BY next playoffs

The goal of a basketball game is to win - not to have a large blowout. point differential is important when you come to measure a team against another team and get an idea of what the overall team is doing - but win% gives you a better understanding of the line-ups that the team plays that are working well - because the win% per player gives you a better understanding of how the players play overall - since these players play with different combination from within the same team. point differential between teams does not have these overlapping roster combination that individual +/- stats have to contend with.

Since Batum plays mostly with our stars (Roy/Aldridge) and less with the 2nd unit - his point differential within the team can be misleading for his overall impact - especially when you realize that Travis had to play heavy minutes as the #1 option on offense with the 2nd unit - next to less effective players like Frye and Sergio. The point differential here is not as much of a reflection on Travis as the win% - I would actually argue that the fact that Batum wins less playing mostly with the first unit vs. Travis's win% - despite the fact that Travis plays a lot with a lesser unit - tells me that Travis is offering a more important role at this time for the team. There is a reason you see Travis more at the end of the games with the stars.

You actually see that, in my opinion, because Travis has a very valuable role on this team - as a go-to scorer. He does not work as a starter because the team usually goes to LMA at the start of the game - but he works well in the 4th quarter as a closer because the team usually does not go to LMA at the end of the game - they go to Brandon and Travis.

What is happening is usually that the team plays structured ball at the start of the game - going to LMA. By the end of the game the opposing team makes the in-game adjustments trying to take the Blazers out of their comfort zone - this usually means better, quicker doubles on LMA, attacking the ball-handlers to make the passes into the post harder and the like - and that's where Travis's unique ability to create a shot for himself comes in handy. If a guy can elevate over a defender and take a shot with a good chance of it going in - there is very little you can do. The reality of the situation is that it is easier to take LMA out of his game with good defense because he can not create a shot for himself in tough situations as Travis can.

Travis plays a lot - because he wins us a lot of games thanks to this unique ability. His win% reflects this. When Batum's offense will progress to the point that it is harder to close him out with good defense (as the Rockets did, for example) - you will see him play more - and his win% will reflect it.

 

I agree with your criticism of +/-, but not your claim that win% does a better job of accounting for that. win% is simply +/-, but drops the magnitude and only retains the polarity (+ or -). I don't think unadjusted +/- (either in the usual +/- form or the cut-down win% form) is a very good tool to evaluate players.

There are Adjusted +/- figures that attempt to account for who the player played alongside and who the player played against (reserves tend to play more minutes against other reserves, for example). I'd much rather look at those, if I wanted to use indirect measures (measures that don't use the player's individual statistics).

 

I suspect that win% does a good job of telling you when someone that is bad for the team plays long minutes and when someone that is good for the team plays too little. If there is a correlation between playing time and win% - and will tell you generally speaking if the rotations work.

Where it will do a bad job is when you compare between players on different positions or when you have to ask a question about the use of a specific player only with a lesser unit.

In this case - we have two players playing in the same position - and Batum actually plays a lot less with the lesser unit. The concerns you raise have some legitimate issues - but clearly not in this case. I can take an issue with them when it comes to Sergio and Bayless - because one played regular minutes and the other only occasional minutes - mostly in blowouts - but in the case of Travis and Batum - it just does not seem like a real concern.

Batum played more of his time with the better team-mates and yet his win% is lower. This, right there, tells you that at this point Travis's ability to provide offense was more important and he was more valuable to the team.

Well - if you can provide these numbers and we can look at them and discuss them - I would be more than willing to waste time thinking about them. In the context presented - between Batum and Travis - win% trumps +/- stats.

I have a real hard time figuring out why Joel's win% is so much lower than Oden's for example - I suspect, however, that this is really a case of the big mismatch Oden provides to other teams - be it other team's first or 2nd unit - and Nate's insistence on protecting him from getting additional fouls and limiting his minutes - Nate might have had no problem letting Joel play less effectively when he was in foul trouble and long minutes when he was tired. I do not however find it surprising that Travis's win% is higher than Batum's - Batum was a nice surprise and a great hope for the future - but his role was limited. He was a garbage man on offense from a wing position and his defense was hit (Wade, Prince) or miss (Artest, LBJ).

 

Unfortunately I believe that is a problem with Nate. He will play guys until you trade them away, and sometimes that is the only way to keep him from putting certain talent on the floor.

 

I suspect the same about +/-. I just suspect neither of them provides any precision.

I disagree that win% trumps +/-, even when comparing players on the same team at the same position. In my opinion, they have the same drawbacks and win% has the further drawback of throwing out a lot of the information. Teams play the games to win...the job of players is to do as much as they can to help a team win, not to win or lose themselves. The magnitude of +/- is highly relevant to how much an individual helps.

