Great deal Durvasa. Much like Gater said and I'm not sure its possible, but is there any way to chart this stuff against the top 8 teams in the west? Truthfully, I really dont care how a guy plays against a team that is nowhere close to the playoffs. Thats not to say those games dont count because they do, but I'm more concerned about how the players grade out against the top tier teams. Prior to last night, Martin has really struggled vs the Mavs for some reason. I know most of the good teams play good defense, but those are the kind of teams the rockets will play if the make the playoffs. Just like they at split with team like portland last year,but even minus roy,the blazers have done the rockets in. Last year, the rockets also owned the thunder,this year its different. I know every year is different and some players advance at a different rate than other. Of course injuries also play a part,but it seems at least last year, we were beating more than our fair share of playoff caliber teams vs this year.
I've really warmed to APM because while I think the metric is incredibly noisy which makes it difficult to compare players on different teams, a lot of people really need to see things like this. Many just can't separate overall offensive or defensive impact from a few pet stats and tend to overvalue/undervalue specific things that catch their eye. It almost seems like APM seems to align more with what the "basketball people" think than what those who attempt to make assumptions on impact based solely on advanced stats think.
We are not at the point to compare with good teams at this moment. Last year we did better against good teams because we had 2 PGs playing at NBA starter level. This year we only have 1 playing at that level. Just look at the games Lowry went down last season, we did worse in those games than we are doing now. One decent PG can't get the job done in this league. And that is the case for Rockets nowadays.
I'm not sure why you'd want to do that- adjusted plus minus is designed to adjust for EVERY player on the floor, including the opposition, so a better performance against the Lakers or Celtics will necessarily be already reflected in the data set. Second - you seem to be saying that it's basically the playoffs that matter and not the reuglar season. Fair point - but why bother using regular season data to prove playoff worthiness at all? Why not just use playoff data (the answer of course is that the sample set is much smaller and probably more prone to errors). Third - if I'm not mistaken, as a mathematical function - I would assume that most players (all?), from Dwyane Wade to Trevor Ariza - are going to have lower plus/minus vs. good teams, by virtue of the fact that htose teams will naturally have better overall plus/minus ratings - because that's what makes them good. Sure, in the first half of the season, as the season wore on, the Rockets regressed to the mean, finishing 18-27 vs. the top 16. That's better than this year (8-20 so far) but by no means good.
durvasa, does "adjusted" account for opponents? If so, the complaint by GATER and leebigez about opponents is not valid. And since the numbers are adjusted to teammates on the floor, the complaint about Battier and Hayes being "lifted" offensively by players like Martin is not valid either. I think the offensive values of Battier and Hayes are greatly underrated by those who sneer at statistical analysis because if you don't look at deeper numbers, you tend to just notice the obvious. As durvasa pointed out, this is not strictly a measurement for "who is the best player". It is a measurement of how valuable the player is with the team he plays for. And if you have enough sample size, that should give you a picture of how valuable the player is in general.
I don't see how you can take out Martin and then call the other guys can't win without him. Kobe couldn't win withou Shaq/Gasol, does that make him useless? +/- is all about who does better with the current mix of players. You find better bombinations with what you have in the hand. Not assuming you have or don't have somebody. The chart shows Lowry/Battier/Hayes make positive impact on both end of the floor with current group of players. Of course +/- can be totally different if we have a Lebron coming off the bench. But the reality is we don't. Those players are doing better than rest of the players in general.
It does adjust for strength of teammates and opponents. But there are synergistic effects between teammates which undoubtedly are real in basketball that aren't necessarily captured. Similarly, some defensive matchups are more favorable than others, irrespective of "strength" as captured by an opponent's APM. Maybe a player like Lowry excels when he's playing with certain types of offensive players, and really struggles when he's playing with other types. If the coach tends to put him in good situations -- matches him up with players he plays well with for instance -- then that would not only help his +/-, it could also help his adjusted +/-. Or take Chuck Hayes. Chuck's adjusted defensive +/-, while pretty good the last few years, has actually dipped considerably from where it was when he was playing a lot of minutes at PF (his natural position) next to Yao. Does that means his adjusted +/- was overrated back then? Depends on how one interpreted it. Within the role he played backed then, I would say he excelled defensively. But if you change the makeup of the lineup and adjust roles, now Chuck's APM doesn't look as good. And there's a good reason for it -- he's not as good a defensive player when he's tasked with protecting the rim.
