ANALYZING THE SIXERS

Posted by: Jeff McMenamin
01/02/13 8:44 am EST

A relatively new and popular way of analyzing sports teams and players is through the use of advanced statistics.

Around the NBA, you’ve seen teams hiring their own stat guru’s to better evaluate talent based on their findings.  Early in December, the Memphis Grizzlies hired ESPN’s highly popular analyst, John Hollinger, as their Vice President of Basketball Operations.  At the beginning of the season, the Sixers hired Aaron Barzilai, creator of the website basketballvalue.com, as the team’s director of analytics.

The most famous stat-guy in the front offices of an NBA team is Daryl Morey, general manager of the Houston Rockets. He was the first NBA general manager to be hired through the Moneyball concept of thinking.  This past offseason, he famously signed Jeremy Lin and Omer Asik while also trading for James Harden.  The Rockets are now one of the most intriguing and exciting teams to watch in the League, currently holding the NBA’s second-best offense.

Being relatively new to the subject of advanced statistics I couldn’t think of a better way to both enlighten myself about the topic and to give readers a better understanding as well then to interview one of the more popular number-crunchers out there today.

Meet Daniel Myers, administrator to the widely used basketball analytics forum apbr.org, which was started by Dr. Dean Oliver, a former statistical analyst for both the Seattle Supersonics and Denver Nuggets.  The forum features conversations between statistical experts such as Mike Zarren, who over the summer had interviewed to become the Sixers general manager, Joe Sill, who created the regularized adjusted plus-minus system and serves as an analytics consultant for the Washington Wizards, and Jon Nichols, who owns the website basketball-statistics.com and who serves as the manager of basketball analytics for the Milwaukee Bucks, among others.

Myers, being a big fan of the website wagesofwins.com when it first came out, started creating his own advanced statistics which have been refined and perfected over the years.  The advanced stats that he created are advanced statistical plus-minus (ASPM) and value over replacement player (VORP).  The full explanation of these statistics can be seen here.  By clicking on the Sixers logo through this link, you can see the breakdown of how each Sixer ranks this season based on his analytical models compared to the rest of the league.

NBA stat expert Daniel Myers’ charting of the ASPM and VORP ratings which he created.

The data shows that Thaddeus Young, Jrue Holiday and Jason Richardson have all played above the league average in their positions this season.  It also shows you that Jrue Holiday is the most crucial piece to the Sixers offense, while Thaddeus Young has shown to be the most irreplaceable player on the Sixers roster.

Defensively the data tells you that Jrue Holiday, Evan Turner and surprisingly Nick Young have been the most productive on the team so far this season.  It also shows that Evan Turner and Dorell Wright have both played average as a whole this season while Lavoy Allen and Spencer Hawes have played widely below average.

Myers explained to me why fans should be taking systems like the ones he created more seriously in terms of evaluating a basketball player.

“Obviously you can’t watch all the games.  Think of stats like looking at a bridge,” said Myers. “I can look at a bridge and tell you by looking at a lot of them whether or not they’re going to stand up or not, but there are ways to more accurately determine reality based on mathematical formulation.  And there are ways similarly in basketball to more accurately explain things because the human eye is fallible.”

“The numbers help provide a non-biased perspective on players or really on anything if the statistics are done properly.”

The advanced stats which I’m most familiar with are WS/48 (wins shared per 48 minutes), Roland Rating, RAPM (regularized adjusted plus-minus) and VI (versatility index).  Before going further in my interview with Daniel, I cleared up which of these stats are the most accurate and what I should take into consideration when looking at these numbers.

WS/48: “I would call the best box score stat available. I’m also partial to the one I created as well, but to the widely used stats out there that’s the best of the bunch.  The biggest issue with offensive ratings and win shares is that they don’t account for usage enough.  Offensive rating is looking at how efficient a player is with the shots they take … There is a value to taking many shots, despite what wins produced or wins shared might tell you about a player.”

Roland Rating: “For what it’s worth I would kind of ignore Roland Rating.  Even Roland Beech its creator doesn’t even care too much about it.  Roland Rating isn’t an adjusted plus-minus at all, which creates extremely high variance. Non-adjusted plus-minus is borderline unusable.”

RAPM: “RAPM is fairly useful, but it doesn’t stabilize extremely fast…If you look at Turner’s RAPM rating for example, you wouldn’t be able to really judge what Turner has been doing this season that effectively. RAPM uses the data from the trailing years, so Turner’s rating in that case wouldn’t change very fast.  Joe Sill, the original developer of RAPM, and Jeremias Engelmann his protege, evaluate a players data from his oldest season to the newest.  When a new season begins Engelmann will regress that players data towards the number from the previous year, instead of regressing the data towards the mean. It can be a little funky when you’re talking about young or old players where their data is trailing up or down. With Evan Turner you’re seeing an under-representation of what he’s doing because he (Engelmann) doesn’t factor in age with his data.”

