On/Off +/- Numbers for OKC-HOU series

It's a stupid stat......theres no reason for it....it tells you almost nothing.
Another stupid stat is Assists....Beverly comes down the floor, makes a great pass to Asik and asik misses the lay in....NO dime, and whos fault...its not a live all stat

 

It is because when he is off the court, the team is overall doing BETTER than when he is on the court.

To prove worth ideally you'd like to be + when on the court, and the team is - when you are off the court.

It proves that your team scores more points while you are on the court, and the other team scores more points while you are off the court.

Hope that makes sense.

 

I think, disregarding the super obvious stuff like LOL GREG SMITH and such, that one of the more interesting points is to compare to their season stats and wonder what changed. 82games has the season stats, though I think they are on a slightly different basis than these.

For a major example, look at Delfino. He was the team leader in +/- for the season, and that met the eye test - though he did get minutes against opposing benches and thus wasn't the team MVP, he was a critical piece in our playing well and allowed us to win the bench battle with regularity.

Since his elbow injury, he hasn't looked like the same guy, and dipped pretty steeply in these playoffs from his season WS or PER. Get the same level of performance from his as we did in the regular season, and it's a different-looking series.

I will say, I'm surprised that Perkins lineups so thoroughly killed Ibaka lineups. I was snickering at Brooks all series for playing Perk over Collison, guess I've been listening to Bill Simmons too much.

 

They increased the variance doe.

 

Pretty much Rockets starters held their own besides Smith and Lin. Rockets bench got outplayed. DEREK FISHER led their team in FG% AND 3 point %, all that needs to be said. (And Smith and Lin was part of that bench later, too)

A hidden ingredient to OKC winning in game 6 was Perkins being out of the game TOO early after 4 minutes and bringing in Collison. It forced Brooks into making a good coaching move. (And also Kevin Martin finally showed up, finally offset any advantage Rockets had)

I made a thread about preferring pre or post trade roster for the playoffs. Rockets wouldnt have Garcia and 95% no Brooks without the trade. They proved to be decent veteran help in a pinch.

But would still have Cole Aldrich who's not that good but much better defensively as a true center than Greg Smith. (He'd be playing his old team that'd be a motivator but you don't know that at the trade deadline). Patterson could play spots minutes at center (still better than Greg Smith there). Marcus Morris could be the 3/4, let Parsons handle Durant. Oh well.

 

you can't tell that right away?

 

Awesome so from this I can gather that Collison was 5 times more valuable than Durant on the Thunder.


BRILLLLLLLIANT

 

+/- is a useless stat except for seasons of play. For such a sample size, it doesn't mean anything.

 

See, you're trying to justify why you're seeing something you don't want to see. In the playoffs, there are no Bobcats, Magic, or Kings to play against. You're not catching a decent team like the Hawks or Pacers on a back to back. Your opponent every night is as well rested as you are, they have ample time to study the tape to take away strengths and they are guaranteed to be a great opponent.

+/- in the playoffs shows what it needs to show. Weaknesses are exposed and Lin & Smith during their time were terrible.

 

Morey, doesn't look at these stats... thinks they're useless.

But maybe you could bring it to his attention and get him to change his mind.

 

It doesn't matter if you are playing the same team. With a sample-size as small as 6 games one game (or one series of plays) can have a huge effect in skewing the numbers.

For instance, using Lin as an example, Most of his minutes were in the first 2 games when Westbrook was playing. The first game in particular would greatly skew his +/- negatively.

Not saying he played well. I think quite the opposite actually, but Lin detractors trying to use this as proof that Lin is terrible are using pretty flawed data.

Greg Smith on the other hand was completely worthless.

 

But it could also mean that the opponent suddenly got hot, hitting anything they threw up while his teammates suddenly got cold even though they kept getting good looks. Or some BS called by the refs totally changed the momentum while he was sitting.

 

I dont need stats to tell me that Greg Smith could not stop Charles Jones from scoring.

 

Stop trying to correct them. It makes it easier to see who does and who doesn't know what they're talking about.

 

or it proves that your team has better bench players than the opposing team?

 

I know you are trying to put it in context. I am just trying to say that the "small sample size" complaint was not without merit.

For example, Harden played brilliantly in game 5. He didn't do very well in all the other games. Should we look at that one game or the other 5 games as the expected norm for him? Or is it the average? If we had lost game 4, we would never have witnessed the great game he played in game 5. Then people would say that he choked under playoffs pressure or overrated or whatever.

Actually, I have always advocated some kind of +/- as the truest measure of a player's worth to the team. But it is awfully difficult to sort out noises.

 

Or that Jeremy Lin played worse than Rafer Alston.