Stupid PER Question

[Wizard Mentor]

The Background:
It has been argued in another thread that Webster's PER would increase by playing with better teammates. Bayless also suggested he'd have more assists had he better teammates in Vegas.

The Conundrum:
It seems like there can only be one successful shot attempt per possession. Ditto for assists. So, doesn't one player's PER decrease another player's PER, other things being equal?

The Scenario:
Last season the Blazers averaged 99.4 pts and gave up an average of 94.1 pts and won 65.9% of their games. If we replayed the season, but kept those numbers, how would we maximize the average PER/48?

 

[Denny Crane]

With better teammates, they would make shots when he passed to them, upping his assists. They'd catch passes slightly off the mark, saving a turnover. They'd pass him the ball in good position to shoot, upping his FG% and efficiency. They'd not pass the ball at his feet which would make it easy for him to turn it over. They'd not waste the shot clock and give him the ball to manufacture a shot with little time to get off a good one.

&c

 

[Boob-No-More]

Simple - playing with better players means more high quality possessions that end in a higher percentage scoring opportunity. If you are surrounded by good players, your team will have fewer turnovers, more assists, more rebounds, more steals, etc. All of which lead to more possessions and more quality scoring chances. PER compensates for pace, but it does not compensate for playing with a bunch of guys who constantly turn the ball over, get abused on the boards and can't make an open shot.

For example, if you play with a bunch of guys who can't hit a wide open shot, you're probably not going to rack up a lot of assists. If you play with a bunch of dead-eye shooters, you will see your assist total increase as they are more likely to make the shot when you pass them the ball. If your assist total goes up, so does your PER.

Similarly, if you play with a great passer/playmaker, you will get more easy scoring opportunities (open shots, easy baskets in transition etc.). Go back and look at the PERS of guys like Shawn Marion and Boris Diaw. They both posted career highs in PER when playing with Steve Nash. And, before you dismiss this by saying, "sure, everybody puts up big numbers in D'Antoni's fast paced system", remember, PER factors in pace. Not only did those guys put up the best raw numbers of their careers, they also produced at their most effecient when playing with Steve Nash.

Good rebounders and good defenders also mean more possessions and more easy scoring opportunities.

Playing next to a superstar who is the focus of the other team's defensive game plan, and draws constant double teams means more unguarded open looks and more easy scoring opportunities for everyone else. Horace Grant, Scottie Pippen, Toni Kukoc all had the highest PERs of their career splaying beside Michael Jordan - and all saw their PER drop by 3 - 4 points upon being traded. John Paxson, who was never a great player, saw his PER drop to single digits when Jordan left to go play baseball. I'm sure there are many other examples of players benefitting from playing with a superstar who diverts defensive pressure and nets them higher quality scoring opportunities.

So, yeah, I do think being surrounded by quality players can help a players PER - especially if that player's role doesn't change dramatically. Obviously, a player's individual stats can benefit by going from a 7th man on a great team to the number 1 scoring option on a crap team. For example, Tony Campbell saw his PER jump fropm 12.5 - 17.1 when he left the Lakers for the expansion Timberwolves. Of course, he also saw a dramtic change in his role as his minutes increased four-fold from 787 to 3164 and his FGA/A skyrocket from 5.5 to 19.3. Obviously, this greatly expanded role allowed him more opportunites to pad his stats. For a player who stays with the same team and plays a similar role, such a huge increase in production is highly unlikely.

BNM

 

[andalusian]

With better teammates, they would make shots when he passed to them, upping his assists. They'd catch passes slightly off the mark, saving a turnover. They'd pass him the ball in good position to shoot, upping his FG% and efficiency. They'd not pass the ball at his feet which would make it easy for him to turn it over. They'd not waste the shot clock and give him the ball to manufacture a shot with little time to get off a good one.

Possibly. But, with better team-mates his usage% might go down, his touches will go down, his shot attempts will go down... etc...

It really depends on the player. If we are talking about a PG that can distribute well - or someone that is is a matchup problem but can not manufacture his own shot (big, strong centers, for example) - better team-mates will likely help his own performance. But, a wing that does not have the handle to create for himself and his team-mates is less likely to see a spike in his production with better teammates. He might even see it go down.

My gut feeling is that until Webster becomes a guy that is a matchup nightmare whenever he gets the ball or can create well - the teammates around him are not going to make much of a difference...