VI: “It’s a good way to characterize a player.  I don’t think it provides a very good understanding to how valuable a player is to his team.  A stat like wins shared however does include things like versatility so it’s a more accurate representation of that.”

Here’s how the Sixers are rated in terms of each statistic:

Player

WS/48

Roland

Off RAPM

Def RAPM

VI

Thaddeus Young

0.13

7.8

2.46

1.07

6.4

Dorell Wright

0.1

-0.4

-1.16

0.41

6.9

Jrue Holiday

0.09

5.8

0.16

0.54

9.2

Spencer Hawes

0.09

-4.3

-1.5

0.03

7.7

Royal Ivey

0.09

-4.7

-3.22

-0.88

4.3

Lavoy Allen

0.09

-2.7

-1.6

1.49

5.9

Jason Richardson

0.08

3.5

0.1

-0.67

5.9

Kwame Brown

0.08

-8.3

-1.72

0.47

4.7

Arnett Moultrie

0.08

-10.9

-1.7

-1.57

N/A

Evan Turner

0.07

-0.2

-1.46

-1.06

8.3

Damien Wilkins

0.06

-5.5

-2.31

0.61

N/A

Nick Young

0.05

-2.7

-0.84

-2.13

5.2

Maalik Wayns

-0.16

-20.1

-1.86

-1.88

N/A

As Myers pointed out earlier, WS/48 is probably the best widely-used stat to judge a player.  Obviously Kwame Brown, Royal Ivey and Arnett (D-League) Moultrie haven’t been more valuable to the Sixers’ success this season than Evan Turner has, so why does Turner sit in 10th on the team in WS/48?

“He’s (Turner’s) a very versatile beast.  The main issue is that he just doesn’t shoot very well,” says Myers. “(Andre) Iguodala is the type of player that Evan Turner should try to become which is someone who doesn’t have to shoot a million shots.  He should focus on doing all the little things more and play good defense instead of worrying so much about his offense.  In my opinion Turner should be a lower usage player.  He should be filling in for rebounds and assists, but probably taking fewer shots.  The problem with the Sixers is that there are just not a lot of highly efficient players on the team.  If he did take less shots though you’d see his rating go up.”

Turner is second on the team in shots attempted with 421.  Even while missing four games due to injury, Holiday still leads the team with 465.  As Myers had said about WS/48 before, there is a value to taking many shots despite what wins shared might tell you. He talked about how Holiday’s rating is slightly higher than Turner’s because of the amount of assists he has compared to his shot attempts.

“When I run regressions to develop stats myself, it’s interesting to look at assists and usage.  The value of a players assists is directly related to how many possessions he uses otherwise,” said Myers.  “So if a player is dishing out five assists but only taking a couple of shots, those assists are typically not very valuable.  But for a player who is taking 15 shots a game, those assists are of high value.  Using Kobe for example, he’s not dishing it unless somebody is wide open. So when his assists occur it’s generally an indicator of that value. (Holiday) not only gets a lot of assists, but he has a very high usage rating.  He’s using 26 percent of the possessions on the team.”

As a team the Sixers are 21st in the League in offensive efficiency.  Their point differential of -2.3 is also 21st.  It’s hard to really thrive as a player in WS/48 with a team as inefficient as the Sixers.  Using the Dallas Mavericks as an example, who have the 22nd ranked team offensive efficiency rating in the NBA, you can see how a teams inefficiency can really have an affect on a players WS/48.  Of players on the Mavs who have played more than 500 minutes this season, just five have a WS/48 average of above .07 compared to the Sixers who have seven.

Now compare these two teams to the Oklahoma City Thunder who are ranked first in the League in offensive efficiency. Of the players who have played more than 500 minutes this season for their team, six players have a WS/48 rating of above .12.  Thaddeus Young is the only Sixer with a rating of above .12.

Myers talked about how the loss of Lou Williams in the off-season for the addition of Nick Young has taken a considerable toll on the Sixers offensive efficiency.

“Lou Williams was the best offensive player on the team last season.  I was really surprised when they got rid of him and brought in Nick Young in the off-season, because Nick Young is notoriously inefficient,” said Myers.  “Most of the advanced stats people would agree that he (Nick) probably shouldn’t even be playing in the NBA.  Unless you want to put him deep on your bench as somebody who can come in and chuck up a lot of shots when you’re way behind.”

