PPP v. eFG?

by Anything but Gatorade on Feb 8, 2012 9:26 AM EST

This isn't my typical "sky is falling" type post, but rather one that I thought was interesting: In last night's game against Florida, Kentucky (awesomely) held Florida below 40% eFG, which is a really, really good number, right? Yet when I checked out our PPP figures from the game, Florida registered at .98 PPP while Kentucky rang in at 1.32 (REAL GOOD). My question is, does the relatively low number of possessions have to do with why Florida shot so poorly yet seemed to be efficient in terms of PPP? I know KenPom leans a lot more on PPP or adjusted offense when compiling his ratings than he does eFG%, so I'm wondering what's really the better metric?

Tweet Comment 4 comments | 0 recs |

Do you like this story?

Comments

I may be on the path to answering my own question

Wow, the guy who asked this was really profound and insightful! I think this might be the way to explain the two metrics: eFG% is basically made fgs + .5 x made 3s/field goals attempted. This number, then, is influenced by numbers of makes and misses only. PPP, points per possession, is measured first by calculating the number of possessions in a game, which is done like this: Possessions = .96 * (FGA − ORb + TO + (.44 * FTA)) and then you take the number of points scored in the game and divide.

The common figure in both of these numbers is field goal attempts. In the upcoming game against Vandy, Vandy actually shoots a better eFG% but their offensive efficiency is only 1.09 PPP or 109 AdjO (you just multiply PPP by 100, not sure why there’s much of a difference). I think the reason their PPP is higher is because FGA doesn’t take whether the shot was a 3 or not into account – if you make more 3s, your PPP is going to be higher. Florida didn’t shoot well against Kentucky, but they did make 6 3s, which means their PPP figure is going to be higher than it would have been had the Gators made 2s instead.

Basically, you can have a higher PPP even with a low eFG% if you make three point shots, which Florida and Vandy both do. Strangely enough, Kentucky doesn’t rely a lot on 3s (just over 20% of the offense) yet still maintains one of the highest efficiency ratings in the country thanks to making 53% of their two point attempts (FGM – 3pt FGM)/(FGA-3pt FGA). 53% is a pretty good number, which probably accounts for the excellent efficiency.

The problem with quotations on the internet is, you don't know whether they're accurate. - Abraham Lincoln

by Anything but Gatorade on Feb 9, 2012 1:34 PM EST reply actions

STAT SHEET IS WRONG!

They didn’t calculate the possessions figure right! Using the formula above, UF had 65.28 possessions (how is that possible…whatever) and scored 58 points, so their PPP was .88. Much better! Meanwhile, Kentucky’s was 1.31, which is pretty awesome.

The PPP formula makes it unlikely for both teams to have the same number of possessions. Florida, using the Possessions formula, had about 5 more possessions than Kentucky did, which is weird, but obviously scored less.

This stream of consciousness FanPost was brought to you by Anything but Gatorade, where neurotica is normal life.

The problem with quotations on the internet is, you don't know whether they're accurate. - Abraham Lincoln

by Anything but Gatorade on Feb 9, 2012 5:02 PM EST reply actions

Statsheet has two possession formulas that they say they use.

But in reality, they use .475 as the “fudge factor”, not .44, and the .96 is not used in college basketball.

The Florida game had 60 possessions. If you run the formula for both teams, one comes to 59.7 and the other to 60.225. Both round to 60.

A Sea of Blue -- Kentucky Sports for the Discerning Fan

by Glenn Logan on Feb 10, 2012 6:30 AM EST up reply actions

Sort of

I’m not sure where you got the possession calculation above, but it’s not quite right. Generally it’s:

(FGA – OffRB) + Turnovers + .475*FTA

but the decimal in place of FTA’s can vary depending on where you check. That figure is calculated for each team and then the two are averaged to get the result for the game. In this case, there were exactly 60 possessions in the game which matches the estimate used by Statsheet and KenPom.

The thing to remember is that a possession doesn’t end until the other team gets the ball back – that means either a defensive rebound, turnover, made field goal, or made free throw. So it is possible to, for example, miss 5 consecutive shots but get the offensive rebound each time and then hit a 6th shot. Your PPP would be 2.0 (or 3.0 if it was a 3 point shot) but your eFG% would be 16.7% (or 25% if it was a 3).

This is essentially what happened with Florida – they missed a ton of shots (low eFG%) but also got a ton of offensive rebounds (41.9% ORB%) which helped them score enough points to have a higher-than-you-would-expect Offensive Efficiency.

Statistics is the part of mathematics that God didn't invent. - Ethan Smith

by JLeverenz on Feb 9, 2012 7:22 PM EST reply actions