'12-'13 PG Pure Point Rating Leaders

[jr lucosa]

I've been trying to find the pure point rating leaders of this past season and haven't been able to find it anywhere, so came up with the numbers for the 30 starting pg's last year. (All may not be starters but I believe played the most minutes at that position for his team.) Pure Point Rating helps show how a player is performing in the traditional PG role of facilitating and limiting turnovers. Here is the definition of pure point rating from Basketball Reference

PPR: Pure Point Rating = 100 * (Lg Poss per 48 min / Tm Poss per 48 min) * (((AST * 2/3) - TOV) / MP). Developed by John Hollinger, PPR is a better "ballhandling" metric than regular assist-to-turnover ratio because it properly weights assists & turnovers relative to each other, and it presents the two in a per-minute form.

So with that I took the league average for possessions per 48 minutes this season which I had at 95.88 and filled in the rest. Here's how the numbers came out for the 30 Point Guards, highest to lowest.

1.12.74 Chris Paul
2. 9.23 Rajon Rondo
3. 8.63 Greivis Vasquez
4. 7.34 Tony Parker
5. 6.70 Deron Williams
6. 6.63 Kyle Lowry
7. 6.39 Goran Dragic
8. 6.28 Ricky Rubio
9. 6.15 Jameer Nelson
10. 5.92 Jeff Teague
11. 5.81 Ty Lawson
12. 5.78 Steve Nash
13. 5.71 John Wall
14. 5.09 Mike Conley jr.
15. 4.84 Brandon Jennings
16. 4.70 George Hill
17. 4.66 Nate Robinson
18. 4.62 Mo Williams
19. 4.53 Russell Westbrook
20. 4.33 Jrue Holiday
21. 4.25 Darren Collison
22. 4.16 Ray Felton
23. 4.00 Kemba Walker
24. 3.90 Steph Curry
25. 3.53 Damian Lillard
26. 3.47 Jeremy Lin
27. 3.26 Isaiah Thomas
28. 3.10 Mario Chalmers
29. 2.10 Kyrie Irving
30. -0.13 Brandon Knight

Note that this is not necessarily a way of deciding who is the better player, but more who is executing the role of traditional or "pure" PG better.

More info on PPR

First, the statistic takes into account that a while an assist is typically worth two points, the passer can only take credit for getting the ball to a shooter who has gotten himself open and then makes the shot. 33% of the credit times two points=.66. This is left to stand as a lesser positive value to turnovers, which statistically cost a team about a point each. Then, the stat is adjusted for both tempo and minutes played, so both offensive style and playing time are eliminated as independent variables. In summary, the stat creates "a single numeric representation of a player's ability to handle the ball and create positive shot opportunities for their teammates,"

Thoughts? Corrections?

[tsherkin]

This is a weighted, per-minute AST:TOV with no nod towards the type of turnovers and no contextual concession to role such as shooting volume. Pure Point isn't a mega-valuable stat because it doesn't really tell us anything we couldn't learn in a second from TOV%.

And again, while Hollinger says he accounts for "offensive style," what he really means is that he's adjusting for pace and has no connection to what kind of role a player is playing on the team. There are systems which generate assists in volume and those which do not; AST:TOV is OK as a peripheral nod, but again, it's not an enlightening stat to any degree.

Look at that list: what did you learn there which you didn't already know?

Now, for giggles, look at that list and match it against TOV% and see what kind of percentage difference you find in the ranking. I bet it won't be all that large. More particularly, it doesn't actually tell you ANYTHING about a player's "ability to handle the ball and create positive shot opportunities for [his] teammates."

It does, however, tell you that the guy is comparatively low-turnover and produces a decent amount of assists... all stuff we can already see rather plainly.

[jr lucosa]

Pure Point isn't a mega-valuable stat because it doesn't really tell us anything we couldn't learn in a second from TOV%.

Sort of. It's an easy way of looking at A:TO ratio with pace adjusted. I'd rather look at one stat than three to get the same information. Mike Conley has nearly the same assist rate as Brandon Jennings but turns the ball over a tad bit more. He's ranked higher than Jennings because pace is taken into account and you can see that by looking at one stat instead of AST% > TOV% > Pace and then trying to do all the math yourself from there.

Look at that list: what did you learn there which you didn't already know?

There are a couple guys who I thought were better facilitators than others, but I did not know the gap was as large as it is near the top. Probably because I never looked at AST%, TOV% and Pace all at once and then tried to create a list out of it.

look at that list and match it against TOV% and see what kind of percentage difference you find in the ranking. I bet it won't be all that large.

