Archive for the ‘Stats Geekery’ Category

sloan sports analytics conference

Revenge of the Jocks: They Can Still Beat Us Up So Be Careful
Moderator: Will Leitch

Money, success, the respect of your peers — none of these will help when a professional athlete wants to fight you in the arena parking lot. This panel will discuss coping strategies such as avoiding eye contact and conducting interviews in pairs.

Panelists: Josh Smith, Kevin Garnett, Rasheed Wallace, Zach Randolph

#YES or #NO: The Ascendancy of the Twitter Poll as Cultural Barometer
Moderator: Darren Rovell

When Twitter launched in 2006, its creators could never have known the amazing ways the technology would be used: Organizing Occupy Wall Street protests, galvanizing the spirit of the Arab Spring, and most importantly, polling mass audiences of indiscriminate users on various sports topics.

The king of the online poll, Darren Rovell, shares the rules behind the secrets in front of the insights he gained from questions such as: “POLL: What song should the Northwestern band play at tonight’s halftime #AntsMarching #MovesLikeJagger #Soulman.”

Panelists: Darren Rovell, Richard Deitsch

To Retweet or to Favorite: Answering THE Social Media Question of Our Generation
Moderator: Jesse Eisenberg as Mark Zuckerberg

A million favorites is cool. But you know what’s cooler? A billion retweets. OR IS IT?

Panelists: Mark Cuban, Jack Dorsey, Joe Mande, Chelsea Peretti

Boner Down: Advances in Uniform Aerodynamics
Moderator: Tas Melas

As uniform technology trends toward tighter compression pants and shorts, male athletes are at high risk of awkward man-part mishaps. Can advances in fabric science lift and conceal these pesky pop-ups or are athletes destined to forever tuck under?

Panelists: Henrik Rummel

Numbers Analytics: How High Can a Human Count?
Moderator: The Count

What is the highest number? Some say 24, some say infinity, some say infinity +1. Will we ever know? Is there a compelling reason to find out? Where are my keys?

Panelists: Nate Silver, Malcolm Gladwell, Joshua Foer, Daniel Pink

Pi vs. Pie: A Special Luncheon Event (extra charge)

For every decimal place of Pi that autistic savant Daniel Tammet recites, competitive eater Takeru Kobayashi will consume a slice of Sara Lee Oven Fresh Cherry Pie®.

Participants: Daniel Tammet, Takeru Kobayashi, Maggie James, Don Povia

Others panels briefly considered:

Manalytics: Why are so many of us men? And why are we so awesome? High five!

Vanalytics: Would this be the best way to transport people to and from the conference?

Tanalytics: “Beyond the base tan.” Presented by Natural Sun salon.

They obviously existed a long time before this ever happened, but ever since Michael Lewis wrote about Shane Battier for the New York Times, advanced statistics have become an increasingly important part of basketball analysis. Get the right broadcaster and you’ll hear about points per possession instead of points per game, comments about how pace can affect raw team numbers or such and such a player’s rebound rate. There’s even occasionally a ticker on ESPN that lists the league leaders in PER. My dad emailed me about it once.

And while some coaches eschew this extra information that can be used to make their teams better, the real smart coaches have bought in to this stuff. The Spurs’ Gregg Popovich is one of those really smart coaches and he’s totally on board with one specific statistic. From the San Antonio Express-News:

“That score at the end of the game is huge,” Popovich said. “I don’t think it’s ever been wrong. It’s proven to be the most consistent stat.”

I did a little digging and it turns out that Gregg Popovich is right about this (surprise, surprise) — every team that has been leading at the end of each basketball game has come out with a win. There is a HUGE correlation between leading at the end of a game and who wins the game.

Does it surprise anyone that it would be the Spurs that figured this out? No way. They’re such a smart team — and they have an organization that is so committed to winning and finding new ways to do things — that it only makes sense that they would be the franchise that truly cracks the code on statistics. No wonder they’ve been good for so long, they just keep trying to have the lead at the end of every game. Very smart.

Color me zero percent surprised that Reggie Miller would be anti-advanced stats and pro-humans who play basketball. That’s not the least bit surprising to me, and I’m guessing I’m not alone in that.

However, I AM surprised that he called a “pocket protector” a “pocket square” while trying to make fun of nerds — who he actually called “lab geek rats,” which nope — because this guy wears pocket squares on the regular. Having been familiar with Reggie Miller’s oeuvre for quite some time, it’s a pretty safe bet that he regrets that mistake because being lumped in with all those dorks is probably the last thing he wants. A guy with a belly button tattoo would never be that uncool.

(via Basketball Prospectus)

Another day during the NBA season, yet another reason to talk about how “unclutch” LeBron James is. After LeBron missed two free throws and he shot none of the Heat’s three field goal attempts in the final minute of the Heat’s 78-75 loss to the Pacers last night, we had a whole new opportunity to declare his poor performance in game-deciding situations. As always, ESPN commentators like John Buccigross could be counted on to join the fun.

Ah, yes, the ol’ “cherry-pick numbers to fit a narrative” tactic. I know this technique because I’ve been guilty of it myself in the past — most notably two months ago, when I showcased the vast difference between Kobe Bryant’s and LeBron James’ performance numbers over the past three seasons, in the regular season or playoffs, in the fourth quarter or overtime, with 0:05 or less remaining in the game. Unsurprisingly, that post generated a lot of discussion, but do cherry-picked stats with a small sample size really prove anything? Let’s see how the narrative can change when we move the goal posts to a few different locations.

Read the rest of this entry »

On Friday night against the Utah Jazz, LeBron James put up one of those statlines where you would know it was him without even looking at the name attached to it: 35 points, 16-for-24 shooting, 10 rebounds, six assists, three blocks and no turnovers. Ridiculous, right? Too bad the Heat still lost the game by a point, and when LeBron passed out of a double team to a wide open Udonis Haslem for the potential game-winning shot — and Haslem missed the shot as the clock expired — the buzz on Twitter was mostly about how LeBron “wilted in the clutch” yet again.

The “LeBron isn’t clutch” narrative annoys me because too many people narrowly define clutch as “making the game-winning shot”. If you watched that final play and you’re completely objective about LeBron, you should agree with me that LeBron made the right play. LeBron was double-covered, Haslem is a decent mid-range shooter, and Haslem was open. Just because Kobe Bryant probably would have taken the shot in that situation, that doesn’t make it the right play.

Of course, what would be defined as “the right play” for most basketball players doesn’t apply to Kobe Bryant — according to some people, anyway. If you replace LeBron with Kobe and Haslem with Pau Gasol on that same play, I imagine most Lakers fans would want Kobe to take that shot. And the reason why they want Kobe to take that game-deciding shot is because of his reputation as the best clutch shooter in the NBA.

This reputation doesn’t just exist in the minds of fans. In a January survery of NBA General Managers by, 48.1 percent of the respondents said they would want Kobe “taking a shot with the game on the line”. In last year’s survey, Kobe was named by 78.6 percent of the respondents. Most likely, the GMs feel this way about Kobe for the same reason his fans do — he’s made a lot of memorable buzzer-beaters in his career. He’s also missed a lot of those shots, and I’ve long believed that the quantity of his misses gets overlooked in this narrative.

Luckily for us, recently launched their “Shot Finder” which claims to track every shot from the 2000-01 season through to this season — up to the February 28 games, as of this writing. The table below shows the results of my Shot Finder query with the following criteria: regular season or playoffs, fourth quarter or overtime, 0:05 or less remaining, shot to tie or to take the lead. The Shot Finder doesn’t let you query across multiple seasons, so I did it for this season and the previous two seasons and added up the numbers for 35 of the most prominent players in terms of overall fame/ability or their tendency to take these shots. If you think I excluded a worthy player, feel free to look him up yourself.

Read the rest of this entry »

Just a few days ago, Metta World Peace complained that Mike Brown wasn’t playing him enough because Brown is “a stats guy” and that “stats are for people who need stats.” Clearly, World Peace doesn’t need stats, as he told CBS Sports’ Ken Berger:

“If I could count how many times another team went away from the best player when I was on him, I’ve got to be like No. 1 in the league,” World Peace said. “That’s not a stat, and coach doesn’t … you would have to play basketball to feel that. When Phil Jackson was here, that’s why I was in the game, because he understands that.”

Interpreted for clarity — “Mike Brown doesn’t play me because he’s a nerd who never played professional basketball and doesn’t understand that what I bring to the table can’t be measured by stats.” It’s a very roundabout, passive-aggressive way of throwing your coach under the bus. Very Lakers of him.

But Mike Brown is not to be deterred. No sir. When one of his players criticizes him for being too stats-focused, he just doubles down on the stats. From the AP:

“If I were him, I’d be frustrated, too,” Brown said. “I told him: `I don’t take anything personally. I’m OK with it. But if was a stats guy, Metta, you wouldn’t be playing at all. I mean, look at your stats. And Synergy says you’re the 192nd-best defensive player in the league. So if I was just a stats guy, the guy who should be playing at the small forward spot is Devin Ebanks — because he’s shooting better than you or Matt (Barnes).”’

What a great comeback. He takes Metta World Peace’s criticisms and uses them as a very sharp zinger, like some sort of tai chi insult master.

There’s nowhere World Peace could take this because Brown is saying, “Yeah, I use stats a lot but even when they say you’re terrible, I still play you.” What can you say to that? Flawless victory for Mike Brown.

Of course, even though Mike Brown proved to be a wise zingster, this is still a coach and his player fighting about playing time in the media. That’s not the best thing for a team that is trying to figure out how to win games in a tough, tough conference. But still, when you have that sort of crack in your arsenal, you have to use it.

After all, Mike Brown wouldn’t want to deprive us of the hilarity of showing that a guy who thinks he’s the best defensive player is really the 192nd-best, would he? Of course not.

Dirk Nowitzki

How do I know I’m a certified NBA stats geek? Any time a player has an exceptional performance like Dirk Nowitzki’s virtuoso 48-point game on Tuesday, I typically go right to the Player Game Finder on to see how that performance ranks historically. In this particular case, I learned that only two players since 1991 have scored over 40 points with 15 or fewer field goal attempts in a playoff game — Nowitzki with 48 points on 15 attempts and Terry Porter (!) with 41 points on 14 attempts in 1992. If that doesn’t put into perspective how awesome Dirk’s performance was, I’m not sure what could.

Because Nowitzki has never won a championship ring, he isn’t widely considered to be one of the elite post-season performers of the modern era. It’s difficult to argue against this perception because, after all, rings are ultimately how we measure players. As ridiculous as it seems to many that people continue to compare Kobe Bryant to Michael Jordan, those comparisons will never go away if Kobe ever captures that sixth ring.

Nowitzki’s dominance in this year’s playoffs has me ruminating about where he ranks among players in the post-season, starting with the beginning of the Magic-Bird Era. Magic and Bird both made their NBA debuts in the 1979-80 season, so that seems like a good starting point for the sake of this discussion. In terms of how we compare players, I went with “Win Shares per 48 minutes” (WS/48) among players who have appeared in at least 50 playoff games and averaged at least 30 minutes per game.

In a nutshell, Win Shares are meant to measure how much a player contributes to his team’s success. The method for calculating this stat is exactly as convoluted as you’d expect, but as a reference point it helps to know that an average player should have a WS/48 score of around 0.1 — Jason Terry, Landry Fields and Rudy Fernandez were the three players to play at least 1,500 minutes and finish with exactly that score this season.

That’s enough of an introduction — let’s get to the money chart. Since the 1980 playoffs, 29 players have played in at least 50 games, averaged at least 30 minutes per game, and have a career post-season WP/48 score of at least .150.

Read the rest of this entry »