Basketball Prospectus: Unfiltered Everything Else is Fluff.

January 25, 2012

Why my top 25’s about to look weird

Filed under: Uncategorized — John Gasaway @ 11:59 am

I’ve been writing about college hoops now since the Harding Administration, but this season I’ve been asked to do something that I’ve never done before. This season I’m filling out a top 25 each week.

If you’ve ever wanted to yell at a writer about how incomprehensibly stupid their top 25 so obviously is (guilty), I would ask that you do one thing. Create your own top 25. You’ll find that it’s an intrinsically unsatisfactory exercise. What is plain and indeed obvious to the naked eye can’t be captured by a listing of teams from No. 1 to No. 25. Invariably there’s a top tier of elite teams that everyone rightly recognizes as Final Four material, and then there’s a big drop-off in quality before you get to the next set of programs. But of course in rankings there’s no such thing as a “big drop-off,” you just go from No. 8 to No. 9. Yell at your own rankings all you want, but someone has to be the No. 9 team in the country.

My main objection to top 25’s as they’re traditionally crafted, however, has less to do with their innate nature and more to do with how pollsters rank teams. The problem with any top 25 is that it’s little more than a list of the longest current winning streaks, combined with teams that are losing and which were likely overrated to begin with. (Did you know that Louisville is ranked No. 25 in this week’s Coaches’ Poll? That’s astounding to me. The Cardinals have recorded the Big East’s tenth-best per-possession performance to date.)

Now, there’s nothing wrong with winning streaks. Kansas, to take one example, has a pretty good one going, and I believe the Jayhawks really are one of the best teams in Division I. But once we surrender to the notion that a team stays where they are in the rankings until they lose, we’ve forfeited about 75 percent of our discretionary freedom of movement right there.

Which is why I’m going to employ a different method for ranking teams. I figure if I’m going to the trouble of tracking thousands of possessions played by 157 different teams I might as well use what I have, right? And every year starting right now, in late January, I feel like a critical mass has been reached in terms of possessions played. I feel like I now have enough information to dissent from the conventional wisdom on a given team with some degree of confidence. The dissent may well turn out to be wrong, of course, but at least my erroneous dissents will be less boring than parroting what everyone else says. Anyway that’s the hope.

So much for the white paper, let’s look at some actual instances where the world will think I’m nuts.

Kentucky
The Wildcats will not be No. 1 in my top 25 that comes out tomorrow. Shocking, I know. UK is plainly one of the best teams in the country, but you don’t need John Gasaway to tell you that the Cats haven’t lost since December 10. You know that already. The question that needs answering is: among these best teams in the country, who’s been performing at a level better than anyone else? That’s a close question, but my answer there is not to be found in Lexington.

Murray State
I think the proper reward for the Racers is my hearty congratulations, offered here, on an amazing 20-0 season. I do not think, however, that people truly believe MSU could go .500 on a neutral floor against, say, the six teams currently ranked right below them in the polls (UNLV, San Diego State, Florida, Creighton, Indiana, and Marquette). I think it’s simply the “0” in “20-0” that’s driving this ranking, which is more of a reward and a pat on the back than a true estimation of the team’s heft. Well, I’m going to offer my true estimation: It’s close and I’m open to persuasion but right now the Racers are not a top-25 team.

Mississippi State
The Bulldogs have been far more impressive to humans than they are to laptops all season long. Fortunately my business card reads “Mediating human-laptop disputes since the Harding Administration,” so I’m qualified to tell you the truth here is somewhere in the middle. Winning at Vanderbilt is nothing to sneeze at, certainly, but then again struggling to beat a struggling Tennessee team on your home floor or losing on the road to Ole Miss are data points as well. MSU’s been outscored by their SEC opponents thus far, and in fact they’re just 12 points away from an 0-5 record in-conference. Rick Stansbury‘s team may well be better than laptops think, but their defense has been frighteningly bad and they are not yet top-25 caliber.

That wacky Missouri Valley
If I feel like I now have enough tracked possessions in the bank to dissent from the conventional wisdom, I really feel that way when it comes to the Valley, where teams are already at the halfway mark in their conference schedules. Creighton is tied for the league lead at 8-1 with Wichita State, but the Bluejays own the bragging-rights tiebreaker there, having recorded a 68-61 win over the Shockers in Wichita on December 31. Now the weird part: Wichita State, even taking that loss into account, is recording perhaps the most dominant performance by a Valley team in the past five years. Nationally the Shockers will of course be filed cognitively under what we’ve seen from the Valley’s best or second-best team in the past, but in truth we haven’t seen a team outscore the MVC by very nearly a quarter of a point per trip, at least not recently. In other words there’s a chance that Creighton and Wichita State are the two best teams the Valley has produced in a long while, and, coincidentally, they both came down the pike the same year.

Keep all of the above in mind when my top 25 drops tomorrow. Also keep in mind this isn’t college football, thank goodness, so none of this matters anyway.

Twitter: @JohnGasaway. Contact: here.

January 24, 2012

What if the Wizards Don’t Get Better?

Filed under: Uncategorized — Kevin Pelton @ 3:14 pm

I’m not quite sure what to say about today’s news that the Washington Wizards have relieved Flip Saunders of his duties of head coach. In this case, relieved seems like an especially apt adjective, as Saunders no longer has to deal with one of the most dysfunctional rosters in recent NBA memory. At the same time, Saunders has to answer for the failure of the Wizards’ young players–particularly John Wall–to develop under his watch, as our Bradford Doolittle chronicled while explaining why Paul Westphal needed to go in Sacramento.

When the coach of a 2-15 team getting outscored by 10.6 points per game gets fired, it’s easy to predict improvement. Regression to the mean all but ensures the Wizards will play better the rest of the way, because almost nobody is actually this bad. Since the merger, just 11 NBA teams have been outscored by double digits on average, and Washington’s record would be a 10-win pace for an 82-game season, threatening the 9-73 record of the 1972-73 Philadelphia 76ers.

