Sabermetrics come to basketball

At this point in their careers, Jefferson > KG.

 

I was referring to the situation when KG went to Boston, not now (KG's a jump shooter like Malone was at the end).

 

Definitely worth it, I believe. Layden and Ballard are both coming into their own, in my opinion, as tremendous take-out guys. Outstanding mix these days of lengthy take-outs and the typical nuts-and-bolts Peter King-type material that is still interesting and gives you a nice big picture.

 

OK that's what I needed to hear. Can have the wife "suggest" to the kids that they get me SI subscription. 13-year-old son will be pleased when the swimsuit edition arrives unless it features Gabby Sidibe, I mean Kate Upton.

 

Really liked the piece. Ballard's had a great year with stories.

It is kind of funny how so far this year -- and it is early -- Kobe Bryant is actually more efficient than Harden, Mr. Efficient.

http://www.basketball-reference.com/players/b/bryanko01.html

http://www.basketball-reference.com/players/h/hardeja01.html

I have doubts about Harden's superstar ability too - there is a difference between being the No. 3 guy and the guy everyone focuses their defense on -- but at the same time it was a great pickup by the Rockets and they did need someone like him, someone who could be that "foundational type player," even if the jury's out on whether he fills that oddly titled role.

The sabermetrics is different in hoops, it's not as easy as doing it with baseball (and I think we talked a lot about this while discussing that Battier piece). For instance, there are those who use PER as the be-all of stats, which is fine. Until you see something like Magic's best season of PER would be the 81st best of all-time.

Winshares, etc., lots of issues too. David Friedman, a blogger, usually has some interesting responses to advanced stats.

http://20secondtimeout.blogspot.com/2012/03/strengths-and-limitations-of-advanced.html

LTL, yeah worth it, although most of the profiles do I think usually end up in the Vault after a week or two; they just don't put them free online right away.

 

What do the advanced stats folks make of the Lenovo plus-minus? Worthwhile? A couple of years ago when the Warriors had Monta Ellis and Stephen Curry, they also had the hard-nosed C.J. Watson. And no matter how on fire Ellis or Curry was, the plus-minus rankings always said the Warriors were better when one of them was on the bench and Watson was on the floor.

PER has always been somewhat difficult to me because so many players are efficient only because of the limited minutes they're playing in the most beneficial situations (i.e. when other second-stringers are on the floor).

 

Plus/minus is the main stat Morey looks at. He's had an in depth interview before about it. I can't remember where I saw it. It might have been a transcript from one of the MIT meetings a few years back.

Anyways, Morey said he relies on dozens of different stats and they aren't everything but if he had to pick one stat it would be plus/minus. Said it isn't reliable game to game but over the course of a season or two it becomes quite reliable, especially in evaluating bench players. He went on about how it isn't just simply plus/minus because you have to adjust for how good or bad the team is. He said his team compares players' plus/minus to those on that players team and have formula's for adjusting it so every player from all teams are easily compared.

 

More proof that it's the players who matter, not the GMs. The teams with the superstars win. Not hard to figure out who they are. Hard to get them, though.