APBRmetrics :: View topic - Scoring Efficiency Ratings

APBRmetrics :: View topic - Scoring Efficiency Ratings - Advancing Hollinger

APBRmetrics
The statistical revolution will not be televised.

FAQ

Memberlist

Usergroups

Profile

Scoring Efficiency Ratings - Advancing Hollinger
Goto page Previous 1, 2, 3, 4 Next

APBRmetrics Forum Index -> General discussion

View previous topic :: View next topic

Author

Message

Ed Küpfer

Joined: 30 Dec 2004
Posts: 783
Location: Toronto

Posted: Mon Jan 24, 2005 11:22 am Post subject:

Nikos: The data you want is at this Yahoo site.

MikeG: I found the error. In row 10398, there is the following record:

Code:

04-11-28 IND SEA harrison,david IND 1 1 Technical 2 9:45

That should read:

Code:

04-11-28 IND SEA harrison,david IND 1 2 Flagrant Type 1 2 9:45

Which means that David Harrison gets one more Flagrant 1 FTA (total = 2) and one fewer technical FTA (total = 0). I've updated the files at Yahoo to reflect the changes, or you can just change it on the copy you have now.
_________________
ed

CrazyFromTheHeat

Joined: 21 Jan 2005
Posts: 31

Posted: Mon Jan 24, 2005 9:26 pm Post subject:

Mike G wrote:

As to the wide range of BFTA/FTA, clearly there are 2 skills at work: drawing the foul, and finishing the shot. Iverson draws fouls and flings up prayers. Shaq tends to score with guys hanging on him.

I agree with this, except that you could say the same thing about other prominent players, and they don't show up on the "low" leaders board. Iverson's total number of attempts really stands out compared to the others on that list.

Mike G wrote:

Shaq might be at the top of the list if his FTA weren't diluted with hackashaq plays. No chance for a bonus shot there.

Good point. In his case, it might be more important to look at total number of BFTAs, where he is second to Stoudamire.

Mike G wrote:

There are quite a range of "ballhandling" fouls still in the mix (undifferentiated from "shooting" fouls.) I suspect these are more significant than differences in "finishing strength".

There's another great study right there - ability to draw fouls. It's all in the play logs, we just need the time and desire to count it. Of course, we could just ask Ed, and he'd probably have it for us tomorrow.
_________________
The Best Miami Heat Coverage
http://heat.mostvaluablenetwork.com/

Kevin Pelton
Site Admin

Joined: 30 Dec 2004
Posts: 970
Location: Seattle

Posted: Mon Jan 24, 2005 10:18 pm Post subject:

To take out the effect of the ballhandling or Hack-a-Shaq fouls, all we need to do is change the denominator from all FTAs to Shooting FTAs, right? Let's see how that changes the leaderboard (I've also reduced minimum FTA to 50):

Code:

Player BFTA SFTA Pct
----------------------------
Jamison 21 98 21.4
JRose 17 81 21.0
Stoudamire 6 29 20.7
JSmith 13 63 20.6
Knight 9 45 20.0
Artest 6 30 20.0
JSmith 12 62 19.4
JuHoward 10 52 19.2
GWallace 15 79 19.0
Brand 35 185 18.9

Only 29 of Stoudamire's 71 free throws have been of the shooting variety.

Code:

Player BFTA SFTA Pct
----------------------------
Cardinal 0 50 0.0
Marshall 0 52 0.0
JrCollins 0 53 0.0
Lue 0 56 0.0
Hart 0 78 0.0
Cato 1 62 2.1
QRichardson 2 99 2.4
ADavis 1 54 2.7
Udrih 1 50 2.9
Ely 2 74 3.8

Shaq, incidentally, is at 15.8%, tied for 33rd with LeBron James ... and Jeff McInnis.

Kevin Pelton
Site Admin

Joined: 30 Dec 2004
Posts: 970
Location: Seattle

Posted: Mon Jan 24, 2005 10:30 pm Post subject:

We can also look at the ratio of shooting free throw attempts to total free throw attempts, which turns out to be pretty interesting:

Code:

Player SFTA FTA Pct
---------------------------
Drobnjak 66 72 91.7
ECurry 148 162 91.4
Mihm 125 137 91.2
Stoudemire 354 388 91.2
Mohammed 120 132 90.9
Ratliff 56 62 90.3
Sweetney 92 102 90.2
Ilgauskas 221 246 89.8
Fortson 165 185 89.2
LWright 57 65 87.7

This bears a suspicious resemblence to the original BFTA% list, for good reason -- it measured more the ability to get a lot of shooting fouls than to really convert fouls while shooting into makes.