Well as I brought up in my first point Jennings is better at keeping TO's down than Conley with the same assist rate yet comes in behind him on the Pure Point list. John Wall is a turnover machine but still tests out near the middle on this list, ahead of a guy like George Hill who has 2/3 the TOV% that Wall does. Isaiah Thomas has a lower TOV% than John Wall and is 14 spots below him.

And again, while Hollinger says he accounts for "offensive style," what he really means is that he's adjusting for pace and has no connection to what kind of role a player is playing on the team.

Yes well this is where using your brain comes in, you probably aren't reading this in the first place or giving a damn about any stat like this if you aren't a pretty big fan in the first place. In today's NBA especially there are ultra athletic guards who are shoot first players really playing the PG because of size only. You as an NBA fan already know who those players are. You know that the Westbrooks of the world are not Pure PG's and therefore will not rate extremely high in a stat that is judging PG's in the traditional sense of getting teammates involved and keeping possession of the ball. If I wanted to judge how a young John Wall or Jeff Teague is doing creating for his teammates and not turning the ball over in comparison to the leagues best while adjusting for pace I can do it all with one stat instead of three as I stated earlier. I can also use this stat to see that a guy like Jameer Nelson had a solid season facilitating for his teammates this year easier than I could by sorting any other 1 stat. I wouldn't use this stat like a mindless monkey and say 'Irving has low PPR, is bad PG'. Obviously you as a person yourself take into account the role the player plays on his team.

[tsherkin]

Sort of. It's an easy way of looking at A:TO ratio with pace adjusted. I'd rather look at one stat than three to get the same information. Mike Conley has nearly the same assist rate as Brandon Jennings but turns the ball over a tad bit more. He's ranked higher than Jennings because pace is taken into account and you can see that by looking at one stat instead of AST% > TOV% > Pace and then trying to do all the math yourself from there.

Right, but again, it's not really effective because it doesn't tell you how the turnovers were related to the playmaking. If the guy is a significant scorer, than that TOV% will be higher and you need to account for that. AST:TOV won't illuminate such things.

Well as I brought up in my first point Jennings is better at keeping TO's down than Conley with the same assist rate yet comes in behind him on the Pure Point list.

He shoots a lot, mainly long jumpers, so a smaller proportion of his possessions end up in turnovers. He passes less and shoots more, so this isn't a surprise.

John Wall is a turnover machine but still tests out near the middle on this list, ahead of a guy like George Hill who has 2/3 the TOV% that Wall does. Isaiah Thomas has a lower TOV% than John Wall and is 14 spots below him.

Sure, but he's also higher up on the raw assist production chain than a lot of the others. Remember, he posted 8.4 AST36 this past season and 16.5 FGA36 as well. So again, you're talking about a guy producing more assists and shooting more often, so that changes his possession distribution and the likelihood of turnovers. You also need to consider which types of plays the player runs more frequently than others, because some are inherently more turnover-prone than others (play types, that is).

. If I wanted to judge how a young John Wall or Jeff Teague is doing creating for his teammates and not turning the ball over in comparison to the leagues best while adjusting for pace I can do it all with one stat instead of three as I stated earlier.

But again, it does a piss-poor job of illuminating anything because it's literally just a look at disconnected numbers. AST/TO isn't a particularly valid stat. You'd be better off trying to look at passing turnovers, play types and the distribution of their turnovers within those play types.

That's what I'm trying to say, this simple stat doesn't tell you anything particularly important because the connection between the two isn't as simple as it is represented within this stat.

Remember, facilitation doesn't have to be based on volume assists, so consideration of the assist stat itself isn't all that valuable depending on who you're examining or in what system they play. Rondo is overrated because of his high volume of assists, while people neglect to consider his pathological aversion to shooting and the presence of scorers and a strong assist-generating system in Boston. Guys who average under 8 apg are often undervalued because they don't produce the sexy box score stats, but someone like Billups was still a very good playmaker even when he wasn't generating assists at that level of production. Someone like Jordan, Pippen, Kobe or Penny were all good playmakers. Look at bigs and you'll often see NEGATIVE AST/TOV ratios, even though some are good.

There are positional factors, role-related issues, style-related issues... This number isn't valuable. You learn more just looking at their raw TOV% than you learn about their AST/TOV ratio.