PDA

View Full Version : Professional Basketball Player Evaluation



Marcus Bryant
04-29-2009, 12:06 PM
The evaluation of basketball talent has been revolutionized, as it has in other sports, namely baseball, by the introduction and use of quantitative analysis over the last decade. Baseball has had a head start over other sports by decades in terms of those who have sought to reveal hidden truths about the game and about players through the use of statistics, beginning with the establishment of the Society for American Baseball Research (SABR) (http://www.sabr.org/) and the development of Sabermetrics (http://en.wikipedia.org/wiki/Sabermetrics), which was based on the legendary work of Bill James (http://www.billjamesonline.net/). This was eventually picked up by others, most notably Boston Red Sox wunderkind GM Theo Epstein (http://en.wikipedia.org/wiki/Theo_Epstein) who relied heavily on statistical analysis to construct the first World Series Champion Red Sox team since 1918 in 2004 (and again in 2007), as was chronicled in Mind Game (http://www.amazon.com/Mind-Game-Created-Blueprint-Winning/dp/0761140182/ref=pd_bbs_4?ie=UTF8&s=books&qid=1241023040&sr=8-4) and Oakland A's GM Billy Beane (http://oakland.athletics.mlb.com/oak/team/exec_bios/beane_billy.jsp), whose use of quantitative analysis in a sport dominated by wizened old baseball men who judged players primarily through eyesight and old baseball "truths" served as the object of the general narrative of the bestseller Moneyball (http://www.amazon.com/Moneyball-Art-Winning-Unfair-Game/dp/0393057658). The famed annual Baseball Prospectus (http://www.amazon.com/Baseball-Prospectus-2009-Essential-Season/dp/0452290112/ref=sr_1_1?ie=UTF8&s=books&qid=1241023219&sr=1-1) has been notably quite accurate in predicting the performance of players in Major League Baseball.

While numbers are an integral part of basketball, the application of quantitative analysis to evaluating player potential and projecting team performance is relatively new. Former Spurs intern and assistant GM Sam Presti (http://www.nba.com/spurs/features/presti_bio.html) (now GM of the Oklahoma City Thunder) was known to have used such analysis in his work in the Spurs front office and is the one acknowledged to have convinced Gregg Popovich and RC Buford to draft Tony Parker with the 28th pick in the 1st round of the 2001 NBA Draft.

There are now a variety of sources for the application of quantitative analysis in basketball. The Association for Professional Basketball Research (APBR) (http://www.apbr.org/) is one, which also hosts an APBRmetrics forum (http://www.sonicscentral.com/apbrmetrics/). In addition, 82games.com (http://www.82games.com/) is a good source for situational stats. databaseBasketball.com (http://www.databasebasketball.com/) and basketball-reference.com (http://www.basketball-reference.com/) are good sources of data. Journalist John Hollinger (http://insider.espn.go.com/espn/blog/index?name=hollinger_john&action=login&appRedirect=http%3a%2f%2finsider.espn.go.com%2fesp n%2fblog%2findex%3fname%3dhollinger_john) has been an early adopter of using stats to understand the NBA and has written extensively on the subject. He maintains Alleyoop.com (http://www.alleyoop.com/) which presents some of his work.

By no means is this analytical work the beginning of the end for basketball scouts, but the ground does seem fertile for those professional teams which are willing to apply it to more objectively evaluate players. In addition, it should bring greater understanding of the game and player performance to fans. This will be a topic we will visit from time to time here in the Think Tank, as well as the more traditional methods of player evaluation.

-MB

Bruno
04-29-2009, 02:15 PM
Nice post. :tu

There is a "new school" GMs who have created statistical analysis software to analyze players. Pritchard and Blazers are said to heavily use that kind of system.

El Jefe
04-29-2009, 02:34 PM
What's really interesting is the amount of statistics being combined with video. It used to be difficult to get tapes on Intl prospects. If you have access to the proper resources, you can now pull up things like the last 10 times James Gist drove to his left, or watch every instance of Tiago Splitter setting up on the right block.

Marcus Bryant
04-29-2009, 04:43 PM
MIT (http://www.mit.edu/)'s Sloan School of Management (http://mitsloan.mit.edu/) hosts an annual conference (http://www.sloansportsconference.com/2009/home/) on sports analytics. The one for this year was held last month. There was a panel discussion (http://www.sloansportsconference.com/2009/agenda/basketball-analytics/) on basketball analytics. While a couple of the panelists will make any Spurs fan gag, overall it looks like the conference organizers were able to bring in some interesting speakers for their panels.

