Kind of new to APBRmetrics. Hoping for some advice.
How would you be able to tell who is scoring more Efficiently.
Player A: 105 Offensive Rating and Usage % of 30
Player B: 115 Offensive Rating and Usage % of 15
Who Is More Productive
Who Is More Productive
Last edited by zsap on Thu Apr 28, 2011 2:06 pm, edited 1 time in total.
Re: Who Is More Efficient
Player B is more efficient by the terms of Offensive Rating. Player A has more production. If this example is from this season, A might be called inefficient by virtue of being below league average team offensive efficiency (107 and change). In some other systems instead of Offensive Rating, player A "might" be considered to have contributed more offensive value to the team than B but it would depend where it set the break-even point for help vs harm. It would at least be closer due to the large difference in usage .
Re: Who Is More Efficient
How do we know Player A has more production? Is it OFF RATING*USG %
Re: Who Is More Efficient
I was a little too quick and simple in my explanation. Thanks for calling for a clarification.
OFF RATING*USG % would give a quick estimate of 'value of total production'. At least the value of the boxscore-based production. The boxscore gives the specifics of the production.
So OFF RATING*USG % = efficiency * usage of possessions = a value of total production. Some metrics handle usage and this product of efficiency * volume of activity completely. Offensive Rating stops short of doing so. This is a rough way to finish the task that I've used and others probably have too. Maybe it could be examined further and possibly refined.
OFF RATING*USG % would give a quick estimate of 'value of total production'. At least the value of the boxscore-based production. The boxscore gives the specifics of the production.
So OFF RATING*USG % = efficiency * usage of possessions = a value of total production. Some metrics handle usage and this product of efficiency * volume of activity completely. Offensive Rating stops short of doing so. This is a rough way to finish the task that I've used and others probably have too. Maybe it could be examined further and possibly refined.
Last edited by Crow on Thu Apr 28, 2011 3:58 pm, edited 2 times in total.
Re: Who Is More Efficient
100 being average - one could quickly assume an average lineup with the first guy would have a 101.5 offensive rating, with the 2nd guy a 102.25 offensive rating. So, I guess I'd say the 2nd guy.zsap wrote:Kind of new to APBRmetrics. Hoping for some advice.
How would you be able to tell who is scoring more Efficiently.
Player A: 105 Offensive Rating and Usage % of 30
Player B: 115 Offensive Rating and Usage % of 15
Re: Who Is More Efficient
This all gets to usage vs. efficiency.
It's likely that Derrick Rose would be much more efficient if his usage was closer to 20% than 35%. Why doesn't his coach cap his usage at 20%?
Why doesn't Tyson Chandler (~70% TS this season) increase his usage from 15% to 35%?
It's likely that Derrick Rose would be much more efficient if his usage was closer to 20% than 35%. Why doesn't his coach cap his usage at 20%?
Why doesn't Tyson Chandler (~70% TS this season) increase his usage from 15% to 35%?
Re: Who Is More Efficient
I'm just trying to figure out who is benefiting the team the most.EvanZ wrote:This all gets to usage vs. efficiency.
It's likely that Derrick Rose would be much more efficient if his usage was closer to 20% than 35%. Why doesn't his coach cap his usage at 20%?
Why doesn't Tyson Chandler (~70% TS this season) increase his usage from 15% to 35%?
Re: Who Is More Efficient
It sounds like the question is changing every post here. You asked who is scoring more efficiently; usage tells you nothing about efficiency. If all you have is offensive rating, the player with the higher offensive rating is the more efficient player.
Re: Who Is More Productive
The problem I have is that quite often the players I give almost no possessions will do well in it and look better than those who I give more to. I want to see who has the best productivity score. I already know who was more efficient. But Player B isn't as beneficial to the team as Player A, because Player A is scoring more points.
Re: Who Is More Efficient
Sure it does (in response to the bolded part). If there were no relation between usage and efficiency, then Tyson Chandler's usage would (should) be near 35%. Heck, he should take all Dallas' shots with a TS% much higher than even Dirk. So why doesn't he? Is he too modest? Maybe he's too nice and doesn't want to show up his teammates? I mean, I think you need to seriously answer this question before moving on and dismissing the relevance of usage. To do otherwise, is simply to continue (contribute to) an ignorant viewpoint (ala Berri and WP).xkonk wrote:It sounds like the question is changing every post here. You asked who is scoring more efficiently; usage tells you nothing about efficiency. If all you have is offensive rating, the player with the higher offensive rating is the more efficient player.
