’s every move on the court. Teams are already paying for the next level of advanced statistics and the early results have wowed statheads at places like the SLOAN conference at MIT.

There’s a great story on the technology at fastcodedesign.com (go read the entire thing).

Now, SportVU systems hang from the catwalks of 10 NBA arenas, tiny webcams that silently track each player as they shoot, pass, and run across the court, recording each and every move 25 times a second. SportVU can tell you not just Kevin Durant’s shooting average, but his shooting average after dribbling one vs. two times, or his shooting average with a defender three feet away vs. five feet away. SportVU can actually consider both factors at once, plus take into account who passed him the ball, how many minutes he’d been on the court, and how many miles he’d run that game already.

A perfect example of this principle is in how SportVU can track assists. Traditionally, the NBA will award a player with the assist if they pass the ball to a player who immediately scores a field goal. It’s a human calculation, a judgment call based more on feel than science (much like a foul). Aside from clear issues of accurate counts, think about everything that generalized “assists” stat misses: What if the best passer in the world is on a team of people who can’t shoot? They’re making the correct play but getting no credit because they’re surrounded by mediocrity. But put this playmaker on a good team, and that team might excel.

SportVU can track these potential assists, and even track the person who passes it to the person who makes the assist. Now scale this level of assessment to every single skill in basketball, and our view of players could change as we know it.

The next level of challenge for coaches is to turn this data into something the players can assimilate. It’s one thing to tell a player “Kobe Bryant shoots X percent less if you force him to go left for two dribbles into a pull-up jumper” and another for a player to really grasp it and apply it in a game.

Heat coach Erik Spoelstra (among others) tries to put these tips in video packages — don’t tell the player, show the player. Most of us are visual learners, if we see it we process it faster.

But the revolution is here to stay, and with companies like SportVU there is only going to be more and more data out there to mine for an edge. And teams will pay and do just about anything to get an edge.

Making the advanced statistics movement relevant to players

Kobe Bryant, Joakim Noah, Luol Deng, Ronnie Brewer, Taj Gibson

There is a fantastic, must read post over TrueHoop by Henry Abbott, timed to coincide with the start of the SLOAN Conference — THE sports statistics conference — going on at MIT in Boston this weekend.

Abbott’s point is a good one: It’s one thing for a front office to have a bunch of advanced statistics in hand and use them in player and lineup evaluation, it’s another all together to explain those numbers to players. Or even some coaches. The new breed of coaches embraces these stats — Dallas last year used numbers extensively to see which five-man units worked best, Miami’s Erik Spoelstra is a numbers guy — but how do you get the lessons learned in those stats to the players?

Abbott has an idea:

Video. Game footage.

In other words, don’t tell me Kobe Bryant takes a lot of insanely difficult shots in crunch time, show me those fallaway leaners against the triple team….

Whatever your point is — that this player should play shooting guard, that that one is worth a ten-day contract because of his D-League rebound rate — stop saying it in nerd-speak charts and tables. Get ye to Synergy and show them all that stuff on game tape.

Use the stats and the spreadsheets more than ever, but behind-the-scenes, like plumbing — not as the final element in the presentation.

Synergy is a reference to Mysynergysports.com, a site that (for a fee) lets you look at video of games broken down by play and situation. I’m a big user, a lot of bloggers and writers are. If you want to see how LeBron James does in transition, you can call up the video of all his transition shot attempts almost instantly (he’s really good at it, by the way). Or you can see what Rajon Rondo does as the ball handler on side pick-and-rolls, or what Lamar Odom does when you force him right off the dribble. You get the idea. Very valuable scouting tool.

But if you’re trying to tell a player to attack off the pick-and-roll and stop pulling up for a jumper behind the screen, showing him a string of video on how he does in that situation (miss, miss, miss) is far more valuable then showing him stats on a page.

Abbott is right, the statistical revolution has to be televised for it to work.

NBA battle over NBA “Moneyball” stats mirrors regular life

1 Comment

We watched it play out in baseball around the time “Moneyball” the book came out. There were the old-school guys and Joe Morgan saying baseball can’t be defined by statistics and the new-school guys like Billy Beane and Theo Epstein winning that way.

Now the movie “Moneyball” is out and it is sparking the same debate in the NBA, where more advanced statistics are becoming part of the debate about players.

Henry Abbott over at TrueHoop saw the film — he would never miss the opening weekend of a Brad Pitt movie — and then talked to stats guys with a few NBA teams. Their reaction was that some of the scenes in the movie where stat guys are have a tough time getting attention is pretty much spot on. Abbott relates this story from a one guy working in the NBA.

I was asked to get involved in a negotiation with a certain player. I did a little homework on the guy, and then went back into the GM’s office, and asked how we should handle the guy’s injury history, specifically a torn ACL that had kept him out of the league for a year-and-a-half.

“He tore his ACL?” asked the GM, sounding surprised. “Where’d you learn that?”

I told him I had just googled the guy. This was in the last couple of years.

He said “ok, you’re going to have to show me how to use this google thing.”

