Dat2U wrote:payitforward wrote:Dat2U wrote:If you go by what he did at the end of the year and the playoffs, Wall looked like a top 10 player. Realistically when healthy he's probably about 15. Last year overall he wasn't in the top 50.
This is getting ridiculous.
He most certainly did not look like a top 10 player in the playoffs. 
The idea that "when healthy he's probably about 15" is just plain bizarre. In his very best year -- much better than his 2d best year & utterly eclipsing any other year in his career -- he still wasn't as good a PG as Chris Paul, James Harden, Russell Westbrook, Mike Conley, Steph Curry or Isaiah Thomas. Or Kyle Lowry for that matter. 
Oh what blasphemy! But if I gave you the numbers without the names, that's how you'd rate them too. & no matter what way you calculate it, no matter whether it's PER or the dreaded WP48 or absolutely any other metric, it always comes out the same. 
& if we composite any 3-year sequence for him vs. any of those point guards, it comes out the same. 
As to last year, overall he wasn't in the top 
150. Look at the d@#ned numbers! You really think John Wall can be playing with significant injury & he's in the top 10% of players in the league? 
How about with his hand tied behind his back? Would that push him down to just above average?
When he's healthy & in his best year, he can crack the top 25 players in the league. Just. Has he shown that he can crack it two years in a row? No. How about 2 years out of 3? No.
For saying these things, I suppose I'll hear that I'm "a John Wall hater." Only, no, I'm not -- he is a terrific player. & in any given moment he can look other-worldly great. But, putting him on a higher level than he deserves does the guy no favors.
 
It's your method of evaluation vs. mines. Frankly I'm not a fan of solely using production measurement roll ups to determine a player's actual impact. So much is missed, especially in terms of a player's gravity, how much attention defenses pay to a particular player which frees up room for his teammates and definitively on the defensive end where fans of productivity measurements offer zero input outside of defensive rebounding, steals & blocks
 
The issues you mention affect events in the game where one player's activities -- that do not leave a stat -- upgrade (or downgrade I suppose...) the stats of another player. 
The implication is that these areas are significant in team success but they give credit (in the form of positive stats) to the wrong person (or too much credit to that person & none to a guy who deserves some & may actually deserves the lion's share).
There are 3 ways to respond to your critique.
1. No roll up is perfect. You don't compare them to "reality." That won't work. You compare them 
to each other in order to discover which one correlates best w/ a specific result -- wins, in this case. They are tools; that's all. If one correlates better than another, use it! 
2. You use statistical regression to determine as closely as you can how much effect these areas of interaction actually have, so that you can weight individual activities reliably based on their actual effects.
3. You track the numbers players put up as they move from one lineup to another, or move through roster changes, or move from team to team. For example, Paul Pierce played one year with John Wall. Was he a very different player from the one he was the year before? Did he do significantly different things, or do things in significantly different proportions, based on what another guy (in this case John Wall) made possible? Or, vice versa, when we look at what John Wall did that year, do we see some "gravity" exerted on his numbers by a "Paul Pierce effect?"
Obviously, the people who come up with these roll up tools understand point 1, & do all the things mentioned in points 2 & 3.
As to my example, Wall & Pierce the single year they played together vs. the many years they played apart, there is absolutely nothing in either guy's numbers that would make you think either of them made the other one either better or worse. 
Of course, it's convenient to have something that can't be quantified, that is just a matter of the eye test applied by someone who "knows basketball," which one can use to dismiss critiques of claims one wants to make -- e.g. that John Wall is one of the top 15 players in the NBA. It's especially convenient, because it features nothing at all that can be proven, nothing that can be demonstrated, & therefore no argument against the claim is valid on its own terms. 
For example, lets say a person points out that in 2016-17, John's best season by far, only 2 guys in the entire league who played meaningful minutes turned the ball over more often than he did, & this is one reason why even that year he wasn't in the top 10 or 15 (tho he was in the top 25) players in the league. There is no reason for you to respond. Who cares what "stat jockeys" think? 
It's as if those turnovers didn't really happen at all, John didn't take the ball out of the hands of his team & put it in the hands of the opposing team more often than just about anyone in the league. That's just a number in a table, a "stat." It's not about basketball at all, in fact it's for people who are fans of numbers instead of basketball fans.
Edit -- the above way too argumentative, Dat. Apologies. Obviously statistical analysis is useful; obviously it doesn't capture everything.