Myers thinks just based on the teams offensive efficiency problems alone, the rest of the season doesn’t look too bright for the Sixers.

“The teams point differential is quite poor,” said Myers.  “It’s never good when you’re being out-scored, especially by a considerable margin.  Unless something drastic happens I see the Sixers finishing with around 36 or 37 wins this season.”

Despite the teams woes both offensively and through injury, they did just have one of their best defensive performances of the season last night to beat the Lakers and improve to 15-17.  With a healthy Andrew Bynum possibly returning in the beginning of February, and close to half a season left for him to play, 20 wins in 40 games doesn’t sound that out of the question.  Even though Myers doesn’t see too much hope for the rest of this season, he believes the Sixers organization could possibly have gotten in Bynum the perfect player to build around.

In the mind of Daniel Myers, Andrew Bynum’s impact on the Sixers could be a cornerstone for the franchise.

“There are two types of players who provide very good value to a team.  Players on rookie contracts and max players,” said Myers. “Both of those players provide significant value beyond what their contract is worth…What I think people need to look for in terms of a player is their salary.  Consider what they produce versus their salary.  The Sixers right now are trying to find those underpaid players.  Jrue Holiday’s very valuable, he’s probably worth $15 million a year and Thaddeus Young is close to that.  Turner’s probably worth about $7 million a year, but you have to realize that these are still young players and Turner in his prime at 26 will probably be worth a couple more million a year.  Holiday will probably be worth about $5 million more once he reaches his prime and Thad at least a couple million more as well.

“Try to recognize how much these players are paid compared to their value on the court.  Bynum could be the cornerstone player for the franchise and if he is he’ll be worth way beyond his max contract.”

The season and franchise after all do hang in the balance of Andrew Bynum’s knees and he said they’re feeling stronger now which is a good sign.  All that can happen now is the waiting game.  If Bynum pans out and returns to the court fully healthy, the team won’t need to play for the lottery to survive.  Not only that, but according to Myers they’ll have at least two more very valuable players in Thaddeus Young and Jrue Holiday locked into deals for the long-term as well.

Since Jrue Holiday’s return from injury, the Sixers have stayed competitive in every game win or lose. It’s hard not to think about what the outcomes would’ve been if big number 33 was out on the court with them during this stretch. In just 13 more games and 29 more days, the Sixers will find out the question they’ve been waiting to have answered all season. Can Bynum play? If he can, the NBA better focus its eyes on the city of Philadelphia because it could be the start of something to be reckoned with for years to come.


 
 
 

24 Responses to “ANALYZING THE SIXERS”

  1. Steve Toll
    2. January 2013 at 10:48

    This is mostly what I have been saying for months, Why didn’t Jeff just interview me?

    Unless this guy charts his own defensive stuff, take it with a huge grain of salt.

    Thad, best player on team. Holiday, most important player. J-Rich, 3rd best player

    Versatility and Quality are 2 different things which this guy fails to understand is his assessment of Evan Turner, who is still not good.

    As for the ridiculous notion of what a players “worth” is, this guy fails immensely. There is a salary cap, each player has a salary, a players worth is relative to what he is paid+produces, and what others are a paid+produce.

    I GUARANTEE that Jeff talked with this guy about Iguodala and purposely left that stuff out because it would agree with everything I’ve written since I came on board AND people would freak when the guy said Iggy was worth 20+ million.

    As for the 76ers. A bottom 10 team, which they are, will not suddenly become a top 10 team with the return of Andrew Bynum and I am sure, that the guy said that as well.

  2. Jim
    2. January 2013 at 11:08

    Steve Toll should read this and explain to me how Hawes is more important that Jrue Holiday

  3. Joe
    2. January 2013 at 13:32

    Good article, I really enjoyed it. It pointed towards (although didn’t explicitly state) something that I am a believer in:

    Diminishing marginal returns – Kobe doesn’t shoot at the highest clip, but he has value because he’s able to get those shots in the first place. If Kobe had a small role in an offense he’d shoot a far better percentage but it would be to the detriment of other players on the team. If Kobe only takes open shots inside 5 feet (for example) then suddenly there’s a lot more shots to go around for World Peace, Gasol and Nash – some of which would be contested jumpers near the end of the shot clock. Kobe scores these comparatively very well, I don’t think the players I’ve named here would. Volume scorers like Kobe are useful in the NBA.