The thing is, the stats can be an effective tool to use along with traditional scouting. That is what happens in Dallas, and that worked out pretty well for them (they love to use a variety of stats, particularly looking at how combinations of players work together). The Celtics and Thunder are stats-heavy teams that are doing well. There are others.

But there is plenty of resistance. Tons of it. Because you can’t define all the intangibles of the NBA into statistics, right? We’ve done just fine without stats like this in the past.

All the arguments struck Dwight Jaynes of CSNNW.com as familiar. Not just from baseball or sports, but from life.

I’m not alarmed when I see that battle between old school and new school because I’ve seen it play out my entire lifetime. I’m old enough to remember how offended people where when I got one of those early phone answering machines — the ones with the little tape recorders in them.

“I’m not going to speak to a recording,” someone in my family said. “That’s an insult.”

Computers? Remember when people hated them? And I mean really HATED them. Some still do, I guess. But I don’t know what I’d do without them. Same for cell phones. My goodness, people were downright offended that others felt they needed them.

Folks, advanced stats in the NBA are here to stay, and that’s a good thing. It’s not the answer to every question, but it can help. It’s a tool. In the end, what really matters is how well you know how to use the tools you have.

How do contending teams perform in the fourth quarter?

Miami Heat v Los Angeles Lakers

Always Be Closing.

That’s the name of the game, isn’t it? Great teams close out strong. They finish well together. They’re able to end games with finality. It’s that ability to lock down in the fourth quarter and finish teams off that separates the men from the boys. Young, weak teams can’t get it done, and teams that want to contend for a title have to be able to.

But how does this year’s crop of contenders compare in that fourth segment?

Using stats from HoopStats.com, I took a look at eight teams that could be considered “contenders” should you carry it out to a fairly liberal degree. And yes, Utah, I hear you, but I needed to make a cutoff somewhere and OKC happened to have more wins when I started this post. For reference, here are Utah’s stats.

Here’s what I found:

The first column of data is the difference between the average points scored in the fourth quarter and opponents’ points scored in the fourth. The second and third columns reference the winning percentages when leading or trailing after three quarters, and the fourth is to provide perspective for the season. All data was before Friday’s night’s games.

Before you start screaming, let’s get some caveats out of the way. The Lakers, as you’ll notice, have a 7.6 point differential and yet get outscored in the 4th. Which pretty much means they establish huge leads and then coast in the fourth when their subpar bench unit comes strolling in and let’s teams make pity runs. Some of it is on account of their inability to close out when down (as evidenced by their .25 winning percentage when down after three quarters). Similarly, there’s no data here to suggest who it is that these teams are battling down the wire. If the Mavericks keep coming back against teams they shouldn’t be down in the fourth to anyway, that doesn’t really inspire a lot of confidence. But still, if we take the fourth as its own game, they’re getting it done, and that’s the point of this little exercise.

What’s notable here is that the Spurs are the best team in the fourth as well as the best team in the league right now record-wise. Compare this with the Heat who are the third best team in the league and tops in point differential, yet fourth in fourth quarter differential, third in win percentage when leading after three, and fourth when trailing. Not great closers.

The Bulls on the other hand are a study in contrast. When things are going well, they’re going really well (second-best win percentage of any contender when winning after three), and when they’re not going well, they struggle to respond in the fourth (fifth-best win percentage when trailing after three. And with a tops differential of +3 in the 4th, it’s clear that the Bulls play exceptionally strong in the final frame when things are going well.

The Magic have real problems here. The worst point differential in the fourth, tied for fourth when leading, and a pitiful .23 win percentage when trailing. To put that in real terms, the Magic come back on their opponent in the fourth less than a quarter of the time. Getting outscored in the fourth is in and of itself and indictment, but when you look at it,the Magic have a very poor chance of winning should they not be leading after three.

The Mavericks best live by the idea of playing on edge the whole time. Nearly no difference in points scored versus points allowed in the fourth (.1 points better than their opponents’ average), the lowest win percentage of any team going into the fourth with a lead, but the second best percentage when trailing. They come back on their opponent half the time, while losing nearly 15% of the time with a lead. Injuries have likely bogged this down as the Mavericks have been in games late, only to fail without Dirk Nowitzki and Caron Butler.

Perhaps most interesting is the Thunder, though, with the second worst win percentage when leading, third worst when trailing, and third worst fourth quarter differential while having the worst total point differential. The Thunder very much seem like a young team that still struggles with closing out games from this data set, despite their reputation as a team wise beyond its years. While this data set is fraught with caveats like the one listed above and well beyond the boundaries of context, it’s an interesting set nonetheless.

Closing in the fourth during the regular season is vastly different than it is in the playoffs, when you can’t rattle off a 10-2 run against the Kings. But this data at least gives us a sense of what the top teams in the league are doing as the game reaches its close.  And while the Lakers have been great overall but bad in the fourth and the Bulls have been solid overall but brilliant in the fourth, both will need to keep working on keeping that extra gear in good condition. It can be the difference in a ring and disappointment when spring comes a calling.