At the same time, the track record of Saunders’ replacement (his long-time assistant, Randy Wittman) does not exactly inspire confidence, to put it politely. Friend of BBP Tom Ziller has the ugly numbers on SBNation.com.

The most intriguing coaching rating available is done by Stats for the NBA, which treats coaches as the sixth player on the court as part of the regularized adjusted plus-minus process, quantifying their effect on players. There are many issues with this process, most notably the assumption that player ability is static throughout the 10-year period during which players and coaches are rated, but it’s a pure measure of coaching ability unlike any other. Saunders, by this measure, comes out as one of the league’s better coaches in this period, which is consistent with his track record in Minnesota and Detroit. Wittman, by contrast, lurks at the bottom of the list with Lawrence Frank (?!?) and a bunch of names who will never coach in the NBA again.

The Wizards are still overwhelmingly likely to get better the rest of the way, but promoting Wittman may not have helped their chances.

Player Skill Set Ratings

Filed under: Uncategorized — Neil Paine @ 12:29 pm

I’ve been thinking a lot recently about “fit” when it comes to building a basketball team.

Too often, statheads are guilty of plugging in players’ metrics and expecting a certain outcome without giving a thought to how the skills of each player mesh together. Dean Oliver’s “Skill Curves” are a simple way of modeling these teammate interactions, but they often fall short because Possession % is not always an adequate explanation for how a player produced points at a given efficiency.

A better way to look at fit would be to grade players’ skill sets and see how certain combinations of skills predict lineup efficiency. That’s a project for a later date, but today I thought I’d take a stab at generating ratings for players based on their skills. Here’s the process:

  • Generate projections for each player. In this case, I used Basketball-Reference’s Simple Projection System and regressed to positional means instead of league means. I also left out the age adjustment, since I want to estimate player skills in year Y, not year Y+1.
  • Decide which skills to measure, and how to measure them. I settled on the following “tools” as base skill sets for an NBA player:
    • Shooting – Pure shooting ability; measured by Free Throw Percentage
    • Scoring – Ability to create & make shots; measured by Points per 36 minutes
    • Floor Game – Passing & ball handling skill; measured by Pure Point Rating
    • Defense – Size + blocks, steals, & other stats that predict defensive +/-; measured by Defensive SPM
    • Rebounding – Ability to grab missed shots; measured by Total Rebounds per 36 minutes
  • Determine a grading scale. First, I calculated each player’s percentile rank (among players at the same position that season) in each category. Then, in true scouting tradition, I set up a 2-8 scale for each skill rating, corresponding to the following percentiles:
    • 2 = 0-3%
    • 3 = 4-13%
    • 4 = 14-35%
    • 5 = 36-65%
    • 6 = 66-87%
    • 7 = 88-97%
    • 8 = 98-100%
  • The average player at a position will have a rating of 5 in a given skill. Remember, players are rated relative to others at the same position, so a 5 Floor Game point guard is still likely to be a better passer/ballhandler than an 8 Floor Game center.

Without further ado, here’s a Google spreadsheet with the ratings for all seasons from 1982-2011.

Some fun notes…

The most “complete” players (highest simple sum of ratings) in the dataset seem to be Larry Bird and Kevin Garnett. In 2005, KG was an 8 (relative to other PFs) in Scoring, Floor Game, Rebounding, and Defense, with a 7 in Shooting, for a total of 39. Bird actually had three separate 39 seasons — 1984, 1985, and 1989; in the former two he was 8 across the board except in Scoring, and in the latter he pushed his Scoring to 8 (with his Defense falling to 7)… Chris Paul‘s 2009 and 2010 seasons were phenomenally complete, as he posted 8s in Scoring, Floor Game, & Defense, to go with 7s in Shooting & Rebounding… LeBron James is known for his all-around talent, which is borne out with 8s in Scoring/Floor Game and 7s in Rebounding/Defense. But the well-known hole in his game is also exposed here: he’s merely a 5 Shooter… Kobe Bryant‘s complete skills seem to have peaked in 2004, when he was an 8 Scorer, 7 Shooter/Rebounder, and 6 in Floor Game/Defense. In 2011 he remained an 8 Scorer (as he’s been every year since 2002), but his Shooting/Rebounding have fallen to 6s, and his Floor Game & Defense only score 5s… Arvydas Sabonis was a surprisingly great all-around player. Even at ages 34-35 in 1999 & 2000, he was an 8 Shooter relative to other centers, with 7s across the board in the other categories… During the prime of his first stint in the league, Michael Jordan was consistently an 8 Scorer/6 Shooter/7 Floor Game/8 Defender, with his rebounding varying between 7 & 8. When he returned from retirement, MJ’s Floor Game & Defense fell to 6, and in the twilight of his Wizards days he was a 7 Scorer/5-6 Shooter/6-7 Floor Game/7 Rebounder/5-6 Defender.

Email Neil at np@sports-reference.com. Follow him on Twitter at @Neil_Paine.

January 23, 2012

The Unknown Alternative

Filed under: Uncategorized — Kevin Pelton @ 2:33 am

In the wake of Saturday’s overtime New York Knicks loss to the Denver Nuggets, SBNation.com’s Tom Ziller had an intriguing column about how criticism of the Carmelo Anthony trade is based on the faulty premise that the performance of the Knicks and the Nuggets since the deal can be compared, which makes little sense given how much better Denver was before the two teams swapped players. I don’t think Tom went far enough. I think there’s another false comparison here, between New York before and after Anthony’s arrival.

The Knicks are, by all available statistical evidence, worse now than they were this time a year ago. However, to assume that they would have continued to play at the same level had they not made the trade ignores a number of key outside factors. First, whether because of injury or simply because he’s lost a step, Amar’e Stoudemire has been terrible this season. I was appalled watching the Phoenix game last week how much difficulty Stoudemire had beating his defender (usually Marcin Gortat) off the dribble, which was a key staple of the New York offense last season. Before the Anthony trade, Stoudemire was responsible for using about 32.0 percent of the Knicks’ plays. If he was doing that with his current .495 True Shooting Percentage, how big a problem would that be?