There was also a panel discussion (http://www.sloansportsconference.com/2009/agenda/careers-in-sports/) on careers in sports, with an apparent emphasis on quantitative roles in professional sports organizations. There was another panel (http://www.sloansportsconference.com/2009/agenda/talent-identification/) on talent identification and one of the panelists was an executive from Chelsea FC of the English Premier League.

Press accounts of the conference can be found here (http://www.sloansportsconference.com/2009/press/). Apparently video from the conference will be made available at a later date.

TDMVPDPOY
04-29-2009, 04:53 PM
Nice post. :tu

There is a "new school" GMs who have created statistical analysis software to analyze players. Pritchard and Blazers are said to heavily use that kind of system.

didnt someone from the rockets use nbalive as a software to analyze the teams players lmao

Marcus Bryant
04-29-2009, 04:55 PM
didnt someone from the rockets use nbalive as a software to analyze the teams players lmao

That would be Daryl Morey (http://en.wikipedia.org/wiki/Daryl_Morey).

Marcus Bryant
04-29-2009, 05:13 PM
Thinking broadly, baseball is a game in which so much of the game is determined by batters against pitchers and vice versa. It would seem that there are limitations in analyzing basketball quantitatively that there are not in baseball. Still, you can look at how a player performs against different teams and whether or not that is significant when compared to his performance against the league.

sabar
04-29-2009, 05:44 PM
I like 82games and the various stats they calculate.

http://www.82games.com/0809/FGSORT7.HTM

Marcus Bryant
05-02-2009, 12:00 AM
Bill Simmons: Never come between an NBA stat geek and the truth. (http://sports.espn.go.com/espnmag/story?id=4011524)

Marcus Bryant
05-02-2009, 12:49 AM
BasketballProspectus.com: There Is No Secret Sauce (http://www.basketballprospectus.com/article.php?articleid=656)

Marcus Bryant
05-02-2009, 01:05 AM
The Wages of Wins Journal (http://dberri.wordpress.com/)
Basketball Prospectus (http://www.basketballprospectus.com/)

Chieflion
05-02-2009, 02:29 AM
didnt someone from the rockets use nbalive as a software to analyze the teams players lmao

It did not calculate injuries. (Tracy McGrady)

wildbill2u
05-07-2009, 12:34 PM
Somebody needs to use a ouija board or whatever software makes them happy to evaluate the current Spurs roster. Our role players are simply not skilled enough, not athletic enough, not in their right position, etc. for the team to be successful.

Watching the role players for the other teams in the playoffs simply underscores how we've been kidding ourselves about the Spurs role players. My God, how we could use a Birdman or Bass or Landry or....

Finally, the one attribute that role players must offer--hustle and focus--simply wasn't there for the most part compared to the Chicago bench, the Denver bench, the Rockets bench, the Mavs bench, etc.

MasonThe question I have is why the coaches and fans here at Spurstalk kept talking about Mason as the quality player we needed as the last piece of the puzzle.

1. He was meiocre at best on defense, only made to look acceptable when he took Finley off the floor.

2. His ball handling/dribbling was awful (the image that will always remain with me was his stumbling fall down the open lane against the Mavs). Forget the PG experiment. He shouldn't be here long enough to develop a skill that he obviously has no talent for.

3. As a shooter he was inconsistent. That's not acceptable for him in our offense. Forget about his late heroics in a couple of regular season games. He disappeared when the real pressure--the playoffs--came around.

BONNER

1.He improved his defense somewhat is about all I can say. He does hustle most of the time.

2. His offense was inconsistent and hesitant. It doesn't matter if you lead the league in 3pt. percentage if you won't take the shot when open.

3. Rebounding negligible, especially offensive rebounding, but that may be accounted for by the fact that he was put out on the perimeter so much to spread the floor. Can't rebound from 23 feet away.

GOODEN
1. He's a good'en on offense. Bit of a ball hog, even when playing with any of the Big Three. When the ball goes into his hands, it's a black hole where nothing ever comes out.

2. Little D. Not too smart. Might be better next year.

3. Rebounding nothing to write home about.

BOWEN & THOMAS

Gave us all they had left in the tank I guess, but that isn't enough to win minutes on a championship run in the tough Western Conference. Father Time visits us all. They need to retire with honor.

HILL

The Best of our role players because of his upside--but let's not get carried away with dreams of superstardom. Everybody loves a kid coming out of the late picks in the draft with no props who makes the team--but he's not EVER gonna be as good as Tony Parker. And he may NOT be a PG after all.

UDOKA

The fact that Udoka played his way back into the rotation tells you that our role players aren't up to par with the rest of the West. He's not real bad, just not as good as we need.
http://www.spurstalk.com/forums/images/Style_Templates/DefaultStyle/statusicon/user_online.gif http://www.spurstalk.com/forums/images/Style_Templates/DefaultStyle/misc/progress.gif

With Mahimi, Oberto and Marcus at the end of the bench we have nothing but question marks there.

We could lose all of these guys and pick up equally flawed players but trying to get better ones, ah, there's the rub.

mountainballer
05-08-2009, 05:55 AM
good post.
I agree that ALL role players of this current Spurs squad are removable and not crucial to the success of the team in any aspect.
the last role player of that kind was Bruce Bowen in the 2006-07 season. (BB in 2008 was still quite useful, but no longer that good to be called vital)

Mason was overrated on this board for most of the season. I still think he was a nice addition (maybe slightly overpayed), but no way he is the kind of starting SG a contender would need.
(I tried to argue in December, that Mason would be a nice trade bait, because his value will never ever be that high again. almost all people called this crazy)
additionally to the many limitations he displayed, it was also obvious, that he doesn't have the endurance to survive a whole NBA season playing starter minutes. he broke down in the 2nd half of the season like a rookie, who hits the wall.
overall, he is a nice player to have on the roster, but only in the role of the 8th or 9th man, playing 15-20 minutes. (this would also allow to keep him fresh enough to get something out of him in the PO).
but if he is a needed piece for a trade that makes sense, just do it and don't even hesitate.

Udoka on the roster is ok, if we consider his salary of just 1 million. that's the salary for a 3rd stringer and you also need those 3rd stringers. but neither will he ever replace Bruce, nor will he be a useful part for the regular rotation. 11th or 12 man, that's ok.
I would still not re sign him, even if he signs for this kind of money for just one year. Spurs should try to get someone like Quinton Ross or Dahntay Jones for similar money instead. those signings would have much more upside.

Marcus Bryant
05-08-2009, 07:55 AM
http://mybroadband.co.za/vb/picture.php?albumid=65&pictureid=276

Solid D
05-09-2009, 09:06 PM
I wonder if statistical & quantitative analysis would have seen the real player behind the numbers of a 6'5" shooting guard out of Loyolya: 26.3 ppg 9.5 rpg and 1.1 apg. (?)

FromWayDowntown
05-09-2009, 10:09 PM
I wonder if statistical & quantitative analysis would have seen the real player behind the numbers of a 6'5" shooting guard out of Loyolya: 26.3 ppg 9.5 rpg and 1.1 apg. (?)

I have always figured that was part of a very subtle tanking effort that paid off in 1987. Had they picked the next guy in line that year, they might have made the playoffs in 1987 and turned into the Utah Jazz, only without John Stockton.

Marcus Bryant
05-19-2009, 10:21 AM
http://online.wsj.com/article/SB124269925054533267.html

Textbook Management? Hardly.
Assembled Largely by Instinct, the Denver Nuggets Keep Winning; Mastering a 'Curious Business'

By MATTHEW FUTTERMAN
The Wall Street Journal
MAY 19, 2009

The Boston Celtics have the former NBA all-star Danny Ainge calling the shots. The Houston Rockets have Daryl Morey, a statistical whiz with an MBA from MIT.

The Denver Nuggets, meanwhile, manage with a five-headed beast whose most influential voice comes from a guy who broke into pro basketball as a financial planner. Somehow though, as the NBA's conference finals begin, the Nuggets look like the deepest and possibly most dangerous team standing. With 54 wins in the regular season and convincing victories over the New Orleans Hornets and Dallas Mavericks in the first two rounds of the playoffs, they now take on the Los Angeles Lakers on Tuesday night in Game 1 of the conference finals, the team's first in 24 years.

http://s.wsj.net/public/resources/images/PJ-AP779_SP_NUG_DV_20090518200835.jpg

Few people are more surprised than most of the folks who run the Nuggets. They don't describe their success as the inevitable result of a carefully designed strategy. Rather, in an era when sports executives like to play themselves off as masters of mathematical analysis and risk management -- and in a year when most NBA teams chose fiscal prudence over expensive superstars -- the Nuggets are an anomaly. They owe their success to a bizarre combination of luck, good health, opportunism and a management strategy that is more six-shooter than Six Sigma.

"I promise you there isn't another franchise in the league that works the way we do," says Bret Bearup, the franchise's "adviser," a sort of diplomat without portfolio who derives his substantial influence over the team's moves from his 11-year friendship with owner E. Stanley Kroenke.

Mr. Bearup, a former financial planner for athletes, isn't the Nuggets' only contrarian element. With few exceptions, (last season's champion, the Celtics, acquired two mid-career superstars in the off season of 2007) most NBA franchises build slowly, adding parts each year to an expanding core. Since 2002, dating to former general manager Kiki Vandeweghe's regime, the Nuggets have had a revolving door, with stars Marcus Camby, Allen Iverson, (the first two picks in the 1996 draft) and Kenyon Martin (the top pick in 2000), among those who have tried out for the role of franchise lynchpin. Only Mr. Martin remains.

"You try not to get frustrated because you know it's a business," says Carmelo Anthony, the team's sublime 6-foot-8 forward who is averaging 27 points per game in the playoffs. "But my past five years, I had a lot of different players I played with."

Also, in an era when numerous NBA teams horde draft picks like precious family jewels, the Nuggets trade theirs like baseball cards. They sent 10 picks away in the past five years to acquire the current roster, including three future first-round choices to land Mr. Martin.

http://s.wsj.net/public/resources/images/OB-DR938_NUGGET_DV_20090518200959.jpg
Getty Images

Nuggets owner E. Stanley Kroenke and his managers value chemistry and camaraderie in the roster.


"I wish I could tell you we knew it was going to turn out this way, but we really didn't," says Rex Chapman the team's vice president of player personnel.

Finally as nearly all sports executives rely increasingly on statistical analysis, Mr. Chapman says the Nuggets act on instinct rather than spreadsheets. "When you are talking about chemistry and camaraderie, you have to," he says.

That chemistry was on display Wednesday night, when Mr. Martin provided the exclamation point on the Nuggets' four games to one conference semifinal victory over the Mavericks. With one minute to go and the Mavericks struggling to survive, guard J.R. Smith drove the lane and gave up an open eight-foot jumper so he could loft a pass to Mr. Martin, who grabbed it in mid-air and slammed it home. That sort of teamwork has turned the Nuggets from selfish first-round losers last season, when the Lakers swept them from the playoffs, into a scary-good bunch this time around.

"We had players before, but it's a team now," Mr. Martin says. "They were just trying to find a nucleus that worked."

For years little did, even as Mr. Kroenke subsidized one of the league's highest payrolls and lost millions of dollars in the process.

Talented offensively at nearly every position, the Nuggets had little leadership or defensive prowess until they persuaded the Detroit Pistons to trade point guard Chauncey Billups for Mr. Iverson four games into the season. Mr. Billups, who is playing in his seventh consecutive conference final, is the sort of baggage-free floor leader and franchise cornerstone that rarely becomes available in today's NBA. "I just know how to win, and that's really all it is," Mr. Billups says in a voice barely above a whisper.

But Detroit wanted to cut back and begin rebuilding by freeing up space under the league's salary limits to make a run at a top free agent this off-season or next. Mr. Iverson's contract expired at the end of the season. Mr. Billups, on the other hand, was owed nearly $40 million through 2011, an obligation the Nuggets, his hometown team, are thrilled to have, even though just months before they gave up Mr. Camby, in part to save money.

"Basketball is a very curious business," says Mr. Kroenke, a leading developer and entrepreneur who has met his match since buying the team in 2000.

"We've had some hiccups," says Mr. Bearup, an affable former Kentucky Wildcat who stands roughly 6-foot-9 with hands like slabs of meat. He met Mr. Kroenke 11 years ago at the elite ABCD Basketball Camp. At the time Mr. Bearup had a business plan that he acknowledges pushed the limits of NCAA regulations. He would recruit the best scholastic basketball talent to his summer all-star teams, which he used as a tool to establish relationships with players who would later hire him as their financial adviser.

At the camp, Mr. Bearup noticed a quick-footed guard follow a steal with an impressive drive to the basket. "The kid's good," Mr. Bearup recalls saying to the stranger sitting next to him.

"That's my son," Mr. Kroenke told him.

Now Mr. Bearup serves as Mr. Kroenke's basketball consigliere, the bridge between the owner and a basketball operations staff that includes vice president of basketball operations Mark Warkentien, a basketball lifer with more than a quarter century of experience in college and the pros. Head Coach George Karl, a 21-year NBA veteran head coach, also chimes in, and so does Mr. Kroenke's son, Josh, and Mr. Chapman. But the final approvals rest with Mr. Kroenke, and the last person he speaks to is Mr. Bearup.

"We don't do anything without first running it by Bret and Stan," Mr. Chapman says.

It was Mr. Bearup's idea last summer to pick up forward Chris "Birdman" Andersen, who'd been suspended for drug abuse for most of the previous two seasons. With his spiked-hair and heavily tattooed arms, Mr. Andersen has become a cult-hero in Denver, where the team store sells 4,000 Birdman T-shirts per week. Mr. Andersen was a leader in the league in shot blocks this season.

"I don't know any team where one guy has all the answers," Mr. Warkentien says. "We went from 45 wins my first year to 50 games last year to 54 this season. That shows you something is working."

Indeed, but even Mr. Karl is still adjusting to the surprising success.

"Oh yeah, I just knew it would work," he says, rolling his eyes. "I'm a historian of the NBA, and the NBA's history says you don't speed up the process and go from a bad team out of the playoffs to one in the position we're in."

Write to Matthew Futterman at [email protected]

Printed in The Wall Street Journal, page D8

DPG21920
05-19-2009, 11:06 AM
“…if one team is good enough to warrant beating another in 55% of its games, the weaker team will nevertheless win a 7-game series about 4 times out of 10. And if the superior team could beat its opponent, on average, 2 out of 3 times they meet, the inferior team will still win a 7-game series about once every 5 match-ups. There is really no way for a sports league to change this. In the lopsided 2/3-probability case, for example, you’d have to play a series consisting of at minimum the best of 23 games to determine the winner with what is called statistical significance, meaning the weaker team would be crowned champion 5 percent or less of the time. And in the case of one team’s having only a 55-45 edge, the shortest significant “world series” would be the best of 269 games, a tedious endeavor indeed! So sports playoff series can be fun and exciting, but being crowned “world champion” is not a reliable indication that a team is actually the best one.”
http://dberri.wordpress.com/

I found that to be very interesting. Although a team being what they claim to be "55%" likely to win is not statistically significant, that 55% shows a team is better. That is why the "it's all about match-ups" argument has arisen.

barbacoataco
05-20-2009, 01:57 AM
One important difference--- Even though baseball is a team sport, when the batter faces the pitcher it is essentially a man-to-man competition. Basketball on the other hand is much more of a true "team" sport. This makes evaluating statistics much harder. For example, a player on a team with Steve Nash will benefit from his skills by getting more open looks etc. I know in baseball there is an aspect where you might see better pitches if protected by a better hitter.

Also, in baseball, all hitters hit the same number of times. In basketball your best players take more shots, or ball hogs take more shots.

The statistical analysis of on the ball defense is not what it should be. I think it is possible. There should be stat that shows the way Bowen shuts down players.

The biggest difference IMO goes back to my first point. In baseball you really are collecting pieces of talent that are kind of unrelated. In basketball, team chemistry is more than the sum of its parts. Look at the way Chauncey Billups changed Denver's whole team this year. Bringing in a new player in baseball does not change your whole team like that, it just gives you one more hitter.

