That's all well and good that there are so many players 28 and older who are first-rate contributors in the playoffs. I don't know how that's relevant to Green. Those players were almost all playing at their current level well before turning Green's current age. We're talking about the likelihood of Green progressing from a solid player to a great player. Bringing in the likes of Lebron, Bosh, Carmelo, Wade, Duncan, Ginobili, Ray Allen, etc into the discussion does nothing because they were all better players than Green at a much younger age.
The article Lar33 linked to wasn't claiming that players are through at age 26, just that most are unlikely to get better after that age. Of course there are exceptions but it appears that you're arguing that there are effective players past the age of 28. No one's disagreeing. I don't think anyone is claiming that Green's going to be out of the league at age 28 or marginalized into a 9th man. What are you even going at?
The article's 'analysis' doesn't tell you any such thing. It simply reflects that the productivity numbers on average do not get better after that age.
But it is convoluting two large effects. One, the fact that the 'average' player doesn't even have a career that goes past a few seasons. Two - long term, even the 'stars' tend to have suppressed output simply by the fact that the league reorganizes to put veteran stars on more talented teams.
The first effect means that any player who has an extended career is, by definition, not average. He may have average numbers as a role player that just survives. But that doesn't make him 'average'.
The second effect means that you can't take the declines predicted by the 'average' productivity numbers as truly indicative of the individual's decline in talent. There is a sharing the load effect caused by the reorganization. A most distinctive illustration of that effect is the numbers of all three of Pierce, Allen & Garnett. Their individual productivity numbers dramatically dropped when that team formed. They didn't just suddenly become less talented all at the same time and so quickly.
The league is set up to put most of the top young talent on lesser teams where they have artificially boosted productivity numbers. After a few years in the league, they tend to get reorganized onto stronger teams (either they get traded or others get traded to them). Individually, they will exhibit a step down in productivity in those situations. In the aggregate, this shows up as 'general declines in productivity' for the average player.
The article's numbers are ultimately just a reflection of the aggregation of those multiple factors and tells you almost nothing about the likelihood of an individual player getting better or worse over his career track.
If you want to try to predict how a player will do, you should look at the actual player and consider the changes in context you expect him to be in in the future compared to what he was in in the past.
Probably the biggest factor to look at that may or may not change will be USG%. Green's effectiveness at getting the ball in the hoop compares favorably with other 'power small forwards' - his eFG% is typically just over 50% and that compares with guys like Carmelo, Pierce, Gay, George, etc. eFG% is a stat that tends to stay fairly consistent with players over time, after they have a couple thousand shot attempts under their belt. That is one of the true 'he is who he is' sort of stats. Yes, it will vary, but not by any tremendous amount. It tends to reflect the profile of shot types they take.
However, Green's USG% has typically been much lower, for a wide variety of reasons. Most of those guys I just named have typically had USG% rates no lower than 26%. Some were quite a bit higher. Green's career high was last year at just 22%. Most of his career it has been below 20%. It's not surprising that the one year it popped up to 22, so did his various productivity numbers and even the over-revered PER.
So a lot of all this debate boils down to expected USG% of Green. If, in Stevens' system, given the players around him, Green's USG goes up to somewhere above 25%, then it is not unreasonable to expect his productivity numbers to jump accordingly. If his USG stays closer to 20% or lower, then they won't.
Did you even look at the study? It's very brief and the numbers cited are Wages of Wins. Now, I'm not a big proponent of the stat or whatever but it's not built just on the raw averages of a player's career. It doesn't reward a player because he put up bad numbers on a bad team if his efficiency numbers are low. It takes into account efficiency, although it seems to have a problem with a minimum baseline of usage and minutes. But the point is, the stat does not simply say: Great player with lesser averages on a great team, thus a declining player. It says no such thing. Your argument is completely off base.
You are also hung up on the brevity of the "marginal" player's career and how that distorts the study. It doesn't distort the study because if the "marginal" player is out of the league by age 26, then he can't affect the numbers afterwards. If you think marginal players are out of the league by age 26, leaving the true survivors in the data set, isn't that relevant to Green's possible career progression?