 

[Boob-No-More]

Possibly. But, with better team-mates his usage% might go down, his touches will go down, his shot attempts will go down... etc...

It really depends on the player. If we are talking about a PG that can distribute well - or someone that is is a matchup problem but can not manufacture his own shot (big, strong centers, for example) - better team-mates will likely help his own performance. But, a wing that does not have the handle to create for himself and his team-mates is less likely to see a spike in his production with better teammates. He might even see it go down.

My gut feeling is that until Webster becomes a guy that is a matchup nightmare whenever he gets the ball or can create well - the teammates around him are not going to make much of a difference...

Webster is not good at creating his own shot. As such, he should benefit by playing with teammates that will create opportunites for him and get him more wide open looks and easy scoring chances.

Specifically, I stated that Webster should benefit from playing with Greg Oden and Andre Miller. He has never played with a center (Theo Ratliff, Joel Przybilla, Jamaal Magloire) that demanded double teams the way Oden does. And he has never played with a PG (Telfair, Jack and Blake) that is good at running the break, or creating easy scoring opportunities for their teammates. I think sharing the court with Oden and Miller will help make Webster a better, more productive, more efficient player.

BNM

 

[Minstrel]

The Conundrum:
It seems like there can only be one successful shot attempt per possession. Ditto for assists. So, doesn't one player's PER decrease another player's PER, other things being equal?

If what you're asking is whether PER is a zero-sum game, it isn't. On any given possession, you can improve your PER by making an assist, scoring points, getting a rebound (and also by not committing a turnover when you have the ball). By getting an assist, you are leading to a teammate making a shot. For a given set of players, playing more efficient, effective basketball will tend to improve everyone's PER, to the extent that the production is spread around.

Now, of course, if you are on a team with three great play-makers and scorers, it will be hard to get assists/points which will negatively affect PER. But not as much as it will negatively affect raw stats, because PER is based on efficiency...so even if you don't score a lot, if you score efficiently it will help your PER.

The Scenario:
Last season the Blazers averaged 99.4 pts and gave up an average of 94.1 pts and won 65.9% of their games. If we replayed the season, but kept those numbers, how would we maximize the average PER/48?

That's pretty hard to say. In fact, since PER is drawn from the individual player's percentage of total assists and rebounds and turnovers, you probably can't vary the PER gained from those events if you fix the raw offensive and defensive numbers. However, PER's scoring element is based on efficiency rather than number of points...so conceivably, if the team scored that same 99.4 points per game, but on better shooting (which would require a lowering of pace...fewer shot attempts, but more makes), the average PER/48 would go up.

However, that's not a very good way to evaluate teams. It's a good individual metric, but teams should be looking to maximize points scored and minimize points allowed. Not average PER/48 of its players.

 

[andalusian]

Webster is not good at creating his own shot. As such, he should benefit by playing with teammates that will create opportunites for him and get him more wide open looks and easy scoring chances.

I am not sure I buy this argument generally. Looking at usage% - Batum, playing next to better players last year - had a lower usage% than Webster had - and had a higher PER. Did Batum get a lower usage% because he is not as good a player as Webster or because these better players just had him as a lower-priority? My gut feeling is that playing next to better players that can finish as well as these guys did last year would bring Webster's usage% down - just as it did to Batum... with the result being lower PER - they just did not need to use him that much on offense and the same would happen to Webster - whose long-range shooting percent is roughly equivalent to Batum's - but relies a lot more on jump shots (more than 80% of Webster's shots in his 3rd year were jump shots, Batum being more versatile was lower than 66%).

Next year will be interesting and the real thing that could help Webster, imho - is Andre Miller pushing the ball - because if there is a place where Webster is actually very decent on offense - it is finishing the break since he is mighty fast in a straight line. A combination of Greg Oden gobbling rebounds on defense with Andre Miller pushing the ball could help him - but, before we get too excited about it - let's remember that the key is not Oden, it is Miller. Pryzbilla is and was a better defensive rebounder than Oden and Webster played next to him - what was missing was a PG to push the ball. I just do not see Oden's presence on offense being such a big difference for Webster - he will not get that many more open-long-range shots than Batum got. It does not matter if the defense collapses on the interior to free the long-range shooters because of a big, burly center or a sneaky small-guard going to the rim as happened with Roy.

I will repeat - I hope he blows up - but I do not really think Webster will be a big contributor in the upcoming year.