Code:

Player SFTA FTA Pct
---------------------------
DJones 7 50 14.0
Atkins 29 86 33.7
Nash 35 102 34.3
Boykins 41 115 35.7
Hart 28 78 35.9
RMiller 27 70 38.6
Stoudamire 29 71 40.8
Posey 23 56 41.1
Ridnour 32 74 43.2
Van Exel 22 50 44.0

See, I did know something! Damon Jones isn't really that bad at converting BFTA opportunities; he's just only been fouled in the act of shooting only four times all season. (That's once every 329.5 minutes.)

Incidentally, there still must be some errors in the NBA's data. I'm not sure how exactly one shoots free throws on double technicals or offensive fouls. Not a big deal, since there's only a handful of them, but still the case.

Ed Küpfer

Joined: 30 Dec 2004
Posts: 783
Location: Toronto

Posted: Mon Jan 24, 2005 10:45 pm Post subject:

admin wrote:

Incidentally, there still must be some errors in the NBA's data. I'm not sure how exactly one shoots free throws on double technicals or offensive fouls.

That's sweet of you to blame the NBA for the errors, but I assure you they're almost certainly mine. WRT to double-Ts, I'll comb the logs now to find the mistakes. Updates here.

However, the offensive fouls are real: if the team is in the penalty, the defensive team will shot FTs. Check out the action at the 0:19 mark of the 4th in this game.
_________________
ed

Ed Küpfer

Joined: 30 Dec 2004
Posts: 783
Location: Toronto

Posted: Mon Jan 24, 2005 10:59 pm Post subject:

Error found! Row 23350 reads:

Code:

05-01-04 PHX MIN outlaw,bo PHX 1 1 Double Technical 4 5:09

Should read:

Code:

05-01-04 PHX MIN outlaw,bo PHX 1 1 Shooting 4 5:09

That is the only Double-T. Files at Yahoo site updated to reflect changes. Further corrections gleefully accepted.
_________________
ed

Kevin Pelton
Site Admin

Joined: 30 Dec 2004
Posts: 970
Location: Seattle

Posted: Mon Jan 24, 2005 11:11 pm Post subject:

That's a mistake in the play-by-play. Offensive fouls are not team fouls, thus no shots can be attempted on them:

Quote:

PENALTIES: The offender is charged with a personal foul. The offended team is charged with a team foul if the illegal contact was caused by the defender. There is no team foul if there are personal fouls on one member of each team or the personal foul is against an offensive player.

http://www.nba.com/analysis/rules_12.html?nav=ArticleList

Ed Küpfer

Joined: 30 Dec 2004
Posts: 783
Location: Toronto

Posted: Mon Jan 24, 2005 11:17 pm Post subject:

admin wrote:

That's a mistake in the play-by-play. Offensive fouls are not team fouls, thus no shots can be attempted on them

Ah well, what do I know. Luckily they only make up 8 of the FTAs in my sample.

In other news, I've uploaded a zip file containing all the logs I used to compile the data, if anyone wants to have a go. It's at the Yahoo site.
_________________
ed

HoopStudies

Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA

Posted: Tue Jan 25, 2005 12:45 pm Post subject:

Ed Küpfer wrote:

In other news, I've uploaded a zip file containing all the logs I used to compile the data, if anyone wants to have a go. It's at the Yahoo site.

There are other errors with these logs, most pretty minor.

Then, of course, there are times when the wrong person is credited with something. This is not correctable because it's also done in the official stats, but fortunately rare. Well, it's rare for things like assists and rebounds, but it still seems to me that a fair number of steals are credited to the person least responsible for it.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.

Ed Küpfer

Joined: 30 Dec 2004
Posts: 783
Location: Toronto

Posted: Tue Jan 25, 2005 3:15 pm Post subject:

HoopStudies wrote:

- The score every so often goes berserk, going from 13-15 to 21-8 on one shot

Yeah, those are strange. This is one of the reasons mining PBP logs is such a chore for me. I wonder if the ESPN.com logs have this problem.