I never said anything about 'bad players' putting up inflated numbers on bad teams. I said, specifically, _good_ players putting up inflated numbers on bad teams. I even referred to the category as 'top young talent on lesser teams'. Perhaps you are the one who isn't reading things carefully.
WOW is most definitely influenced by USG and good players get more USG with a lesser surrounding cast.
My disagreement is not with the WOW stat or the study in the paper. As I said all along, it is a reflection of an aggregate affect. My disagreement is with trying to use it as any sort of predictive model for an individual. It is not useful for that at all.
I don't get why Harden is relevant to Green. Harden was a unrecognized star before he got to Houston because his per-minutes production was outstanding. His stats from PER, True shooting percentage, eFG, assist percentage, Offensive-Defensive rating split and WS/48 were elite. None of Green's numbers have ever been elite.
I'm not sure why you are bringing Harden up in this comment because in my comments above to which you are responding, I made no reference to Harden. Harden's phenomenal scoring efficiency (points per FGA) derives primarily from his ability to get to the line at a ridiculous rate. Ultimately, he and Green are different players - different positions, different roles on a team, providing different types of value.
I did comment elsewhere to note that Harden's volume increased in accordance with his increased USG. His USG% jumped by 38% (from 21 to 29) and his scoring per 36 jumped 26% (from 19.3 to 24.4).
What are you even saying about Green's increased usage? You wrote about how having more opportunities and thus putting up better averages doesn't mean a player is better. Are you saying Green will be that kind of player? Based on his overall stat line, Green's more likely to fall into the category of a player overrated by raw averages.
To your last comments, on the results of Green getting more USG: One of the more consistent attributes of a player is eFG%, because it reflects both the shot-type distribution of the player ('at rim', jumpers, 3PT) and his efficiency at making them. Once a player gets a few thousand shots under his belt, his eFG% tends to stay pretty constant.
As I pointed out, Green has actually been very consistently a solidly efficient shooter from the floor, as measured by eFG%, in comparison to other SFs. His lifetime eFG% is a solid 49% and even much better if you drop his rookie season (44%). Since he has been with the Celtics, his eFG% has consistently been over 50%. For comparisons, Pierce' career eFG% is 49.9%.
Also, while Green is not as point-efficient per shot as someone like Harden (because he historically hasn't gotten to the line at anywhere near the same rate) he's actually been fairly efficient at generating points off his shots compared to the average SF. His career PTs/FGA is modestly good at 1.2, but it has gone up each of the last three seasons. Over his last two seasons, it is 1.25. Since he has been a Celtic, it is 1.28 and last year it was 1.29, which is excellent. Both Pierce and Carmelo were at 1.30. And as I mentioned elsewhere, in the playoffs this last season, he rated at a fantastic 1.44 per FGA. Small sample, certainly. But the trends are all very positive.
So, based on shooting efficiency and point-per-shot generation, Green is overall, a shot-efficient scorer.
You keep citing PER and points per minutes - but neither of those alone are complete indicators of scoring efficiency. Both are USG sensitive. You can be very efficient at your shooting, but if you don't take enough shots per minute, both those numbers will be suppressed. PER has a lot of other flaws, but I'll leave that alone for now.
To post high per-minute scoring, you need to be both shot-efficient AND to have high enough USG%.
Green IS shot efficient. But due to the nature of the teams he has been on, he has never had high USG rates.
It is fair to assert that maybe he was not aggressive enough at taking his own shots. And maybe that will prove an innate flaw.
But one can look no further than the aforementioned Harden to see that USG% rates are, indeed influenced by context. There are, of course, many other examples that show that player USG% rates are very much contextual. Ray Allen's USG% dropped from 29.5% in 2006-07 to 21.6% in 2007-08.
So USG% very clearly is extremely sensitive to the team context around a player.
The context around Green is in flux. We can't be certain whether he will get an increase in USG% in this new situation. But if he does, it is reasonable to project, based on his shot efficiency trends, that his scoring will increase (without necessarily needing to play 45 mpg).
Whether that will be enough to raise him to the level of a 'star' player or not remains to be seen.
