The Roland rating is a great example of that. Using Battier specifically, it would appear to put him in between to superstars in terms of contribution – and I don’t think anyone really believes that. But the failure of the model is better captured by simply looking at the front page.

The top-13 in +/- come from a total of 3 teams – Cleveland, L.A. and Orlando. Cleveland alone has 6 players in the top-20. Some of these players – such as Ben Wallace – are not good. They’re just buoyed by their strong teammates, as mentioned. And by definition, any player on a good team will likely end up rated higher than great player on a bad team – as the latter will almost certainly be in negative territory.

While I haven’t wrapped my head around everything Rosenbaum did, a quick scan of the lists indicate his adjustments still lead to many peculiar rankings.

So I would wager that a GM that relied on traditional metrics would likely end up making better decisions than one who focused on the Roland ratings(and the various derivatives of it). And I think that’s the big challenge here – as you noted, there’s a long way to go in measuring collaboration in the NBA. But incremental steps in that direction might actually make a team that follows them worse.

But that’s not to say it isn’t worth exploring – if a team manages to make the full “leap” to effectively measuring and comparing collaboration, it will be a definitive competitive advantage…

Visibility into a common, vetted set of statistics (plus raw data) for any player can provide as close to an apples-to-apples market comparison when it comes to identifying and rewarding top individual performers. No such consistency exists for the enterprise; Company A typically does not have relative measures of its PM’s performance compared to his/her counterpart at Company B. More or less blind when it comes to the outside world.

Then of course is the flexibility in the sports world to directly act on these captured individual performance metrics (trades, free agent signings, contract extensions, minor league reassignments). I’m sure many business leaders would love to drop a ‘DNP-CD’ on a poor performer, but there are often clear and necessary barriers to this.

So I think this really starts with establishing good business and team performance metrics – and most enterprises have a lot of work to do there before they can effectively move this down to the individual-level.

But then let’s be honest.. it’s all about the playoffs.