Re: Who Is More Productive
100 is not fixed as the current league average in Oliver's Offensive Rating. The scale tp my understanding is essentially points produced from all offensive activity per 100 possessions used. The NBA average has been above 100 since the late 70s. The current average NBA team offensive efficiency and the weighted average of the Offensive Ratings of all players in the league might vary some now. When the formula was created I assume the weights were chosen so the team and player averages ended up being real close to each other or matched.
To find rough net value of production for a player, one approach would be player OR - league average OR, whatever that is, maybe close to average of team offensive efficiency, times volume of offensive activity, * usage.
But player leadership on usage and the possible value of certain high usage players to teams remain topics, as Evan noted, that might call for other approaches (perhaps more dynamic approaches) for integrating usage and efficiency data.
To find rough net value of production for a player, one approach would be player OR - league average OR, whatever that is, maybe close to average of team offensive efficiency, times volume of offensive activity, * usage.
But player leadership on usage and the possible value of certain high usage players to teams remain topics, as Evan noted, that might call for other approaches (perhaps more dynamic approaches) for integrating usage and efficiency data.
Last edited by Crow on Thu Apr 28, 2011 4:07 pm, edited 3 times in total.
Re: Who Is More Productive
EvanZ, I would beg to differ. It's semantics, but I think we need to have clear definitions for the terms.
I define efficiency as the scoring rate. The points per shot, or offensive rating, or any other "rate" stat.
On the other hand, production should account for the usage.
In response to the topic itself: in my latest project, I've put together a "game value" metric, that takes simple box-score stats and calculates a value (in points above replacement) for that stat line.
There is a threshold (as ASPM found also) below which additional usage really doesn't help the team at all. I have it as about 0.83 points per possession, where a turnover counts as -.63 points. This scale is similar to offensive rating, so if I just use the offensive rating posited in the initial post, here's what I'd come up with:
ORtg of 115 and usage of 15%: value would be (1.15-0.83)*.15 = 4.8 points / 100 possessions.
ORtg of 105 and usage of 30%: value would be (1.05-0.83)*.3 = 6.6 points / 100 possessions.
(Value is value over replacement, roughly)
This is just a rough value; it should be possible to figure this out more accurately for offensive rating, as indeed has been done on this forum.
Check out Eli W's post here: http://www.countthebasket.com/blog/2008 ... fficiency/
There's also a bunch of stuff in the Google Cache on threads from this site regarding this discussion: (look at the "cached" versions): http://www.google.com/search?q=site%3As ... =firefox-a
I define efficiency as the scoring rate. The points per shot, or offensive rating, or any other "rate" stat.
On the other hand, production should account for the usage.
In response to the topic itself: in my latest project, I've put together a "game value" metric, that takes simple box-score stats and calculates a value (in points above replacement) for that stat line.
There is a threshold (as ASPM found also) below which additional usage really doesn't help the team at all. I have it as about 0.83 points per possession, where a turnover counts as -.63 points. This scale is similar to offensive rating, so if I just use the offensive rating posited in the initial post, here's what I'd come up with:
ORtg of 115 and usage of 15%: value would be (1.15-0.83)*.15 = 4.8 points / 100 possessions.
ORtg of 105 and usage of 30%: value would be (1.05-0.83)*.3 = 6.6 points / 100 possessions.
(Value is value over replacement, roughly)
This is just a rough value; it should be possible to figure this out more accurately for offensive rating, as indeed has been done on this forum.
Check out Eli W's post here: http://www.countthebasket.com/blog/2008 ... fficiency/
There's also a bunch of stuff in the Google Cache on threads from this site regarding this discussion: (look at the "cached" versions): http://www.google.com/search?q=site%3As ... =firefox-a
Re: Who Is More Productive
A number of the threads listed on the first few pages of the link mentioned by DSMok1 have already been recovered and posted on their own (just browser the forum pages) or in the miscellaneous recovered thread, which now has a list of topics by thread page in the first post for easier finding of what you want to read. I grabbed and posted a few more that were missed previously and seemed notable on this general topic.
Last edited by Crow on Thu Apr 28, 2011 4:59 pm, edited 1 time in total.
Re: Who Is More Productive
Am I calculating OFFRTG right. I'm doing (PTS/POSS)*100
Re: Who Is More Productive
For a team, that's right. For a player, it's WAY more complicated.zsap wrote:Am I calculating OFFRTG right. I'm doing (PTS/POSS)*100