What's in a bench?

The folks over at 82games.com released a trio of very interesting articles today. And I promise, in making up for my mini-vacay, that I will post a commentary on each of them.

First up, was a look at starter/bench stats across the NBA this season. They accumulated totals for minutes, possessions, points, rebounds, assists, steals, blocks and plus/minus and split it by a team's starting five and the rest of their team. They went on a game-by-game basis so, for example, Manu Ginobili's production is credited to the starters when he starts, and to the bench when he comes off the bench (complicated I know, but I felt the need to clarify).

Does bench production or is, in colour commentator speak, "a deep bench" a good indicator of overall team success? Well, Portland, San Antonio, and Toronto are 2nd, 3rd, and 4th in bench minutes per game (roughly 90-96 minutes per game (2nd parentheses inserted to mention that's out of 240 available minutes... so they're devoting roughly 40% of their floor time to bench players)). So those three teams are doing well, but then again, Seattle is leading the league in bench minutes per game, and they're clearly not doing well. But you probably play your bench a lot when you're losing by 20 by the end of the third in half your games.

Also weakening the "deep bench means more wins" argument are the teams at the bottom of the list when it comes to bench minutes per game. Washington, Phoenix, and Orlando are 2nd, 3rd, and 4th in fewest bench minutes per game (roughly 66-71 minutes per game (2nd parentheses inserted here to mention that's one-third less than the teams that use their bench the most and that's only approx. 27 to 30% of a team's available floor time.

What else, the authors of the article seem to forget about that pesky little thing called context with one comment they make. They say, "Reggie Theus is getting some praise for his handling of the Kings this season, and one characteristic he's brought to his coaching so far is that his team leads the league in fewest bench minutes per game."

Well, you probably don't play your bench very often when your three top players (Kevin Martin, Ron Artest, and the former King Mike Bibby) missed a combined 70 games (46%) of your early season sked. So the point they try to make there is (pretty much) useless.

Also, I think the counting stats (points, rebounds, assists, etc...) aren't all that helpful because teams rely on their bench for different things (depending on your starting five). When Ben Gordon is coming off the bench, you're going to depend on him (and in turn, "the bench") for scoring. But some teams just want their bench players to come in and do a job defending and rebounding the basketball.

That being said, I think (and this shouldn't come as a surprise coming from me), the plus/minus stat for the benches is a valuable tool. Here is a list of the top benches (when sorted by plus/minus):

1. Utah

2. Boston

3. L.A. Lakers

4. San Antonio

5. Detroit

6. Toronto

7. Denver

8. Portland

9. New Orleans

10. Houston

11. Orlando

12. Cleveland

13. Golden State

14. Philadelphia

You have to go all the way down to the 14th ranked bench before you find a team with a losing record. "Wow!" You're saying to yourself. "Chris has really found a great stat. When will he ever stop? Wow, Chris really is awesome." (at least that's what I can only assume you're saying)

But, we've talked about this before, does this really pass "the laugh test?" Does anyone really believe Boston has the second-best bench in the league (Eddie House can't be stopped!)? Well, no, the problem with assigning a plus/minus to any bench player in the NBA is that they play with starters a lot as well, so Eddie House might just be riding Kevin Garnett's coattails into a gaudy plus/minus ranking (in fact, I'm sure that's exactly what Eddie is doing... man, what am I hammering Eddie House today???).

That being said, I think Boston's bench--namely James Posey, Tony Allen, Leon Powe, Glen Davis, and sure, even Eddie House--are on the court mainly to defend. And they do it well, and that (defence) isn't a stat that is measured well, so maybe those of us making fun of Boston's bench should reconsider our evaluation based on these rankings.

Or maybe not.