    At present, although all of the stats are useful in different areas, none seem to capture this fact very well. It’d be fascinating to see a stat that can truly take account for this, i.e. would it be possible to estimate the detriment to the team if D12 were taking all of Kobe’s shots, and vice versa? I think this is the next step in ABPR Metrics, and I think if someone could come up with a stat that accounted for this then the field of study could really move forward as it would be accounting for the fact that basketball is a game of five-on-five, not five games of one-on-one.

    Kind regards,

    Joe.

  4. DSMok1
    2. January 2013 at 13:35

    Hey, Steve–thought this would get a comment from you.

    1) Defense is very hard to measure, and is best done by charting. Even then, if you don’t know the assignment, attributing credit is rather difficult. None of the box score stats capture more than half of what goes on at the defensive end.

    Take all metrics with a big grain of salt regarding defense.

    2) Versatility and Quality are certainly two different things. Evan Turner is, in my opinion, a below-average NBA player.

    3) The conversion of a metric to a salary valuation reckoned dollars is not particularly hard, and is common knowledge in sports statistics circles. The trick is the validating and choosing the right metric.

    4) Iguodala was probably worth around $20 million, I agree.

    Daniel

  5. DSMok1
    2. January 2013 at 13:42

    Joe:

    Like this?

    http://www.48minutesofhell.com/spurs-stats-skill-curves-matt-bonner-usage-efficienc

    Daniel

  6. Steve Toll
    2. January 2013 at 14:38

    Joe,

    Kobe killed the Lakers last year because of his shooting and this year, he is not. That should give you an idea of things.

    Daniel,

    I agree with all your points there. Not sure how much you’ve previously followed Philadunkia but I am a one man band of sorts.

    People are prolly gonna go berserk over your ET and Iggy comment but it is obviously spot on.

    This article validated much of what I have said about the 76ers and I thank you for that.
    Even a healthy Bynum isn’t going to make this team a contender and the team better grab one helluva player outta the lottery this year

  7. DSMok1
    2. January 2013 at 15:22

    As Daryl Morey has long said–the worst place to be in the NBA is a 40 win team. You aren’t good enough to contend and aren’t bad enough to get the high value (underpaid) top lottery picks.

    The only way out of the treadmill is either to get a max player (that is still underpaid) or blow it up/tank.

    The Sixers bought a lottery ticket with Bynum. Iguodala is a great player, but he was paid like one. To contend with a salary cap, you must have some value. OKC has it with both underpaid max players and underpaid rookie-scale players. Miami has it with underpaid max player(s?).

    Bynum is either a) a potential underpaid max player or b) a ticket to blow up the team when you let him go.

    Just so long as the Sixers don’t end up with a chronically injured OVERpaid max player.

    Daniel

  8. Datruth4life2.0
    2. January 2013 at 15:29

    There are lies, damn lies, and statistics. I beleive in advanced stats, but I believe more eyes more. Jrue and Turner are blossoming, Thad is an undersized 4 who is not a starter on a championship team and the 76ers have the worst frontcourt (without Bynum) in the NBA. I don’t need advanced stats to tell me that. And when he said Turner should try to be like Iguodala, I said, WTH! Does that mean he wants him to be an overpaid role player that shoots 50 percent from the free throw line and dribbles the ball off his foot in crunch time while complaining that he doesn’t get the credit he deserves? No thanks.

  9. Steve Toll
    2. January 2013 at 16:54

    Datruth4life2.0

    Fact or Fiction: Holiday is likely to have a peak higher than the primes of guys like Andre Miller, Bibby, Devin Harris, Calderon, B-Diddy?

    Fact of Fiction: ET is the exact same player as his first 2 years except he hasn’t regressed to the mean on his absurd 3pt shooting?

    Fact or Fiction: Your have NBA League Pass

  10. Jon in LA
    3. January 2013 at 01:18

    “Why didn’t Jeff just interview me?”

    If this isn’t the most telling thing about Steve Toll’s posts here, then I don’t know what is.

  11. Steve Toll
    3. January 2013 at 08:43

    Jon in LA,

    Twas’ sarcasm, I don’t do interviews

  12. Joe
    3. January 2013 at 12:06

    Daniel,

    Thanks – that was exactly the sort of thing I was thinking about – had no idea this had been done yet! A little surprised at the author’s choice of functional form as I wouldn’t expect the skill curves to necessarily be straight lines, but he clearly knows his stuff and it was an engaging read.

    Steve,

    Thanks also for your comment. I was actually using Kobe as a hypothetical example – obviously he has had seasons where there probably were opportunity costs to his shooting at such high volumes, however I suspect in his good seasons this was not the case. Now I think about it though – out of interest, how many games do you think the Lakers would have won last year if Jodie Meeks started at SG for them?

    All the best,

    Joe.