As for presenting Adjusted +/- stats, I can't seem to find recent ones. Back in 2005-06, 82games.com had them, but they don't seem to have them since then. I personally don't tend to use +/- (adjusted or otherwise, and I consider win% to be part of +/-) because NBA stats people say that you need about three years worth of data before they start to become relatively reliable. I think +/- is an interesting concept but I'm still a bit skeptical about how well it isolates a single player's value. I prefer direct measures, like PER or win shares, augmented by observation (especially for defense).

At least, as far as publicly available metrics go. It's entirely possible that teams have great proprietary systems that use +/-. Roland Ratings are interesting in that they mesh direct measures like PER with +/-.

 

My guess is that +/- is a lot more volatile to blow-outs (either way) that win% - again, neither is perfect - but when you look at the same position within the same team - win% seems a lot less volatile to edge situations - in the specific case of Nic. vs. Travis and who they play with - it makes no sense at all that Travis who plays a lot more with low-% players like Frye and Sergio is over-utilized when his win% is higher than Batum who plays most of his time with our 3 best win% guys (Roy, LMA, Blake).

As I said - there are very good reasons to question both philosophically - but since a blow-out is worth the same weight in the win/lose column as a small-margin victory I would be more interested in win%, it is less volatile to blow-outs and in the specific case discussed - there is every-reason to suspect it provides better feedback given the consistent play of Batum with the 1st unit.

I disagree. What matters is winning the games, not the score margins. You go wins first, score-margins when there is no clear answer. While you throw some information with win% - you also throw out a lot of outliers that sully the big picture...

 

Jack was the point guard at the end of a large number of games last season. And I thought he started at least part of the season.

At any rate, Nate played him in crunch time which allowed Jack to lose at least 4 games for Portland last season. Nate would have never stopped playing Jack in crunch time as long as he was on the team.

As for Sergio Vs Jack Vs Blake: That is like comparing shit with nuts in it versus shit with corn in it. No matter then difference, you are still talking about shit.

 

It's possible he did lose us 4 extra games, but I'd take it over the 8 Sergio would have lost us.

 

Chicken and egg situation here, are we as good a 4th quarter team because Outlaw was in in the 4th or was Outlaw that good because he was there in the 4th...

The reality of the situation is that Outlaw mostly played with the starters in the 4th and we were a good 4th quarter team... this reflects on both sides.

What Nate did is really a smart thing - Outlaw is really effective as a scorer - but his value as a scorer is especially high when he can take these impossible "shit they broke our play let's throw it to Travis" situations - so Nate plays him with the starting unit only at the end of the game when the opposing team has made their adjustment to stop whatever else Portland was doing - while going to the structured offense through LMA at the start of the game.

The funny thing is that Outlaw's stats are better individually (PER) and in team wins (win%) - Batum is a more consistent player (he will not dominate a team and he is not going to kill you with stupid mistakes) and he does not need to score to be effective - these are his advantages.

Travis can win you a game and can force the other team to pay attention to another scorer to make Roy's job easier - he is more valuable at this stage at the end of close games.

I have no doubts in my mind that overall Batum is a better defensive player - especially when it comes to consistency. I suspect that if you need to have one defensive possession you need to stop - Travis will be just as effective as Batum or even more so - but he does not play defense as consistently and will hurt you with some dumb plays.

I hate it when people resort to "my eyes tell me" and "it's correlation without causation". I am more than happy to acknowledge problems with statistics and the fact that they are not the end-all of everything - and I am more than happy to require an eye test in addition to statistics - but the fact of the matter is that we have a pretty capable coaching staff which seems to go with what the statistics tell you. At this point it comes down to "the statistics do not agree with my position - so my eyes are better than the coaching staff".

This coaching staff is very successful with these guys and the statistics do not support your position. When these things align - no offense, I am going with the professional eyes and the stats over the "correlation without causation" catch-phrase.

 

Jack started 3 games at the start of last year. After that he was moved to the bench because Blake deserved it. Nate was right about Blake over Jack last year and he was right about Jack vs. Sergio. They might have both been shit - but one of them was done in a bathroom with proper ventilation and running water - the other was done on the kitchen table during dinner...

I am really not sure why we even bother going down this road again - but even this year - where Sergio was much better than he was last year - Sergio's PER is lower (12.5) than Jack's was last year (13.1) - don't get me started on Sergio's PER last year (under 10), his win-score this year (1.6) is lower than Jack's last year (3.0) and let's not talk about Sergio's win-score last year (it was actually negative - I can't for the life of me think of another case when someone had a negative win-score and people were clamoring to have him replace someone that had a positive win-score).

 

You can argue until you are blue in the face about who is better than who. The fact is, none of them are getting it done. Period. Getting it done is what it's about. Not whose getting it done better than who.

 

So - if you can not become Bill Gates rich it does not matter if you can have a nice house in the suburbs (or an apt. downtown - as you wish) or if you are homeless living under a bridge? Please...

 

No because your example is moot. Bill Gates is getting it done. Your just being a dumb ass.

 

I am clearly missing the point - but could you please specify what is it of mine that is not so smart in the buttocks region?