HoopStudies wrote:

- Sometimes a player is listed with a score change and you know they must have made a shot but it isn't shown.

Those aren't really a problem, unless you're logging the types of shots taken.

HoopStudies wrote:

- They are inconsistent in their naming, somewhat understandably. If Cleveland is playing Seattle, Lebron James is listed as L. James to avoid confusion with J. James. But he's listed as just James in most games.

The convention for NBA.com logs, if anyone is interested, is that if there is more than one player with the same last name on both teams playing, their surnames will be preceded by their first initial, a period, and a space. Otherwise, only their surnames will be shown.

What makes it difficult is some teams, during the course of the season, have two players with the same surname playing for them -- but not at the same time. For example, Steve Smith has played all season for the Bobcats. Until Dec 14, he would show up in the log as "Smith." After Dec 14, Theron Smith began playing for the team, and Steve's name was displayed under "S. Smith." At the end of the season, looking back at the games prior to December 14, you would have no idea which "Smith" it was to which the logs refered. You would have to look at the boxscores to see which one played. PITA.

HoopStudies wrote:

- I've noticed once that their labeling of what period it is is wrong.

Never seen that one. I'll keep an eye out.
_________________
ed

HoopStudies

Joined: 30 Dec 2004
Posts: 705
Location: Near Philadelphia, PA

Posted: Tue Jan 25, 2005 4:08 pm Post subject:

Ed Küpfer wrote:

HoopStudies wrote:

- The score every so often goes berserk, going from 13-15 to 21-8 on one shot

Yeah, those are strange. This is one of the reasons mining PBP logs is such a chore for me. I wonder if the ESPN.com logs have this problem.

The QC on this one is straightforward, fortunately.

Ed Küpfer wrote:

HoopStudies wrote:

- Sometimes a player is listed with a score change and you know they must have made a shot but it isn't shown.

Those aren't really a problem, unless you're logging the types of shots taken.

Depends on what you're doing. It causes me QC problems.

Ed Küpfer wrote:

HoopStudies wrote:

- I've noticed once that their labeling of what period it is is wrong.

Never seen that one. I'll keep an eye out.

I remember a case where the 1st period was labeled as the 2nd period. I think they then had 2 fourth periods.

Sportsline and ESPN both have different info. ESPN has no info on shot type, but has distance. Sportsline has shot type but little of several other things. Sportsline has shot charts that are nice, too.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.

CrazyFromTheHeat

Joined: 21 Jan 2005
Posts: 31

Posted: Tue Jan 25, 2005 8:48 pm Post subject:

admin wrote:

Damon Jones isn't really that bad at converting BFTA opportunities; he's just only been fouled in the act of shooting only four times all season. (That's once every 329.5 minutes.)

The trick with this is, there's a reason why DJ has only been fouled four times in the act of shooting all year. If DJ believed that he could get to the line regularly by driving to the basket regularly, then he would. He's not refusing to come inside the arc just because he's so good out there, he's doing it because he knows he has little chance of success in traffic.

A similar scenario would be saying that Shaq might be a good three-point shooter, we just don't know because he hasn't tried any. But we know that Shaq stays close to the basket because that's his best chance of success, and that if he started breaking the backboard with threes then he would diminish his value. DJ is the same way. He doesn't accumulate BFTAs because it's not what he does well, not because he isn't given an opportunity.
_________________
The Best Miami Heat Coverage
http://heat.mostvaluablenetwork.com/

Dan Rosenbaum

Joined: 03 Jan 2005
Posts: 540
Location: Greensboro, North Carolina

Posted: Tue Feb 01, 2005 3:06 pm Post subject:

Ed Küpfer wrote:

Here's how the types of FTs break down:

Code:

FTAs
TYPE 1 2 3 TOTAL

Away from Play 35 4 0 39
Clear Path 28 0 0 28
Delay Technical 4 0 0 4
Double Technical 1 0 0 1
Elbow 0 4 0 4
Flagrant Type 1 0 77 0 77
Flagrant Type 2 0 12 0 12
Hanging Technical 12 0 0 12
Illegal Defense 341 0 0 341
Inbound 0 24 0 24
Loose Ball 3 1002 0 1005
Non Supported Technical 2 0 0 2
Offensive 0 8 0 8
Personal 8 6628 3 6639
Punching 2 0 0 2
Shooting 2473 18578 303 21354
Taunting Technical 6 0 0 6
Technical 446 0 0 446

TOTAL 3361 26337 306 30004

The FTAs that represent the end of a possession -- Shooting, Loose Ball, Away from Play, and Personal -- sum to 13259, or 44.2% of the total. Nice to have that FTA coefficent confirmed.