  13. Steve Toll
    3. January 2013 at 12:15

    Daniel,

    The 76ers didn’t buy a lottery ticket with Bynum. They purchased the opportunity to never win 50 games with a “SUPERSTAR”. My immediate response to the trade was 3rd out of 4th and could soon be 4th in the deal. Vucevic and Harkless are NBA players. Jason Richardson is almost 32 years old and signed through 2015 at 6 mil. The team has no cap room and 0 cheap talent, Maalik and moultrie are just cheap. That is why they can’t miss on this years lottery.

    Without Bynum, they are a bottom 10 team. With Bynum, they are not a top 10 team.

    The team must grab a guy who can contribute in the Lottery this upcoming draft because the cap situation doesn’t improve going forward

  14. steven
    3. January 2013 at 13:03

    I just read that Sacremento is considereing trading Cousins for Vucevic and Harkless….so we posibly could have kept Iggy and had Cousins..and could have kept LOU….that would suck…Cant we trade Hawes and turner for Cousins?

  15. steve toll
    3. January 2013 at 15:36

    Joe,

    I actually got crushed last year for saying Meeks > Kobe last season on the Lakers…..

    Steven,
    The Magic don’t deserve such good fortune.

  16. Jim
    3. January 2013 at 19:52

    Steve:

    You’re calling Holiday the most important player on the team and yet you were ready to let him walk in free agency. Wow.

  17. Steve Toll
    4. January 2013 at 10:43

    Jim,

    Bargnani is the most important player on the Raptors because when he plays suck but when he is out they are good.

    If you think that paying near max this offseason if they didn’t do his 4yr-41mil extension to Jrue “can’t crack .100 WS48″ Holiday is a good idea, well, good luck with that.

  18. Jim
    4. January 2013 at 13:56

    Steve,

    Here are your words: “Bargnani is the most important player on the Raptors because when he plays suck but when he is out they are good.”

    None of that makes sense.

    In case you were trying to compare the impact each player has on their team, the Raptors have won more with Bargnani out of the line-up and the Sixers lost all of the games Holiday missed.

    I was in favor of the extension because I’m confident that Jrue would have received a higher offer from another team in the offseason. You’re the one that would have let the team’s most important player go without compensation.

  19. Joe
    4. January 2013 at 15:52

    Steve Toll,

    I know – that’s why I’m asking.

    Have a great weekend,

    Joe.

  20. Rob
    5. January 2013 at 14:10

    People keep saying this team isn’t as good as last year, I don’t think that’s true. If it weren’t for last years shortened season the sixers wouldn’t have opened 20-8. After that they were average at best. Then rose goes down in the playoffs and we steal a playoff win, if it weren’t for the continuity from the team last year while other teams figure out how to play with eachother we don’t make the playoffs. That being said I think this team with Bynum is a top 4 team in the east.

  21. Steve Toll
    5. January 2013 at 17:26

    Rob,

    The team makes the ECF last year if DC doesn’t play Turner in the Boston series. He was historically awful.

    While it is highly unlikely (less than 10%) that the current team is top 4 in the East with Bynum. They would still only be the 12th best team in the NBA.

    Last years team was a top 6 team that MASSIVELY underperformed in the regular season because DC was terrible

  22. Rob
    6. January 2013 at 01:56

    Steve toll,
    Evan turner single handily brought us back from 20 in game 5 if I’m not mistaken, hits stats were not efficient, but he willed the team back and really set the tone for the rest of the team. If he doesn’t play that never happens and we lose 4-1

  23. Rob
    6. January 2013 at 02:08

    Last years team for the first third of the season might have been a top 6 team, while others were gaining chemistry from the lockout shortened season(no preseason, not many practices between games). To say that the sixers were a top 6 team last year is laughable, thunder bulls heat spurs celtics clippers lakers. Not to mention all the tier 2 teams that would beat us matchup wise, pacers grizzlies jazz Knicks. I’m sure I forgot one or two, doing this off te top of my head, those teams would put us around 7-8th seed either conference… Right where we finished. You can hate on dc all day long, players play the game. What would you have done differently? Your hall of fame player iggy would still not have gotten past the first round without a timely injury from d rose. That’s not very good for a hall of fame player. You know what your stats can’t argue… Wins/losses. Last years team was virtually the same the year before it and a 7th seed playoff entrance. I could make an argument that if every coach was better than they actually are their teams would perform better, but the zen master can’t coach every team and at some point it’s on your players, specifically your so called hall of fame player.

  24. steve toll
    8. January 2013 at 15:32

    Rob,
    Keep on Trucking

Leave a Reply