Blackjack
05-21-2009, 12:58 AM
College Basketball stats and analysis

statsheet.com/ (http://statsheet.com/)

poop
05-21-2009, 11:35 AM
LOL give me a break, this has gone WAY too far.

just watch the fucking games, analyze the box score for a few minutes and you can figure it out.

relying on vast, abstract mathmatical formulas isnt going to do anything special.

spurster
05-21-2009, 12:06 PM
I think the question of whether one player like TD, Nash or Billups makes their teammates better is not that hard of a statistical question. Surely, one can make correlations comparing when TD is on floor vs. when he is on the bench. This would have to be done over enough games so that individual defenses (do they double TD or not?) average out. Why do you think Manu is so often used as the 6th man?

Marcus Bryant
05-21-2009, 12:50 PM
LOL give me a break, this has gone WAY too far.

just watch the fucking games, analyze the box score for a few minutes and you can figure it out.

relying on vast, abstract mathmatical formulas isnt going to do anything special.

Tell that to Red Sox Nation.

Marcus Bryant
05-21-2009, 01:02 PM
One important difference--- Even though baseball is a team sport, when the batter faces the pitcher it is essentially a man-to-man competition. Basketball on the other hand is much more of a true "team" sport. This makes evaluating statistics much harder. For example, a player on a team with Steve Nash will benefit from his skills by getting more open looks etc. I know in baseball there is an aspect where you might see better pitches if protected by a better hitter.

Also, in baseball, all hitters hit the same number of times. In basketball your best players take more shots, or ball hogs take more shots.

The statistical analysis of on the ball defense is not what it should be. I think it is possible. There should be stat that shows the way Bowen shuts down players.

The biggest difference IMO goes back to my first point. In baseball you really are collecting pieces of talent that are kind of unrelated. In basketball, team chemistry is more than the sum of its parts. Look at the way Chauncey Billups changed Denver's whole team this year. Bringing in a new player in baseball does not change your whole team like that, it just gives you one more hitter.

True. In baseball, it is much easier to isolate offensive performance (and defensive, as it relates to the pitcher) as so much happens at the plate.

One of the key insights of Sabermetrics was that on-base percentage is a much better predictor of runs scored and ultimately, wins, than simply batting average. Not to mention slugging percentage as opposed to batting average. So while a singles hitter with a high batting average may seem like a good offensive player, the moderate power hitter with a low batting average but who draws a lot of walks and hits for power is indeed a greater contributor to a team's success.

The main thing is that the quantitative analysis of baseball simply has a head start on its application in other sports. The technology is certainly there to capture all kinds of stats from basketball games, complete with video, and mine that data for insights and the hidden values of players. Think about it, you can know instantly how well a player performs against 1on1 defense in a certain arena, when his team is down 10+ points versus being ahead 10+ points, with 10 or fewer seconds left on the shot clock, against a certain defender.

Ultimately, there are aspects of the game that won't be changed by statistical analysis. But personnel management and coaching decisions will be informed by it. It is certainly worth challenging the conventional wisdom of a game. Is it really worth risking an out for a runner to attempt to steal second base? When you're down 4 with 30 seconds to go should you attempt a 2 point or a 3 point field goal?

DPG21920
05-21-2009, 01:12 PM
There are all kinds of applications to go along with limitations. There will always be two ends of the spectrum: those who use their instincts and those who rely solely on data.

The best bet is probably somewhere in the middle. You saw the Rockets have a thick packet of data on Kobe for Battier to study. It broke down into countless categories: Kobe off the dribble, off a pull up, going to his left, his right, which spots on the floor...

I remember seeing something in football that a statistical analysis showed that going for it on 4th down (or calculating the odds of achieving a 1st down if you have 4 plays to get it) is much higher than you would expect. So it was concluded that more teams, based on statistics should be going for it on fourth down. It had it broken down by position on the field and yardage needed. Yet many teams do not go for it.

There are times to play the odds and times to use your instinct.

barbacoataco
06-04-2009, 05:24 PM
That point about going for it on 4th down is a good example of where coaches do not do what the stats show. According to the numbers if you're past your own 30-40 yard line it is better to go for it on most 4th down situations. But fans would call you crazy.

I think one of the things they are using stats for in basketball is analyzing tendencies of offensive players.

I can't stand it when fans compare two players and say something like, "they both are 20 ppg 8 reb players so therefore they are more or less equal. Leaving out FG%, TO's, defense, offensive rebounding, clutchness and everything else.

I have thought for many years a good stat in basketball is Points per Shot. This shows the overall efficiency of a player including FT's, 3 ptr, FG%. It answers the most basic question of offense.