This is in response to a question that I got from DeanO through e-mail. When comparing teams, I compute true shooting percentage separately for two point attempts and three point attempts. But this raises a question. How do I allocate free throws between two point and three point attempts?

Two-point attempt free throws (possessions in parentheses)
Clear Path and Flagrant Fouls: 117 (0)
Shooting - 2 shot: 18,578 (9,289)
Shooting - 1 shot: 2,460 (0)
Total: 21,155 (9,289)

Three-point attempt free throws (possessions in parentheses)
Personal - 3 shot: 3 (1)
Shooting - 3 shot: 303 (101)
Shooting - 1 shot: 13 (0)
Total: 319 (102)

Unallocated (possessions in parentheses)
Away from Play - 1 shot: 35 (0)
Away from Play - 2 shots: 4 (2)
Technicals, Punching, and Illegal Defense: 814 (0)
Elbow: 4 (2)
Inbound: 24 (12)
Offensive: 8 (4)
Loose Ball - 1 shot: 3 (0)
Loose Ball - 2 shot: 1002 (501)
Personal - 1 shot: 2 (0)
Personal - 2 shots: 6,628 (3,314)
Total: 8,530 (3,835)

-- I tended to try to overallocate to two pointers, allocating all clear path and flagrant fouls to two point attempts.
-- I allocated 13 Shooting - 1 shot fouls to three pointers assuming that the odds of making a three-pointer when fouled was half that of making a two-pointer when fouled. This is only 13 free throws, so it is not a big deal whatever I do.

Now clearly there are some screwy categories here, and I am sure a handful of free throws are misallocated, but I think this is basically right. In essence, most free throws are fall in the two point attempt and unallocated categories. There are very few in the three point attempt category.

My argument from here is that the unallocated free throws should be allocated to two pointers and three pointers by their proportions of overall shots. If two pointers are 75% of field goal attempts, then 75% of the unallocated free throws should be allocated to two pointers and 25% should be allocated to three pointers.

This would differ by team and would be different on offense and defense for a given team, but for an example, we can do this with the league averages.

As of January 31, about 80.6 percent of field goal attempts in 2004-05 were two point attempts, leaving 19.4 percent as two point attempts. Thus, this would result 6,878 of the unallocated free throws (and 3,092 of the unallocated possessions) being allocated to two pointers and 1,652 of the unallocated free throws (and 743 of the unallocated possessions) being allocated to three pointers.

Thus, after allocation here are the totals (with possessions in parentheses).

Two-point attempts: 28,033 (12,381)
Three-point attempts: 1,971 (845)

Note that this results in a possession multiplier of 0.442 for two point free throw attempts and 0.429 for three point free throw attempts.

gabefarkas

Joined: 31 Dec 2004
Posts: 1303
Location: Durham, NC

Posted: Tue Feb 01, 2005 5:08 pm Post subject:

how do you guys handle FTA's where the team retains possession after the FT's? like technicals or flagrants and such? is it two possessions, or one?

Dan Rosenbaum

Joined: 03 Jan 2005
Posts: 540
Location: Greensboro, North Carolina

Posted: Tue Feb 01, 2005 5:15 pm Post subject:

gabefarkas wrote:

how do you guys handle FTA's where the team retains possession after the FT's? like technicals or flagrants and such?

is it two possessions, or one?

I think the consensus is that technicals count as zero possessions. In general, the possession of the ball does not change from one team to the other during a technical, flagrant, etc. Counting them as zero possessions is one of the reasons why free throw attempts count, on average, only 0.44 possessions.

Display posts from previous:

	APBRmetrics Forum Index -> General discussion	All times are GMT - 5 Hours Goto page Previous 1, 2, 3, 4 Next
Page 3 of 4

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum