2014 Sloan Sports Analytics Conference: In-Game “Innovations” With George Karl and Daryl Morey

via flickr | research451

via flickr | research451

The MIT Sloan Sports Analytics Conference is built on a foundation of forward thinking. The sports world converges on Boston for a weekend every winter with one all-encompassing goal in mind: To broaden understanding of the games we love through avenues previously unexplored.

The intentions of “In-Game Innovations: Genius or Gimmick?” – a panel comprised of George Karl, Daryl Morey, Bill James, Nate Silver, and Kevin Kelley – were undoubtedly to further such discovery; progress is bred by new curiosities. Mere minutes into the hour-long discussion led by ESPN’s Kevin Arnovitz, though, it became evident the speakers were focused on “innovations” that challenged the very definition of the word.

Karl and Morey are two of basketball’s foremost pioneers of the analytical age. Their teams stress efficiency above all else, leading to an invariably large discrepancy between shot attempts that carry higher expected outcomes and those that hold lower ones. The thinking behind that strategy is unquestionably sound on the surface, and the public majority is finally trending toward similar awareness below it, too.

But that shift still hasn’t shielded Karl and Morey from consistent criticism. The 2013 Coach of the Year is out of a job, after all, and some league followers remain lukewarm on the surging Rockets despite league-best play since the calendar year. Why?

The belief that in basketball, real innovation isn’t innovation as we know it.

Coaches tend to want to control the action on the floor. All coaches overuse strategies which give them the illusion of control. Chaos is actually very useful; if you’re on offense and you can create chaos it will lead to a breakdown and you can score. The instinct [for coaches] to try and control the action actually can work against you. I think that is one of the things that makes great coaches in basketball [more] than any other sport: tolerance for creating chaos. – Bill James

There’s no discord with regard to defensive pandemonium. General ball/player pressure and the creation of miscues are universal positives not only because they limit or prohibit the opposition’s chance to score, but for the possibilities that aggression immediately gleans on the other end of the court. “Points off turnovers” is a meaningful statistic for a reason other than the explicitly obvious: The highly efficient transition opportunities that such a quick change in possession often yield.

But we’ve been slow to embrace the effects of overall offensive ‘chaos’ the way we have on defense or fast-breaks. The panelists agreed that line of thinking is flawed.

When I first came to the Rockets we started tracking the success of Jeff Van Gundy’s play-calls. Two-thirds of the way through [the season] we had enough data to say which play-calls are better than most? So we had HORNS, we had this, we had that, and then I said, ‘What’s this one at the top – the most efficient one?’ And [Van Gundy] said, ‘It’s called random. That’s when the play breaks down and we just set a random screen.’ That might tell us something. – Daryl Morey

Letting players simply play – if you have the personnel, of course – is sometimes the best course of action. Basketball is an intuitive game at its core, one rooted in creative spontaneity. Certain types of structured offense, though, clash head-on with those unique tenants. The overarching theme of “In-Game Innovations” was a consensus belief that hands-off coaching is the surest approach to winning games, especially when a team lacks the singular star conventional wisdom says is necessary to become a champion.

I got fired last year. The gimmick was our style works in the regular season but not the playoffs. I don’t think my team could beat anybody playing slow-down, possession basketball in the playoffs or the regular season. The style of playing fast and quicker, I think, will get more contagious… There’s great coaching in the NBA, and playing fast takes out concepts and philosophies of defense; it’s a reactionary game. If you’re trying to score in the first seven to 10 seconds it’s a reactionary game as opposed to a structured game. If you’re going to try to play power basketball against the best big guys in the world and LeBron James and Kevin Durant, in a seven-game series they’re going to figure out how to beat you. – George Karl

In addition to sharing a preferred reliance on field goal attempts at the rim and from beyond the arc, Karl and Morey are two of the league’s biggest proponents of shooting early in the shot-clock. The Nuggets ranked no lower than sixth in pace during Karl’s nine seasons in Denver, and the Rockets have emphasized similar tempo since Morey developed a roster equipped to do so. While we can safely assume that both have prescribed to the belief that fast offense is better offense for many years, Karl and Morey insist that pace should be an even greater priority due to the recent metamorphosis of NBA defensive doctrines.

The Thibodeau/Van Gundy defenses are just too well designed, especially with the new defensive three-second rule – you can’t beat them just playing halfcourt, slow-it-down. You’ve got to get something early before they can get set. – Morey

To put it simply, modern defenses are too big, too fast, and – perhaps most importantly – too smart for teams to successfully subsist on a healthy diet of slow, scripted offense. Teams need to do all they can to prevent them from establishing the game’s tempo. And it’s those realizations that have in part spurred the rise of so-called ‘non-traditional’ lineups that eschew established positions and favor nuanced skill above most other aspects.

Offensive basketball is a game of execution, spacing, and decisions… The teams that run all these plays – going from one side to the other and shooting late in the shot-clock – it’s been proven statistically that’s not efficient offense unless you’re a great defensive team. I just think the game is going to no-position basketball players, quick-decision basketball players, and smart decision-makers. I kind of fell into this [line of thinking with Denver]. We made the Carmelo Anthony trade and got Raymond Felton. .. Then I started playing [Felton and Ty Lawson] together. And it wasn’t because I was this guy that was so smart; I just decided to play my best players. And putting two point guards on the floor has become a fad in the NBA. Now I believe you really want an Andre Iguodala on the team because he can play like a point guard at 6’6’’ and can play the 1, 2, 3, or 4… It’s not that we want to play with [smaller players] on the court; we want good and effective decisions… You want to get into that random area, and who’s going to be the best in the random area? The great athlete, but he’s got to make good decisions and he’s got to understand spacing. – Karl

We think of in-game innovations as a coach managing the nightly landscape of his team: which lineups he’ll play, which mismatch he’ll exploit, and which ways he’ll go about manufacturing the surest means to victory. But basketball is meant to be played freely and coaches are predisposed to micromanagement; something’s got to give. The panelists clearly believe in application of the former over the latter, and in doing so seemingly speak out against the stated goal of the panel.

Is stressing less manufactured innovation actually innovation? Karl, Morey, and company answered with resounding, convincing confirmation.

Follow Jack Winter on twitter.




Jack Winter