Showing posts with label PPWS. Show all posts
Showing posts with label PPWS. Show all posts

Sunday, November 18, 2007

Analysis: Individual Stats after 2 Games

Since Pomeroy hasn't posted individual stats yet on his site (as of this writing), I thought I'd put up some numbers after the 1st two games. Caveats:
  • SMALL SAMPLE SIZE: only 2 games played, so the significance of these stats is sketchy, at best (there's a reason why KenPom doesn't have his posted).
  • METHODOLOGY: I make no warranty on these numbers. I hope I've done all of the calcs correctly, but I've certainly made a math error or two before. I do know that Pomeroy and I get slightly different numbers for team stats, due to differences in how we calculate team possessions. I use the method listed on his website, but he is apparently using something else these days. The differences there are negligible, however, and hopefully that's also true for these individual stats. I did make a change from the original formulae that Dean Oliver published for calculating possessions and ORatings, but I think Pomeroy would have made the same correction (it involves the factor to convert FTs to possessions used).
Offense
## Player Name % Min
% Poss
O Rat PPWS eFG%
OReb% TO% A Rate
21 Sapp, Jessie 65.0%
27.8%
130.7 1.47 75.0%
2.4% 21.9% 19.6%
3 Summers, DaJuan 61.3%
17.7%
98.2 1.08 54.2%
5.1% 22.0% 1.7%
2 Wallace, Jonathan 60.0%
19.1%
154.6 1.44 73.3%
2.6% 6.9% 15.0%
55 Hibbert, Roy 57.5%
34.7%
119.4 1.22 60.0%
19.0% 11.9% 2.5%
33 Ewing, Patrick 56.3%
12.3%
93.8 1.00 50.0%
8.3% 34.5% 5.2%
1 Macklin, Vernon 50.0%
10.8%
132.2 1.00 50.0%
12.5% 0.0% 2.0%
5 Rivers, Jeremiah 48.8%
6.8%
87.3 0.75 37.5%
0.0% 23.9% 3.9%
15 Freeman, Austin 43.8%
22.3%
85.1 0.92 46.2%
3.6% 24.3% 2.7%
4 Wright, Chris 37.5%
29.8%
66.6 0.91 40.0%
4.2% 42.6% 3.1%


Defense
## Player Name
DReb% Bl Rate St Rate
21 Sapp, Jessie
14.3% 0.0% 1.2%
3 Summers, DaJuan
11.1% 2.6% 2.6%
2 Wallace, Jonathan
2.8% 0.0% 2.6%
55 Hibbert, Roy
10.3% 16.8% 1.4%
33 Ewing, Patrick
4.5% 2.9% 4.2%
1 Macklin, Vernon
6.8% 9.7% 1.6%
5 Rivers, Jeremiah
5.2% 0.0% 1.6%
15 Freeman, Austin
9.7% 0.0% 0.0%
4 Wright, Chris
11.3% 0.0% 4.2%

Quick comments:
  • I cut off players with less than 20% of available minutes played (sorry, Messrs. Crawford, Wattad and Jansen).
  • Sapp has been using a lot more possessions than last season (27.8 vs. 18.7), but thanks to his hot outside shooting (5-9 3pt), this has been a good thing. He may be making his case for taking up the slack from Jeff Green's departure. He won't continue to shoot this well, but his turnovers should come down a bit as well. He's also leading the team in defensive rebounding, which won't keep up.
  • Summers is starting off the season a bit like last season, making turnovers without assists to go along. Otherwise, he's not using as many possessions as last year, but is otherwise playing just as you'd expect. Frankly, I'm most interested in seeing his defense improve this season.
  • Wallace can't keep this up. It's just not possible. It's not his shooting (6-10 3pt, 2-5 2pt) that is so extraordinary, but his lack of turnovers that have his ORating so high. Last season, Wallace turned it over on 24.8% of the possessions he used, worst on the team other than Rivers. Of the ~15 possessions that ended due to Jon so far this season, only one was from a turnover. I'll be thrilled if he can just keep his TO Rate below 15%, more than double where he's at now.
  • For all those who've clamored for 3 seasons for the guards to pound it inside to Hibbert, you're finally getting your wish. He's shooting 60% from the floor (15-25 2pts.), which is actually down from last year, and he's still making great decisions (only 12% TO Rate), cleaning up the offensive glass and protecting the paint (17% block rate!).
  • Ewing has 3 turnovers and is 0-2 from 3pt. Otherwise, he's doing exactly what you'd want. I'm not worried, as he's only played 45 minutes (small sample size, people).
  • Warning: TVBP (thinly-veiled brag post). I said last season that Macklin is making the case for more playing time with his tempo-free offensive stats. Good old Vern is not letting me down with no turnovers, strong offensive rebounding and mostly smart shooting (4-7 2pt.; 0-1 3-pt (!?)). If he just leaves the outside shooting to his teammates and pounds the defensive glass a bit harder, he'll be fine. His % Poss needs to come up, but that may be more difficult.
  • Rivers is still struggling with his shooting, but he's taking care of the ball a lot better than last season. He gets his minutes because of his defense anyway (see Gene Smith), and he's only taken 4 shots so far, so there's no reason to worry here.
  • The freshmen are playing like, well, freshmen. Freeman is struggling from the outside (2-8 3-pts.) and Wright is turning the ball over far too much (6 TOs in 14 possessions used). Frankly, I expected no better, and hope that they can continue to get quality minutes so they can develop in time for conference (and post-season) play.