Second, Landry Fields morphed from a budding star to a borderline starter right about the time of the Anthony trade. Part of that, surely, is the way Anthony’s arrival changed Fields’ role in the office. At the same time, Fields was a second-round pick who was playing out of his mind for three and a half months. While Fields’ translated college statistics suggested he was underrated, they still projected him as barely better than replacement level as a rookie. It’s certainly possible that Fields’ last 44 games reflect his true talent level more than his first 50.

Having a true point guard in Raymond Felton would make a huge difference to the Knicks, and their depth would be stronger with Timofey Mozgov in the middle. Still, my suspicion is that a New York team without the Anthony trade would be struggling nearly as much as the current group. Since that team does not exist, however, it’s easy to apply to it whatever expectations best serve your conclusion about the Anthony trade. If you believe the Knicks would be in the middle of the Eastern Conference Playoff race with Felton and Danilo Gallinari, that can’t possibly be disproved.

The unknown alternative applies to all sorts of other decisions, too. I’ve been thinking about it lately in the context of forward/center Darnell Gant taking the final shot of Washington’s 69-66 loss to California on Thursday. Lorenzo Romar drew up the play to free Gant for an open shot despite the fact that he had missed his previous eight shot attempts on the evening, and Gant in fact missed at the buzzer to seal the loss. Inevitably, Romar was criticized for not getting the ball to star forward Terrence Ross, who had hit a three moments earlier to get the Huskies in position to tie.

Setting aside the question of whether Gant’s previous results mattered (aka “Hot Hand” theory), the interesting question to me is what chances Ross would have had of making the possible shot. Ross is making 39.4 percent of his triples this season, but Cal expected him to take the final shot and was prepared to take him away. We also know that percentages dwindle in the closing seconds because time pressures players to take more difficult attempts. Would Ross have had a 30 percent chance of tying the game? 25 percent? 35 percent? That’s the thing about end-of-game decisions: They’re usually a choice between a variety of unpleasant alternatives, which is why a missed shot should really be the expectation in such situations.

You can contact Kevin at kpelton@basketballprospectus.com. Follow him on Twitter at @kpelton.

January 20, 2012

Stats From the All-Time Kobe-vs-LeBron Series

Filed under: Uncategorized — Neil Paine @ 11:09 am

Just for kicks, here are composite stats from all games where Kobe Bryant and LeBron James played against each other:

Team Kobe
Wins: 5
Losses: 11
Offensive Rating: 104.4
Defensive Rating: 107.9
Best Win: +19.8 Efficiency Diff (1/19/2009)

Player G GS Min FG FGA 3P 3PA FT FTA ORB TRB AST STL BLK TOV PF PTS ORtg %Ps DRtg AWS36
Bryant 16 16 595 136 336 18 63 107 124 13 81 82 20 4 51 42 397 102.7 34.1 110.1 5.0
Odom 14 10 523 90 166 7 35 42 57 40 130 47 14 8 19 46 229 124.0 19.2 108.2 7.9
Fisher 10 9 307 32 78 10 33 12 17 3 24 30 11 0 15 20 86 99.0 15.1 109.0 2.7
Bynum 10 8 284 40 70 0 0 22 38 20 71 11 5 13 10 37 102 116.9 15.8 106.9 5.3
Gasol 7 7 265 53 104 1 5 20 29 21 62 15 3 3 7 13 127 115.4 21.6 111.3 6.9
Vujacic 11 0 163 20 46 11 27 9 10 2 18 14 6 1 3 10 60 123.3 13.7 110.1 6.3
Walton 8 3 156 13 35 1 4 4 6 6 23 21 6 1 8 7 31 90.6 15.4 109.2 3.4
Parker 4 4 145 19 42 4 13 2 3 2 9 14 6 1 4 9 44 106.5 14.8 114.5 4.0
Artest 5 4 144 17 46 8 18 3 8 15 29 19 7 0 5 17 45 106.9 19.2 110.1 5.5
Farmar 7 0 102 8 25 4 15 0 0 2 9 8 3 3 9 7 20 68.1 15.3 109.6 -0.1
Brown 3 3 101 9 16 0 0 4 9 7 20 3 1 4 1 11 22 122.0 10.6 110.6 4.0
VladRad 5 3 83 11 26 3 11 2 5 4 19 7 2 2 1 9 27 111.1 14.9 108.1 5.1
Mihm 4 3 82 16 28 0 0 6 9 3 16 2 2 4 3 14 38 115.8 18.3 109.0 4.6
Cook 7 1 80 13 31 4 8 7 8 5 19 4 2 1 4 8 37 101.0 20.2 102.9 5.3
Atkins 2 2 78 6 21 2 10 8 8 0 4 7 4 0 4 3 22 91.7 16.3 110.9 2.2
Butler 2 2 77 12 33 0 2 7 8 5 14 5 7 2 6 8 31 86.4 24.5 102.5 3.3
Brown 4 0 61 5 22 3 9 5 5 1 10 4 1 0 3 5 18 78.3 20.3 111.2 -0.3
Turiaf 4 0 60 4 12 0 0 8 14 6 14 1 0 3 3 13 16 93.0 17.0 110.7 -0.4
Ariza 3 0 58 7 15 2 4 2 2 4 8 5 4 3 3 6 18 112.3 15.8 102.7 6.4
George 3 1 54 10 18 1 3 0 0 6 13 2 4 2 3 5 21 101.1 17.3 91.8 8.2
Barnes 3 1 50 2 12 0 3 4 6 4 7 1 3 2 3 5 8 64.1 17.3 111.0 -0.7
Payton 1 1 42 5 22 0 1 3 4 0 4 4 1 0 6 2 13 54.5 28.2 87.2 -7.3
Jones 2 0 38 2 9 0 4 1 2 5 8 1 1 1 0 5 5 82.4 11.9 108.9 1.1
Evans 2 0 35 1 15 0 1 5 6 6 9 1 1 1 2 5 7 62.4 24.5 113.3 -4.2
Grant 1 1 34 3 9 0 0 0 2 3 9 1 0 1 1 6 6 73.5 13.3 84.4 -1.1
Slava M. 2 1 34 5 14 0 0 3 3 6 9 2 1 0 1 3 13 102.2 20.0 91.7 5.1
Blake 2 0 32 0 6 0 6 2 2 0 2 2 0 0 1 1 2 41.4 11.8 118.2 -3.5
Grant 2 0 27 2 7 0 0 2 4 4 11 0 2 0 1 3 6 82.3 17.1 101.4 4.3
Brown 2 0 20 2 7 0 2 1 2 0 1 1 0 0 3 2 5 50.4 23.5 114.8 -8.2
Morris 1 0 18 0 3 0 1 0 0 0 3 1 0 0 2 1 0 11.2 13.3 117.5 -6.4
Rush 1 0 17 4 7 0 0 2 2 0 2 0 1 1 0 1 10 136.7 16.7 80.2 9.5
Russell 1 0 16 1 5 0 3 1 2 0 2 1 0 0 1 1 3 55.2 16.9 88.2 -4.6
Powell 2 0 14 3 6 0 0 1 2 2 4 0 0 0 1 3 7 99.5 25.8 112.1 1.4
McRoberts 1 0 13 0 1 0 0 0 0 2 2 2 0 0 2 2 0 55.5 15.9 125.6 -3.7
Kapono 1 0 11 1 3 0 2 0 0 0 0 1 0 0 0 0 2 88.7 12.9 126.3 0.3
Sampson 1 0 10 0 0 0 0 2 2 1 5 0 0 0 0 3 2 239.4 4.7 75.5 5.0
Murphy 1 0 9 4 4 0 0 0 0 0 0 0 0 1 1 2 8 129.9 17.8 120.4 10.0
Mbenga 1 0 2 0 0 0 0 0 0 0 1 0 0 0 1 1 0 0.0 35.4 97.8 -27.6
Goudelock 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0 126.3 0.0
Yue 1 0 1 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0.0 62.5 116.3 -48.8

Team LeBron
Wins: 11
Losses: 5
Offensive Rating: 107.9
Defensive Rating: 104.4
Best Win: +18.5 Efficiency Diff (12/25/2010)

Player G GS Min FG FGA 3P 3PA FT FTA ORB TRB AST STL BLK TOV PF PTS ORtg %Ps DRtg AWS36
James 16 16 656 153 352 20 66 109 155 25 120 119 30 11 60 41 435 107.1 34.6 102.7 7.3
Ilgauskas 13 10 387 80 150 1 4 39 54 38 98 19 6 15 21 53 200 113.8 23.2 102.6 6.2
Varejao 11 2 259 28 53 0 0 20 33 20 79 6 8 11 6 26 76 119.3 13.3 101.5 6.3
Gooden 8 8 228 30 69 0 1 8 11 15 64 9 5 10 9 29 68 97.5 16.7 102.7 3.4
Pavlovic 9 4 181 25 56 7 17 11 13 3 17 9 5 2 4 17 68 113.8 15.2 110.8 3.9
Snow 6 3 166 9 15 1 3 12 14 1 16 24 8 2 9 15 31 121.3 10.0 106.5 4.5
Gibson 7 2 144 15 32 8 21 7 8 3 7 5 3 1 3 16 45 127.0 11.6 113.0 3.1
Hughes 4 3 135 16 47 4 13 3 10 2 14 12 5 2 14 16 39 70.7 21.3 105.7 -2.0
Williams 3 3 115 21 48 8 12 13 14 4 16 14 2 1 11 6 63 110.7 26.3 113.8 5.7
Bosh 3 3 111 27 45 0 0 9 14 9 30 3 2 1 7 9 63 117.9 23.9 100.9 7.2
Chalmers 3 2 85 10 29 8 20 4 5 1 7 8 3 0 3 6 32 111.8 18.0 103.7 4.2
Wade 2 2 76 15 40 1 5 7 7 3 10 11 5 2 7 4 38 96.2 31.4 98.0 4.1
Marshall 4 0 74 12 24 5 12 4 5 6 18 3 1 2 4 7 33 118.2 19.1 111.1 6.8
Newble 5 2 74 6 13 0 4 0 0 1 8 7 3 2 0 10 12 119.5 8.4 107.9 3.1
Jones 5 1 70 3 14 3 14 0 0 0 5 4 0 0 2 2 9 70.6 10.2 113.2 -0.6
McInnis 2 2 69 13 21 2 5 3 3 0 4 10 1 0 6 6 31 114.5 18.9 109.6 4.7
Hickson 4 2 65 8 17 0 0 8 9 7 21 2 1 2 3 3 24 113.9 18.0 106.0 7.2
West 2 1 60 4 7 2 2 0 0 0 6 7 3 3 3 4 10 111.0 9.1 96.0 5.3
Wally S. 2 0 57 9 16 5 8 0 0 1 8 3 0 0 0 4 23 154.1 11.4 123.4 7.2
Parker 2 2 51 4 12 2 7 2 3 0 6 2 0 2 1 4 12 86.7 12.9 100.3 1.3
O’Neal 2 2 50 10 17 0 0 4 8 3 13 3 1 1 2 8 24 113.2 21.7 95.1 5.6
Dampier 2 1 40 1 2 0 0 0 0 2 6 1 1 1 5 3 2 37.7 9.6 102.5 -1.4
Traylor 2 0 40 6 12 0 0 0 0 4 8 2 1 0 4 6 12 89.6 19.4 104.8 0.7
Anthony 3 1 38 3 5 0 0 1 2 4 8 0 1 1 1 5 7 122.6 9.3 103.2 3.6
Brown 2 0 37 4 7 1 2 1 2 0 3 3 1 0 0 2 10 140.9 9.4 105.6 5.3
Boozer 1 1 37 4 11 0 0 4 4 3 12 1 3 0 3 4 12 88.8 18.3 85.1 4.3
Wallace 2 2 36 2 5 0 0 1 2 7 10 3 3 2 0 3 5 138.3 10.0 113.4 9.8
Battier 1 1 34 4 11 3 7 0 0 3 6 1 1 1 1 1 11 115.4 15.4 104.1 5.8
Murray 1 1 33 8 16 1 3 4 8 0 8 2 2 1 0 3 21 115.6 26.3 98.0 9.2
Miller 3 0 33 5 10 3 5 2 4 4 9 0 1 0 1 2 15 129.4 17.5 104.2 8.9
Harris 2 0 32 4 10 1 2 6 8 1 2 3 1 0 0 1 15 130.6 18.6 108.1 6.9
Williams 1 1 31 3 11 1 3 2 3 0 7 0 2 1 3 3 9 63.6 19.8 86.4 -1.0
Ollie 1 0 30 1 4 0 0 5 5 0 1 2 2 0 0 3 7 136.3 8.6 96.1 3.8
Jones 2 0 30 2 8 2 7 0 0 1 5 1 0 0 0 2 6 90.3 12.4 101.4 1.4
Haslem 1 0 25 4 7 0 0 0 0 2 8 1 0 0 2 0 8 101.8 17.2 102.5 5.2
Moon 1 0 24 6 7 1 1 0 0 0 1 0 1 1 0 3 13 175.1 9.4 96.1 9.9
Bibby 1 0 22 2 3 2 3 0 0 0 2 1 0 0 0 2 6 216.3 5.7 114.4 5.5
Battie 1 0 21 1 3 0 1 0 0 3 11 1 0 3 1 3 2 81.0 9.5 81.7 6.2
Howard 2 0 20 0 2 0 0 1 2 3 5 1 0 0 0 4 1 93.9 9.2 106.8 -0.1
Wagner 1 0 20 4 13 0 3 0 0 0 2 0 0 0 0 1 8 69.7 24.3 99.9 -4.0
Arroyo 1 1 19 1 5 0 3 0 0 0 1 4 0 0 0 0 2 79.8 14.8 101.6 1.0
Williams 3 0 17 1 7 0 2 0 0 1 3 1 0 0 0 3 2 49.7 21.0 105.3 -6.2
Jackson 1 0 16 4 5 3 3 0 0 0 0 2 0 0 1 0 11 163.4 16.8 126.8 14.4
Cole 1 0 16 2 7 0 0 0 0 0 1 2 0 0 0 1 4 86.1 18.9 110.2 -1.6
Wilks 1 0 14 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0.0 3.1 124.0 -1.1
Brown 1 1 13 1 7 0 2 0 0 1 5 0 0 1 0 2 2 41.0 20.4 87.4 -5.0
Kapono 1 0 13 1 4 1 2 0 0 0 1 0 0 0 1 2 3 64.1 15.2 100.7 -5.0
Miles 1 0 8 0 0 0 0 1 4 1 2 0 0 0 0 1 1 75.9 10.8 98.6 -0.2
Wright 1 0 8 0 2 0 0 0 0 0 0 1 0 0 0 1 0 33.8 10.4 130.7 -6.3
Pollard 1 0 7 0 0 0 0 0 0 1 2 0 0 0 0 3 0 205.1 2.5 110.9 -2.7
Curry 1 0 6 2 3 0 0 2 2 2 3 0 0 0 1 2 6 133.9 38.5 105.3 11.8
Moiso 1 0 5 0 0 0 0 0 0 0 2 0 0 1 0 1 0 0.0 82.6 4.3
Jackson 2 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0 128.7 0.0
Kinsey 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0 132.5 0.0

Email Neil at np@sports-reference.com. Follow him on Twitter at @Neil_Paine.

Projected Leaderboards

Filed under: Uncategorized — Kevin Pelton @ 2:53 am

Using the in-season projection method combining current performance and SCHOENE described in today’s column, here’s a look at our projections for the end-of-season per-game leaderboards.

Player               Tm     PPG  
LeBron James        MIA    28.8
Kobe Bryant         LAL    27.4
Kevin Durant        OKC    26.7
Carmelo Anthony     NYK    24.3
Dwyane Wade         MIA    24.1
Blake Griffin       LAC    22.5
Derrick Rose        CHI    22.4
Kevin Love          MIN    22.3
Monta Ellis         GSW    22.1
LaMarcus Aldridge   POR    21.8

Kobe Bryant came into Thursday (not reflected in these stats) with a lead of a point per game over LeBron James, but the expectation is he’ll fall al little bit off his current pace. So too may Kevin Love, who is currently fourth–though SCHOENE can’t account for the Ricky Rubio factor.

Player               Tm     RPG 
Kevin Love          MIN    14.6
Dwight Howard       ORL    13.6
Andrew Bynum        LAL    12.0
Blake Griffin       LAC    11.2
David Lee           GSW    10.4
Anderson Varejao    CLE     9.9
Kris Humphries      NJN     9.9
Marc Gasol          MEM     9.8
Tyson Chandler      NYK     9.7
Marcin Gortat       PHX     9.5

This is Love’s and Dwight Howard‘s category to lose, though Andrew Bynum could make a run if he keeps playing 35 minutes a night.

Player               Tm    APG 
Chris Paul          LAC   10.1
Rajon Rondo         BOS    9.8
Steve Nash          PHX    9.7
Deron Williams      NJN    9.1
Jose Calderon       TOR    8.4
John Wall           WAS    7.9
Ricky Rubio         MIN    7.9
LeBron James        MIA    7.7
Derrick Rose        CHI    7.6
Kyle Lowry          HOU    7.2

The most interesting names here are Rubio, who could easily finish higher with more minutes, and Kyle Lowry. Lowry was averaging 8.7 assists through Thursday after averaging just 6.7 in similar minutes last season, so this method is understandably a bit skeptical he can keep it up.

Player               Tm    SPG 
Chris Paul          LAC    2.6
Iman Shumpert       NYK    2.4
Ricky Rubio         MIN    2.2
Rajon Rondo         BOS    2.1
Monta Ellis         GSW    2.0
Jeff Teague         ATL    2.0
Mike Conley         MEM    1.9
Brandon Jennings    MIL    1.9
Dwyane Wade         MIA    1.8

The NBA added two world-class thieves this season in Rubio and New York’s Iman Shumpert. Rubio is probably the biggest threat to Chris Paul leading the league.

Player               Tm    BPG 
JaVale McGee        WAS    2.7
Dwight Howard       ORL    2.5
DeAndre Jordan      LAC    2.4
Serge Ibaka         OKC    2.2
Andrew Bynum        LAL    2.0
Andrew Bogut        MIL    1.9
Marc Gasol          MEM    1.9
Roy Hibbert         IND    1.8
Tyrus Thomas        CHA    1.8
Josh Smith          ATL    1.7

DeAndre Jordan and JaVale McGee are both averaging three blocks a night thus far, but the conservative projection doesn’t see anyone finishing that high. Jordan in particular could come back to Earth, having averaged just 1.8 blocks per game as a starter last season.

January 19, 2012

They don’t shoot threes like they used to

Filed under: Uncategorized — John Gasaway @ 1:45 pm

This week I penned a brief but heartfelt salute to perimeter-oriented teams, and in the course of pulling up the information necessary for such an effort I got to thinking about how Ken Pomeroy‘s seminal 2008 essay, “Arc Madness,” really needs an update.

Right now you’re saying: “‘Arc Madness’? Seminal essay? Ubiquitous question marks?” You see, Ken’s piece appeared in our first (2008-09) Basketball Prospectus book, meaning it was thoroughly enjoyed by his mom, my mom, and a hoops fan in Italy that’s been contacting me regularly on Facebook for years now. (Buongiorno!) The rest of you, however, may need an introduction to the essay in question.

Ken was writing on the eve of the three-point line’s relocation from 19.75 feet to its current more robust distance of 20.75 feet. As Ms. Pomeroy’s kid is wont to do, Ken predicted the consequences of that relocation with pitiless accuracy: very slight declines in “both three-point accuracy and the frequency of the shot” are indeed what came to pass with the new line in place.

The coolest part of “Arc Madness,” though, was a chart that summed up the history of the three-pointer from 1987 to the present with Edward Tufte-like efficiency. And, thanks to the information contained on what hoops analysts in white lab coats refer to reverently and knowingly as “NCAA page 44,” we can now bring Ken’s 2008 chart up to date.

Behold the strange, compelling, and begging-to-be-adapted-for-the-screen story of the three-point shot in Division I, 1987 to 2011:

In the first season of the three-point shot, 1986-87, just 15.7 percent of D-I shot attempts were launched from beyond the arc. After all, coaches had offenses patterned on what the rules of the game had been previously. As Ken noted in his essay, Rick Pitino was viewed as some kind of wild three-point-crazed innovator that year when Providence rode Billy Donovan‘s outside shooting all the way to the Final Four — but just 30 percent of the Friars’ attempts that year were threes. That distribution of shots today would put a team in the bottom third of D-I in terms of perimeter-orientation.

The other interesting aspect of that first year is how amazingly good teams were at something they almost never did: shooting threes. In 1986-87 D-I as a whole hit 38.4 percent of its threes, a mark that’s never been equaled since. Maybe non-Providence players that year dared to try the newfangled weapon only if they were really wide-open. In any event, more threes were attempted with each passing year, and within the span of just a few seasons the level of accuracy from out there had corrected downward to (and more or less locked itself in at) between 34 and 35 percent.

Then, prior to the 2008-09 season, the line was moved out a foot, a moment that can be seen clearly in the chart above. For the first time since the three-point shot’s introduction, teams devoted a smaller share of their attempts to tries from beyond the arc. Similarly, three-point accuracy in D-I dipped from 35.2 percent in 2007-08 to 34.4 percent in 2008-09.

Which brings us to where we are today. In each of the past three seasons threes have comprised between 32.7 and 33 percent of attempts from the field. That’s just three years’ worth of reality, of course, but it’s worth noting that we’re looking at the most stable such period usage-wise in the history of the three-pointer. Maybe 20.75 feet is the magic distance. The inexorable multi-decade increase in three-point attempts was at last capped, and long-range shooting no longer threatens to overtake the sport as a whole. In this particular sense, at least, the relocation of the three-point line should be classified a success. NCAA Men’s Basketball Rules Committee, I salute you!

BONUS pro bono assignment editing! I continue to be frankly fascinated by Steve Alford‘s season at Indiana in 1986-87, and I’m convinced it contains all the necessary ingredients for a ripping good Michael Lewis-variety long form piece. The essentials are as follows: Radical new rule change in a star player’s senior year; legendary coach renowned and celebrated for a style that predated said rule change; and, most crucially for our pitch, a national championship season.

Heck, I’ve even sketched in the outlines already:

Imagine you’re a star quarterback in college, and you have been for three seasons. Then right before your senior year the NCAA comes in and changes the rules. From now on touchdown passes of longer than 20 yards will be worth nine points instead of just six. That’s more or less what happened to Alford. For his first three years any shot he made from the field was going to be worth two points and only two points. But then the three-point line was introduced in time for Alford’s senior season. And despite Bob Knight‘s subsequent reputation as a coach who disdained the three-point shot, he at least allowed Alford to embrace the new weapon enthusiastically.

Though he was listed at just 6-2, Alford was a career 56 percent two-point shooter (!) entering his senior season. Then the three-point shot arrived, and Alford devoted 40 percent of his attempts to tries from beyond the fancy new arc. His two-point accuracy plummeted (down to 44 percent) but his overall scoring efficiency went up because he made 53 percent of his threes as a senior, including seven made threes in the Hoosiers’ win against Syracuse in the national championship game. That’s the power of the three-point shot, and Alford was one of the first players in Division I to demonstrate its full potential.

It’s all yours, Mr. Lewis.

Twitter: @JohnGasaway. Contact: here.

January 18, 2012

PET for the NBA

Filed under: Uncategorized — Kevin Pelton @ 8:49 pm

Over the course of this week, our Drew Cannon has introduced a new offensive rating for college players he calls Four Pettinella Score, or PET. Using data from Basketball-Reference.com, I applied the same method to the NBA. Here’s the top 10 so far this season (minimum 200 minutes):

Player             ORTG    Usg    PET

LeBron James        121   33.0   120.0
Kobe Bryant         106   39.7   119.2
Carmelo Anthony     107   35.4   117.0
Kevin Durant        113   31.9   116.5
Louis Williams      118   28.2   114.8
Kevin Love          115   28.7   114.4
Andrea Bargnani     111   28.7   113.3
Derrick Rose        115   27.1   112.9
Kyrie Irving        107   29.5   112.8
Russell Westbrook   101   31.6   112.5

As you can see, PET conforms to conventional wisdom a bit better in the NBA than it does in college. I find this instructive in terms of reinforcing that the biggest issue with any kind of comprehensive college metric is accounting for strength of schedule. Applying a team-level adjustment to player stats may not go far enough.

In the NBA, we don’t really have that problem, so even after just three weeks, seven of the top 10 players were All-Stars a year ago. Andrea Bargnani‘s scoring ability has never been the question, and he’s off to a great start; rookie Kyrie Irving might get there sooner rather than later. The only real fluke in this group is Philadelphia’s Louis Williams, a career 33.5 percent three-point shooter who is making 41.7 percent of his triples thus far this season. (By the way, neo-Peja Stojakovic Ryan Anderson is just outside the top 10, ranking 11th.)

This comes as no surprise to me; Four Pettinella Score utilizes the same logic as the offensive portion of WARP. The way I get to individual offensive rating is somewhat different and I make a slightly smaller adjustment for usage (adding one point of Offensive Rating per point of usage, rather than 1.25), but the only philosophical difference between the two metrics is my inclusion of an adjustment for floor spacing. So I certainly think there’s merit to using PET, especially among players with similar schedules.

Referee Ratings

Filed under: Uncategorized — Neil Paine @ 8:58 am

Random question of the day: Which referees are contributing most to offensive or defensive basketball?

To find the answer, I ran a variation on Basketball-Reference‘s Simple Rating System. Like in the SRS, each team was assigned an offensive and defensive rating that minimized the squared error when predicting the outcome of every game… However, I also tacked on variables for all 60 NBA referees, in an attempt to determine their impact on the outcome, independent of the quality of teams playing or home-court advantage effects (assuming referees’ impact on offense/defense is equally allocated to both teams).

Here were the resulting team ratings (all numbers are through Monday’s games):

Team         Poss       Off    Def   Ovr
----------------------------------------
1.  PHI     1200.4      4.5   -7.6  12.1
2.  CHI     1328.8      1.2   -8.4   9.6
3.  MIA     1200.2      0.8   -7.0   7.8
4.  ATL     1286.1      6.2   -0.5   6.7
5.  OKC     1278.8      6.8    0.9   5.9
6.  SAS     1199.4      5.7    0.4   5.3
7.  POR     1222.8      1.6   -3.5   5.1
8.  DEN     1250.6      4.7    0.2   4.5
9.  LAC      907.7      6.5    2.4   4.1
10. LAL     1363.2     -1.1   -4.6   3.5
11. DAL     1285.2      0.2   -3.1   3.3
12. ORL     1073.7      4.6    1.3   3.3
13. IND     1094.5     -2.0   -5.1   3.1
14. MIN     1201.5      1.8    0.7   1.0
15. MEM     1108.7     -3.2   -4.1   0.8
16. CLE     1118.6      1.1    0.4   0.7
17. UTA     1103.7     -2.0   -2.6   0.6
18. HOU     1197.5      3.0    2.9   0.1
19. PHO     1083.5      1.9    3.3  -1.4
20. MIL     1099.2     -3.4    0.3  -3.6
21. GSW     1108.1      0.0    4.4  -4.4
22. NYK     1208.2     -3.9    0.5  -4.4
23. NOH     1142.0     -5.5   -0.9  -4.6
24. BOS     1051.0     -1.1    4.0  -5.1
25. TOR     1232.9     -1.8    3.5  -5.3
26. NJN     1242.3      0.9    7.6  -6.7
27. SAC     1300.9     -5.7    4.0  -9.7
28. DET     1125.6     -2.4    7.5  -9.9
29. CHA     1305.5     -7.3    4.3 -11.6
30. WAS     1196.8    -11.4    0.8 -12.2
----------------------------------------
    HCA                              5.2
    Lg Constant                    103.2
----------------------------------------

And, of course, the effect of each referee:

Referee           Poss  Rating
-------------------------------
KarlLane          798.6   8.7
MarcDavis         729.9   8.3
ScottTwardoski    905.1   7.7
JamesWilliams     932.3   7.1
JTOrr             891.5   7.1
HaywoodeWorkman   839.5   6.0
KaneFitzgerald    992.2   5.9
KevinFehr         713.4   5.8
MichaelSmith      914.9   5.5
EliRoe            839.0   5.3
OlandisPoole      927.5   4.5
KenMauer          819.0   3.9
ZachZarba        1041.7   3.8
MaratKogut        810.4   3.7
LeonWood         1014.8   3.7
VioletPalmer      804.8   3.7
EdMalloy          919.7   3.0
NickBuchert      1027.1   2.7
JoshTiven         935.3   2.2
KevinCutler       934.4   1.9
MarkLindsay       921.9   1.6
DanCrawford       838.4   1.3
DerekRichardson  1021.5   1.2
GregWillard       900.7   1.1
TreMaddox         721.4   0.8
CurtisBlair       886.5   0.8
BrianForte        828.2   0.5
TommyNunezJr      912.8   0.3
EricDalen         913.3   0.3
MikeCallahan      805.8   0.3
BennieAdams      1016.3   0.2
JoeCrawford       914.8   0.0
TonyBrown         890.8  -0.1
PatFraher         856.1  -0.4
TomWashington     916.7  -0.6
DerrickCollins   1023.9  -0.6
ScottWall         842.8  -0.7
JohnGoble         904.6  -1.2
RodneyMott       1005.3  -1.5
GaryZielinski     845.9  -2.2
MarkAyotte        897.6  -2.7
MontyMcCutchen   1000.2  -2.9
RonGarretson     1001.3  -2.9
DickBavetta       827.2  -3.3
BrentBarnaky      730.1  -3.4
DavidGuthrie      661.6  -3.4
ScottFoster       820.6  -3.4
LeroyRichardson   897.8  -3.9
JamesCapers       838.9  -4.1
EricLewis        1000.4  -4.4
EddieF.Rush       731.3  -4.5
TonyBrothers     1006.9  -4.6
DerrickStafford  1025.4  -5.4
MattBoland        821.3  -5.8
SeanWright        821.3  -5.9
BillKennedy       988.4  -6.4
JasonPhillips     801.8  -6.8
BillSpooner       838.3  -7.9
CourtneyKirkland  895.9  -8.8
DavidJones        911.1 -10.7
-------------------------------

How do you interpret these?

Well, for any given matchup, the home team’s offensive rating (aka the road team’s defensive rating) can be predicted by:

ORtg_H = Lg_Constant + .5*HCA + Ref1_Rtg + Ref2_Rtg + Ref3_Rtg + Home_Off + Road_Def

Likewise, the road team’s offensive rating can be predicted by:

ORtg_R = Lg_Constant – .5*HCA + Ref1_Rtg + Ref2_Rtg + Ref3_Rtg + Road_Off + Home_Def

This means, for instance, that games officiated by Karl Lane feature offensive ratings 8.7 pts/100 poss. higher than average, all else being equal, while games officiated by David Jones have ratings 10.7 pts/100 poss. lower than average.

Although the 2012 season is a small sample right now (making the standard errors on these estimates quite large) those are some pretty big referee effects!

Email Neil at np@sports-reference.com. Follow him on Twitter at @Neil_Paine.

January 13, 2012

Ryan Anderson = The New Peja Stojakovic

Filed under: Uncategorized — Neil Paine @ 4:44 pm

This post is meant to expound a little on a couple of tweets I made in the past day-plus:

@Sportsref_Neil
Orlando’s leaders in % of tm shots taken while on the court? Ryan Anderson (27.5%) and Glen Davis (24.2%). Howard (19.6%) ranks 6th on team

@Sportsref_Neil
More Ryan Anderson fun… Peja Stojakovic best year was 2004: 120 ortg/22 %poss/106 drtg/4.7 spm/182 aws+… R.A. 2012:131/21/105/7.5/229

Orlando’s Ryan Anderson is having a crazy start to the 2011-12 season. His offensive rating is 130.8, and while plenty of one-dimensional jump-shooters have crazy ORtgs in limited touches, Anderson is using 21.2% of the Magic’s possessions while on the court, and firing off a staggering 26.6% of their shots while in the game. Like most guys who take so many spot-up Js (Anderson has bombed 59% of his FGAs from deep), he needs others to create for him — hence the fact that 80% of his made shots are assisted — but this level of usage from a shooting specialist is almost unheard of.

In 2007 and 2008, J.R. Smith posted FGA%s of 25.5 and 26.8, respectively, with 59.3% and 57.9% of his FGAs coming from 3-point range. Smith was a different animal, though — he also had touches/min in the 1.0-1.1 range (Anderson is only touching the ball 0.7 times/min), and just 66% of his FGs were assisted in 2008, indicating a greater ability to create for himself. Smith was also an athletic 6-6 swingman; Anderson is a 6-10 stretch 4 whose athleticism doesn’t impress anyone. Dig deeper in the list of players with high FGA% and 3PA%, and you run into a similar issue, with gunner guards like John Starks & Eddie House showing up.

Perhaps the best comparison for Anderson, then, is another tall shooter who wasn’t especially adept at creating for himself but still launched a ton of shots with great efficiency: Peja Stojakovic.

Stojakovic’s best year by just about every metric was 2004, when he posted an ORtg of 119.9 on 22.0% of Sacramento’s possessions while on the floor. He also had a 106.0 individual DRtg, an SPM of +4.72, and a 182 AWS+ (per-minute Alternate Win Score — the best linear-weights metric — compared to the league average, where avg = 100). With Chris Webber limited to 23 games that year, Peja was the focal point of the Kings’ offense despite taking 40% of his shots from beyond the arc and needing assists on 77.1% of his field goals.

Using Mike Bibby at the point and Vlade Divac to pass out of the high post, Kings coach Rick Adelman crafted a Webber-less offense that ranked 2nd in the league thanks to Peja’s eye-popping efficiency numbers. Likewise, the Magic currently have built the league’s 2nd-best offense around the passing of Jameer Nelson & Hedo Turkoglu, and Anderson’s shooting. And most surprisingly, Dwight Howard has taken a backseat in all of this — Superman currently sits 5th on the team in %FGA, taking only 20.9% of the shots while in the game.

Peja proved that a successful offense can be built around a brilliant shooter who can’t necessarily create looks for himself (or others — Anderson is passing on a paltry 26% of his touches, even lower than Stojakovic’s 35% in ’04). But the jury is still out as to whether Anderson can keep up this level of hot shooting. He’s knocking down treys at a 43.4% clip right now, a rate even Peja was only able to match once (2008, when he hit 44.1% of his threes for New Orleans). And remember, that was a run-down 30-year-old Peja, using only 17.9% of the possessions while on the court. Also note that, while Anderson made 39.3% of his threes with a 122.1 ORtg last year, he was taking only 22.8% of Orlando’s shots while in the game. The leap from is 22.8% to 26.6% is a massive one indeed.

Still, Orlando’s scoring fortunes may depend on whether their newfound offensive centerpiece can continue to be as efficient as a prime Stojakovic. He’s posted Peja-esque numbers so far, so it’s going to be fun to see if he can live up to those expectations over an entire season.

Email Neil at np@sports-reference.com. Follow him on Twitter at @Neil_Paine.

Older Posts »

Powered by WordPress