I guess to summarize:

Agree that offensive efficiency should be based on a player’s outcome as compared with an average teammate.

Can someone tell me what constitutes an individual possession? I’ve tried looking for it in a few places but can’t find a definition.

Re: shot difficulty — obviously a layup counts the same as a triple teamed long jumper so one should always shoot an open layup. But my question is — why and how did the player get a wide-open layup? In the case of someone like David Lee, the reason he shoots such a high percentage is because, yes, he gets offensive rebounds, tips, and putbacks, but mostly because he makes himself available for uncontested shots off drive and dish, and also because he runs the pick and roll well with Duhon or Nate. But what would happen to those easy shots if it were, say, Mardy Collins, who can’t shoot anything past 4 feet running the point? What would likely happen is that Collins’s defender would play off him, sag into the lane obstructing passing lanes, and prevent Lee from getting his open shot. This same concept is exactly why the entire Suns team under D’Antoni fell apart when Nash sat down. And why this year Shawn Marion is shooting a TS 7-10% less than it was 2004-2007 under D’Antoni and with Nash. Has Marion suddenly lost all his offensive ability and become a worse-than-average offensive player? Did Elton Brand completely forget how to play basketball after coming to Philly from the Clippers? Probably not, but the stats tell a different story.

My point is that teammates and systems matter, and that TS%, and really many statistics are highly confounded by them. And why looking at stats blindly, or even just 90-95% like Owen would say, to tell you how good a player will be in particular system is not valid.

That is in fact part of the logic behind the WOW.

One answer to your question is scoring volume. The difference between a Landry and a Stoudemire is that the latter scores 5 points more per 36 this year and scored 10 points more last year. That is a huge difference. And that encapsulates some of what you are talking about re different roles on the team. When Amare can score 25 points per 36 on a 60+ ts%, that is incredibly impressive.

That said, I do think Carl Landry is an excellent player and I find it curious that he can’t get more time and that no one made him a better offer last year. Every stat you can find, including adjusted +/- will tell you that he was very effective last year. It’s pretty obvious when you watch him play as well I think.

1. Brent Barry .6063
2. Amare Stoudemire .6046
3. Steve Nash .6029
4. Dwight Howard .5958
5. Yao Ming .5935
6. Manu Ginobili .5906
7. Shaquille O’Neal .5840
8. Peja Stojakovic .5829
9. Dirk Nowitzki .5819
10. Carlos Boozer .5775
11. Corey Maggette .5756
12. Chauncey Billups .5745
13. Andrei Kirilenko .5737
14. Mike Miller .5732
15. Ray Allen .5724

As you can see all different types of players from 3pt specialists to low post centers are represented.

What makes Lebron’s 59% TS% better than Landry’s 63% is not based on position but because LeBron helps his team more by using more possessions. So while the first 15 possessions that Landry uses are more efficient than the first 15 possessions LeBron uses the fact that LeBron uses another 15 possessions instead of delegating those to his less effficient teammates makes LeBron a much better player.

A players true offensive worth is based on both the amount of possessions a player uses and the efficiency in which he uses them. Every possession a player uses that is more efficient than his average teammate is a well used possession. The further above average that efficiency is, the more helpful each used possession is. So while every possession Landry uses is better than the possessions LeBron uses the fact that LeBron uses twice as many possessions while still maintaining a well above average TS% makes him a much better player.

As for shot difficulty it does not matter. The fact that Nate’s shots are so much more difficult than the average NBA players’ shots does not help the team. The only thing that actually mattters is the bottom line and a wide open layup is worth the same as a triple-teamed fade away twenty footer. So while the difficulty might speak to Nate’s skill it does not mean anything to the results.

I know that is a bit conveluted but I hope I was clear.

the degree of difficulty argument makes no sense… Shaq’s shots ARE easier than Nate’s, which is the main reason his TS% is so high. That and the pretty jump hook :)
Saying efficiency #s don’t reflect the complexity of play, varying types of shots, etc. is like saying the score doesn’t reflect how the team played. Sure – sometimes you play better in a loss, than you do in a win… but does it follow that won-loss doesn’t reflect how good the team is?
And I have never run across ANYONE who doesn’t see a difference between easy shots and tough shots. Why certain people bring up that strawman is beyond me.

What I’m saying is that it is difficult to compare two players with wildly different shot difficulty and role in the offense with exactly the same statistic. For instance– Carl Landry and Amare Stoudemire currently have exactly the same TS% (0.002 apart). Does that mean that if you had them both on your team you’d split the shots evenly between the two since they are equally efficient? So all I’m trying to say is that HOW you get to a particular TS% greatly changes how the TS% should be interpreted. David Lee having a TS% of 65% 2 years ago is worlds less impressive than Steve Nash having the same TS% 3 years running 04-06. David Lee’s TS% of 65% 2 years ago is way LESS impressive than Lebron’s TS% of 59% this year. At least to me it is. Like I mentioned above somewhere in another post, perhaps ratio of a certain player’s TS% to the average TS% of his particular position or role would be much more useful. That way you statistically realize that Chris Paul really is a better offensive player than Carl Landry even though his TS% is worse.

So I maintain that when you are trying to rate a PLAYER, not a TEAM as you are in your example about playing better in losses than wins, you HAVE to see what goes into the genesis of whatever stat you are using — otherwise you see David Lee with TS or 65 and think he is a better player than and not trade-able for one Kobe Bryant, which a certain long-time poster with a name starting with an O and ending with an “wen” not so facetiously suggested last year.