The other side of that is Chuck has been a consistent positive, even when surrounded by not so great teammates. That tells me he can play and he makes a positive contribution. This type of analysis on our team as currently constructed without a true superstar lets you know who is getting it done the best. It reflects what we already know. If you take Lowry, Hayes, and Battier off the floor, we would go practically winless.
Nice explanation of the nature and context of the numbers. +/-, adjusted or not, is not a "holy grail" stat or an absolute measure of a player's "goodness" like some sort of video game player rating. To me, they are more useful in terms of raising interesting questions such as why a team tends to suck when a supposed talented player is on the floor, or why a team's offense seem to run more efficiently when a player who rarely takes a shot is playing.
Doesn't adjusted +/- unfairly punish players with good alternatives on their own team? For example when calculating Scola's rating it would look at how his team fares without him on the floor when let's say Miller is out there instead. So both of these players would be unfairly given lower offensive ratings because they both execute so well on offense such that you don't really notice a difference when Scola goes out and Miller comes in. Correct me if I'm wrong.
True. Although if a player has a consistent rating in it, despite changing roles, you could deduces that they are simply good/bad players? For example: Lowry went from backup PG to starting PG, a dramatic change of roles, causing many stats to change like his FT rate, charges drawn etc. However he is still a consistent positive in +/- through out. Jordan Hill went from playing along side Brad Miller to the starting center. He even complained about the role change, preferring the more touches he gets as a backup. However no matter where he plays, the one constant is the team loses with him on the floor. Brooks went from starting PG who is tasked with running to team to hired gun off the bench. Yet his +/- has been below avg to bad the whole time. From this, along with watching the game, I see Lowry as simply a good player, Hill as a bad player. Brooks is a little more complicated. He is an average player when playing the role of starter. However this year coming off the bench he is clearly uncomfortable or unsure about his role. He has not adjusted and is forcing the action. Under the present circumstances he is simply a bad player. This is because he is below average on offense when he forces things, and really below average on defense per usual. If he could ever embrace, adjust to the off the bench role, I suspect his +/- would improve to show an above average bench player who can help the team.
I think that is on court/off court. A separate stat from +/-. +/- is simply how the team fared with you ON the court, nothing to do with your backup.
I think you can deduce that they have found ways to be useful for their team. And if you can consistently find ways to do so, with different teammates and in different roles, it means you are someone a team should want. It doesn't mean that if you throw 5 guys with good +/- on the court, you'd have a world beater. Take an extreme example: a team with 5 Kyle Lowries on the court at the same time probably wouldn't work, but one Lowry next to 4 other guys is pretty good to have. Like durvasa said, you still gotta look for guys with complementary skill sets.
Of course. 5 Dikembe Mutombos wouldn't work either, despite being HOFers and obviously very good players. I think it means that if you can consistently find ways to help your team, with different teammates and different roles, then it means it is not the teammates and roles, but rather you who is producing it. Meaning a good player. I think that's the same thing you are saying anyways. Actually, scratch that. 5 Dikembe Mutombos WOULD work. In fact, they would never lose, 0-0.
It's expressly designed to control for alternatives and filter out their contribution. Accordingly, Scola wouldn't be penalized for having Landry as a replacement vis-a-vis Amare Stoudemire who would check out for Ronnie Turiaf.
I'd like to see a comparison of the "other players on the floor" with Brooks Vs. Lowry. Probably a considerable task to put that together tho ...
Sam,Choujle- I never said only against playoff caliber teams didnt matter, but when you're team building, you always have a target. I really want to know how big of a jump, up or down, do those players take vs the better competition. If the goal is to enter the playoffs and make noise, if all your best players struggle big time against playoff teams, what kind of noise are you really going to make? If all the rockets have are a team full of guys who murder the poor teams and struggle against the good teams,then they're wasting their time. Let me use the example I've used before,luol deng. Posters on this board have talked about him being overrated and overpaid. They've also said he's ineffecient because he doesnt draw fouls and is mostly a mid range jump shooter. If you look at his numbers across the board in both the regular season and playoffs they are the same. It doesnt matter the defensive team,road game,or competition, he's the same. So by hisory I know deng's output is portable. On the other hand, some players production dont convert in those environment. In fact,they decrease alot and those players really hurt their team. I remembet Cliff Robinson being terrible in the playoffs. He was a tough cover during the regular season and took the sorry pill i the playoffs.
