A reliable measure of scoring efficiency

bigbadstevenson · Post #1 » by **bigbadstevenson** » Fri Feb 4, 2011 5:10 am

Here's my idea for a guage of how efficient a player is at scoring in basketball:
Take the number of points a player scores (not including points scored on technical fouls) and divide it by the number of possessions on which the player shoots the ball (a set of free throws is counted as 1 unless the player makes the preceding field goal, in which case it is not counted).

Here's an example:

Player X scores 15 points, going 3 of 6 from 2pt range, 1 of 4 from 3pt range, and 6 of 7 from the free throw line. Of critical importance is the breakdown of the free throws. In this example, the player had 2 trips to the line that were separate from a made basket (for 4 shots), 2 "and ones" and the technical. The "and ones" are adjoined to a made basket, so they are not counted as a separate attempt. So the denominator is 6+4+2. The rating in this example is 15/12 or 1.25.

On its face, this seems to me to be reasonably viable as an indicator of scoring efficiency that could be included in box scores. This is just one phase of my plan to change fantasy basketball into something more reflective of the actual value a player has to his own team. I'm hoping that some of you can discuss the advantages and disadvantages of using this scheme instead of the percentages that are commonly used currently, and the limitations it has for real coaches.

Jimmy76 · Post #2 » by **Jimmy76** » Fri Feb 4, 2011 6:39 am

TS% is what you're looking for, it doesn't do a perfect job of measuring the number of possessions going into a fta but it's so close it isn't a huge deal

It's expressed a little differently but it's the same idea

I agree it should be included in box scores

mysticbb · Post #3 » by **mysticbb** » Fri Feb 4, 2011 7:30 am

Excluding technical free throws doesn't make any sense, because they are not getting a made free throw by default. They can still miss the free throw.

Anyway, as Jimmy76 said, what you are looking for is true shooting percentage: ts% = pts/(2*(fga+0.44*fta))

The factor 0.44 was determined by checking a tons of play-by-plays. Thus he has some merits and works very well as an overall estimation of the used possible scoring possessions.

ronnymac2 · Post #4 » by **ronnymac2** » Fri Feb 4, 2011 10:02 pm

I have a question. Individual offensive rating is somewhat useful if you take USG into account, correct?

Jimmy76 · Post #5 » by **Jimmy76** » Fri Feb 4, 2011 10:20 pm

ronnymac2 wrote:I have a question. Individual offensive rating is somewhat useful if you take USG into account, correct?

Ortg is basically ts%+ast+orb-turnovers

Its an overall offensive efficiency stat not just scoring

mysticbb · Post #6 » by **mysticbb** » Fri Feb 4, 2011 10:57 pm

ronnymac2 wrote:I have a question. Individual offensive rating is somewhat useful if you take USG into account, correct?

Well, it is useful without usage, if you just want to know how efficient a player is using his opportunities. That's what the stats tells you. Usage is basically an indicator for "creation of scoring opportunities". Thus, if you want to know which player is the "best" offensive player, you have to take usage into account.

Post #7 » by **penbeast0** » Fri Feb 4, 2011 10:58 pm

TS% wouldn't be necessary if they counted fouled while shooting as a shot . . . you get the points for the made free throws anyway and the ball is still going up to face defensive rebounding edges. Not sure the rational for not doing this except that box scores traditionally only use the less accurate fg% rather than points per shot which would replace ts% after this (though as the OP brought up, techs do skew this a tiny bit).

ronnymac2 · Post #8 » by **ronnymac2** » Sun Feb 6, 2011 1:28 pm

mysticbb wrote:
ronnymac2 wrote:I have a question. Individual offensive rating is somewhat useful if you take USG into account, correct?

Well, it is useful without usage, if you just want to know how efficient a player is using his opportunities. That's what the stats tells you. Usage is basically an indicator for "creation of scoring opportunities". Thus, if you want to know which player is the "best" offensive player, you have to take usage into account.

Right. That's why while Tyson Chandler and Nene are more efficient than Lebron/Wade/Kobe IN WHAT THEY DO, but they aren't better because you're comparing apples and oranges- the level the players are on is differentiated by USG.

Kobe/Wade/Lebron have much greater responsibility (as at least indicated by USG) as the foundation of their respective offenses. Nene and Chandler are just side pieces that are mega-efficient at being side pieces.

Am I understanding it right?

mysticbb · Post #9 » by **mysticbb** » Sun Feb 6, 2011 1:43 pm

ronnymac2 wrote:Am I understanding it right?

Yes, basically. You can look up offensive adjusted PM ratings to get a better view about the impact. Bryant, Wade or James have a high offensive adjusted PM, while Nene or Tyson Chandler don't have such a thing.

It is basically a discussion about efficiency vs. usage. I have a metric for that in which I combine the amount of "points created" per 100 possessions with the efficiency. The league leaders right now are Nash, James, Williams, Bryant, Paul. Nene is 29th, Tyson Chandler is 97th. Well, if I some up that value over the whole team I get a 100% correlation to the team overall offensive efficiency. That doesn't mean I distributed the "points created" correctly.

SideshowBob · Post #10 » by **SideshowBob** » Sun Feb 6, 2011 6:22 pm

On the other hand, there's three guys that have posted a usage of over 30%, and maintained an offensive rating of over 120.

http://bkref.com/tiny/YM5uj

Basically, these are some of the best offensive seasons ever, ultra high usage level with extremely efficient offense. FWIW I expect James current season to be on that list by April