+/- numbers - how accurate/relevant? (Luis Scola in game vs. Hornets)

What he wrote was that Marion was the best player on the Suns in the year Nash won MVP and Amare was out. He never said Marion was the best player in the league.

 

I think what you're describing is that a player's +/- is dependent on his teammates, but their impact (as determined by +/-) is also dependent on the player's. That is circular, but you can tease out an "adjusted" +/- using a regression. Many teams have their methods of doing this, including the Rockets and Mavs and more, and the idea is described in some detail in articles at 82games.com. For it to be useful, you really need multiple years of +/- data. One game tells you almost nothing.

 

any stat is meaningless without watching the game. and every stat should be looked at with other ones.

 

It's a natural habit of us game watchers to selectively remember the good things and either ignore or not recognize the bad things.

I thought Landry outplayed him, and the +/- indicates that as well.

Scola was a -6 on the court. Since we won by 11 points, that means we were a +17 when Scola was on the bench.

 

This is exactly right.

 

And Scola played 34 minutes, so that means we outscored the Hornets by 17 in 14 minutes without Scola on the floor. That +17 without Scola is far more interesting to me than the -6 with him. A -6 in 34 minutes could be do to a lot of different things. +17 in only 14 minutes strongly suggests that whoever was replacing Scola on the floor was doing a pretty solid job.

 

does the +/- show who was playing with scola and who was playing with landry? did they run the exact same plays? did they play the same amount of time? did they play against the same defender?

too many unknowns..........

u "thought" landry outplayed scola....

 

What you are not taking into consideration is that Scola was functioning as a Mini-Yao in the 1st half,the others were depending on him to score while they were quiet. Only Wafer and Barry were complementing him on offense.
Ron's ineptitude also contributed to Luis' poor plus minus. When Luis got off the floor the rest of the team got involved which led to a better overall plus minus diff.
If it wasn't for Luis's 1st half effort we wouldn't have even been close to the Hornets by half time,and no stats can disprove that.

 

durvasa, looks like you should whip out that Popcorn game flow thingy so we can further investigate what happened today. That should pretty much explain everything.

 

exactly.

and the same goes to the with- and wothout-____ (plug in a player) win-loss records. unless they play exactly the same opponents with exactly the same players. this ain't tennis.

 

What did I point out that you didn't see? Whether you saw it or not, the Rockets did outscore the Hornets by 17 points when Scola was not on the floor (14 minutes). That's just a fact. The reasons, I would surmise, is that whoever was on the floor, collectively, was doing a fantastic job. That doesn't mean Scola didn't play great himself. Sometimes, the ebb and flow of the game causes runs to happen at certain stages of the game, and I guess Scola just happened to be on the bench through most of that.

 

Did I say anything that would suggest otherwise? I thought Scola played great.

 

That's why I'm saying the plus minus thing isn't accurate all the time.

 

One game means nothing. +/- over a season means something.

Actually over the season, when Scola and Landry played together, Rox loses ground because both are below average defender and we keep giving up layups. When one of them plays with Yao, we are OK.

 

But it was accurate. The Rockets were outscored by 6 with Scola on the floor. It isn't typically an accurate measure of how well the individual played, but it still tells a part of the story that's factual. It's up to people to understand what it's saying, and not draw bad conclusions from it.

 

This is one of the most worthless stats in basketball imo. I say because it has no bearing on what a player is doing to help the team win or lose. It could be garbage time, and the team subs, and 1 team goes on a 8-0, but were down by 30. Even the slug that did nothing will have +8.

 

You could make a similar argument about FG% or any of the traditional box score stats. They're all worthless stats because a guy who only plays in garbage time could take put up big numbers against scrubs.

 

thats why i said stats is overrated. u gotta watch the game to know who is doing well. the stats said mutombo got 3 blocks but we all know how many shots he altered that wasnt even recorded. we all know how he hustled and fought over chandler to grab that rebound. that rebound was much harder to get than a rebund where chris paul missed the 3 pointer and it just fell to artest. does stats tell u that? its still 1 rebound

 

Ok,I hear you. It doesn't tell the whole picture though.