Sunday, April 15, 2007

Analysis: Season Review, Pt. 2

Picking up where we left off in the last post, here are three more sets of stats for the Hoyas players for this season, breaking the season into three chunks: OOC (13 games, Nov. 11 - Dec. 30), BE regular season (16 games, Jan. 6 - Mar. 3) and post-season (8 games, Mar. 8 - 31).


Points per Weighted Shot

PPWS simply divides the points scored by a player by the sum of his field goal attempts and a fraction of his free throw attempts. As the (former) Big Ten Wonk says, PPWS is not the end-all of statistical analysis, but it's easy to calculate and can give an objective indicator of a struggling or surging player.

PPWS = PTS/(FGA + (0.475 x FTA))

A general rule is that a player under 1.0 is having a rough go of it, and a player over 1.3 is doing extremely well. Keep in mind that small sample size can give some curious results.

Player OOC BE Post
Green 1.28 1.17 1.13
Sapp 1.00 1.03 1.06
Wallace 1.32 1.33 1.42
Hibbert 1.34 1.45 1.25
Summers 1.26 1.06 0.97
Ewing 1.34 1.25 1.06
Egerson 1.11 - -
Rivers 1.10 0.71 1.01
Macklin 1.39 1.45 1.22
Crawford 1.10 0.95 1.50
Spann 0.00 1.40 2.00
Dizdarevic 0.00 1.50 2.11
Izzo 0.51 0.00 0.00

First thing that I notice is that the much-maligned guards showed no drop-off in performance as the competition improved, although Sapp's shooting was merely pedestrian, while Wallace was superb. Hibbert and Green slumped a bit in the post-season, but you just couldn't expect Roy to keep up that insane pace from the BE regular season. That stat may indicate just how little competition Hibbert had in the Big East this year (outside of Gray). Summers and Ewing, in their 1st year playing for Georgetown, did struggle to score later in the year. Rivers' offense was a real problem during conference play (he desperately needs a mid-range and long-range shot), but he didn't embarrass himself when it mattered most. And, dare I say, play Vernon Macklin.


Points per Possession Used

Speaking of that old post, I developed a couple of new (to me, at least) statistics then, based on the idea that you can credit - or blame - individual players for possessions used in addition to weighted shots. This is the same stat that was used to calculate possession stats in Part 1, albeit without explanation there.

Poss. Used = FGA + 0.475*FTA + TO - OR

The idea here is that a player ends a possession by shooting, by getting fouled and heading to the line or by turning the ball over. But a player who can get an offensive rebound saves a possession for his team, and should be rewarded. So a new stat, points per possession used [PPP] accounting for all possessions used rather than just shots, is simply:

PPP = PTS/(FGA + 0.475*FTA + TO - OR)

Here we go:

Player OOC BE Post
Green 1.11 1.08 1.15
Sapp 0.95 0.86 1.03
Wallace 1.02 1.12 1.38
Hibbert 1.52 1.66 1.65
Summers 1.08 1.05 0.92
Ewing 1.28 1.25 0.95
Egerson 1.15 - -
Rivers 0.87 0.46 0.67
Macklin 1.35 2.02 1.22
Crawford 0.99 0.87 1.13
Spann 0.00 3.50 2.00
Dizdarevic 0.00 0.75 0.68
Izzo 0.34 0.00 0.00

While this stat gives some interesting results, it tends to bias towards inside players who can accrue offensive rebounds while avoiding turnovers. So, instead of commenting, I'll just go straight to the final stat.


Points + Assists per Possession Used

And this stat is exactly as it reads, although I have come up with a funky acronym (take that Mr. Gasaway).

PAPU = (PTS + ASST)/(FGA + 0.475*FTA + TO - OR)

The idea here is to reward ball-handlers for their extra risk of turnovers that they assume by crediting them for finding teammates in scoring position. Traditionally, that would be the guards, although the sharing philosophy of Georgetown's offensive scheme isn't as heavily weighted that way as most teams (he says without a shred of statistical evidence to back it up).

Player OOC BE Post
Green 1.43 1.34 1.28
Sapp 1.29 1.19 1.39
Wallace 1.32 1.39 1.71
Hibbert 1.67 1.75 1.82
Summers 1.13 1.20 1.08
Ewing 1.52 1.56 1.23
Egerson 1.29 - -
Rivers 1.39 0.71 1.60
Macklin 1.60 2.41 1.22
Crawford 1.14 1.13 1.13
Spann 0.00 3.50 2.00
Dizdarevic 0.00 0.75 0.68
Izzo 0.34 0.00 0.00

I don't have a good enough feel for PAPU yet to make any absolute statements (another post for another day), but it seems that a score above 1.25 is an indicator of solid play.

What I find interesting here is that, for the 6 core players, three played above their season average in the post-season (Hibbert, Wallace and Sapp), while three did not. This is similar to what was shown in the PPWS table above, except that Roy is now being credited for his yeoman's effort on the offensive glass (3.8/g). Other winners are Rivers, although his post-season stat is heavily biased by the 7 assists he recorded mostly in garbage time against Belmont and (dare I say again?) Vernon Macklin.