It’s like Zeno’s paradox argument against stats. Motion doesn’t exist because before you can get from A to B, you have to go half that distance. And half that distance. And half that distance… infinitely.

I always had a problem with Zeno’s paradox because it presupposes the existence of infinity. Infinity is an abstract theory that cannot be proven. Any argument for the existence of infinity that relies on numerical argument is fundamentally flaw because numbers do not exist.

Numbers are the only way to even explain the infinity as numbers can go on forever. But the only reason they can do that is because they do not exist. They are an abstract concept.

For example: You are given a liter of fluid to drink, with the instruction that you must drink it all but you may only drink half of the fluid at a time. Applying Zeno’s paradox you drink 1/2 of a liter, then 1/2 of the remaining fluid leave 1/4 of the original amount, then 1/8, 1/16th, 1/32..ect. The theory is that you would never run out of fluid because there will half of the fluid will always remain.

I disagree. That may work on paper if you use numbers because they are abstract. But 1 liter of fluid is not abstract. at some point in that fluid exercise you will get down to 1 molecule of the fluid-let say the fluid is water and there for the remaining molecule of water is H2O. What happen when you split that last molecule of water in half? It ceases to be water and hence you have consumed all the fluid even though numerically that should not have been able to occur.

Infinity only exists in numbers. But anything that is corporeal or has mass is finite. Zeno’s paradox fails because if you are able to identify two points-A and B, then the distance between those two points is finite. There is an infinite amount of numbers between two points, but the space/mass between two points is finite. Therefore the distance can be traveled.

I’ll prove it two you: Take a book-for those of you that own them ? and hold it out in front of you. Call the point the book is point A. If there is a clear path the ground, call the ground point B. Now drop the book. If the book makes it two the floor then it exposes the flaw in Zeno’s paradox. The book could not have gotten from point a to point b without first travelling half the distance between those two points, then half the distance between that point and the floor, then half the distance between that point and the floor, ect. Yet it reached the floor did it not?

If you come to me with a visual observation, I am liable to disagree with it because the stats don’t support it.

If you come to me with a stat and visual observation, I am liable to disagree because the stat and observation has beem impacted by short term issues that do not reflect the player’s skill and potential contribution.”

So, if I may summarize:

Whatever anyone says, you are liable to disagree.

Correct?

I understand and agree with everything you said.

These thing gets repetitive when someone (like me) expresses a point of view and someone counters it with a stat that for one reason or another I believe falls into the 10%-25% category. So to make my point, I need to explain why I feel that’s the case. “Some” people consider that an attack on the stats (which I find indespensible) and others keep coming back with more amd more stats and missing my point. So it gets very repetitive.

Saying efficiency #s don’t reflect the complexity of play, varying types of shots, etc. is like saying the score doesn’t reflect how the team played. Sure – sometimes you play better in a loss, than you do in a win… but does it follow that won-loss doesn’t reflect how good the team is?

And I have never run across ANYONE who doesn’t see a difference between easy shots and tough shots. Why certain people bring up that strawman is beyond me.

Owen,
I’ll try to explain.
… (endless conversation) …

It’s like Zeno’s paradox argument against stats. Motion doesn’t exist because before you can get from A to B, you have to go half that distance. And half that distance. And half that distance… infinitely.
So eFG isn’t valid because all two point shots aren’t equal. And if you parse it down to between close, inside, and jumpshot, they still aren’t equal because there are different variations between an open 8 footer and a covered 20 footer. And even if you parse it to the covered 20 footers they’re not all equal because not all of them will be the same (covered by Nate Robinson or Dwight Howard?) … infinitely.
However motion exists because over the course of those infinite points the distance is miniscule. Same thing with something like eFG. Over the course of a few hundred shots, the variation in difficulty evens out, so they don’t really matter.
I could argue ad infinitum that your observational methods are flawed. Are you re-watching the games in slow motion? Do you know what plays the Knicks are running? How can you tell if a player went out drinking the night before and didn’t get a good night’s sleep? How slippery is the ball? What’s the temperature in the gym? Maybe on the nights you saw Lee, the ball was slippery in a cold gym and he knew those were lower percentage shots?
Any college freshman taking a psych or forensics class can tell you that human memory is one of the least reliable forms of evidence. That’s why we use numbers, because they are an objective source of information. Relying on observational skills helps fill in the gaps between the numbers, but using them as a source is largely faulty.

I don’t think you get me yet.

1. A wide open short jumper is NOT the same as a 15-18 foot jumper with the best defender on the opposing team covering you, but some players shoot way more of those tough ones than others. It doesn’t all even out. The eFG% stats don’t tell you that. That’s a flaw. So if you want to try to improve your understanding, you can use observation. At a minimum, you’ll be able to indentify the extremes and come to a better conclusion.

2. Without question, memory is quite flawed and also subject to emotion etc…. As since it is, using hard objective stats to supplement visual perceptions is a necessary part of any analysis of a player or team.

I am not criticizing stats and never have. I have often criticized a specific stat because I didn’t think it reflected reality.

I am advocating the use of stats, visual observation, and a subjective analysis of details that might have had an impact on the stats and observation (like short term injuries, trades, system changes, player development etc…) in order to address the weaknesses of any individual method and attain the most comprehensive and accurate view of a player or team.

If you come to me with just a stat, I can see it too, but I am liable to disagree with it because of what I see.

If you come to me with a visual observation, I am liable to disagree with it because the stats don’t support it.

If you come to me with a stat and visual observation, I am liable to disagree because the stat and observation has beem impacted by short term issues that do not reflect the player’s skill and potential contribution.

(by the way, that wasn’t an endless conversation. it was an endless post) ;-)