A fantastic read from two of the sites friends, Arsine Khayoyan and David Dickey.

Basketball fans often groan when they open the NBA page to find that a sports writer’s ad hoc statistical gymnastics have the Jazz favored for the championship, or the Bucks with a higher power ranking than the Spurs.

It’s been said that “an unsophisticated forecaster uses statistics as a drunken man uses lamp-posts – for support rather than for illumination.” So as avid followers of the NBA, we decided to implement some of our skills as trained economists to our favorite team, the Los Angeles Lakers.

When looking at the recent statistical formulas concocted to analyze the game, it strikes us that often times one tries to make what they do appear far more complicated than it really is. This renders them ‘special’ and ‘different’ than everyone else, just like mom used to tell them. At the same time, simple averages, margins of victory etc. are hardly significant in any scientific manner.

Thus we decided to split the difference and do a statistical regression model and see which variables are significant on a scientific level. The variables used are pretty simple; they are comprised of the numbers seen in a box score after each game.

Specifically, we looked at…

• Number of field goals attempted
• Field Goal Percentage
• The difference between the amount of 3-point shots attempted by the Lakers and opponent
• The difference between the amount of free throws attempted by the Lakers and the opponent
• The difference between offensive rebounds
• The difference between defensive rebounds
• The difference between assists
• The difference between steals
• The difference between blocks
• The difference between turnovers
• The difference between the number of fouls assessed to each team

All in all, 11 variables were used for all 82 regular season games plus the 5 playoff games against Utah. These regression results give us better understanding than most of the other statistics used out there today. The most significant variable for a Lakers victory was the difference in turnovers. This range was -13 (Opponent had 13 more turnovers than the Lakers) to 10 (Lakers had 10 more turnovers than opponent). The probability for victory in this scenario is 99.99% and 21.56% respectively. When the amount of turnovers is even for both teams, the Lakers still have a 96% chance of winning, ceteris paribus.

The next most important variable was the difference in assists. Obviously, the more assists the team has, the more baskets are being made and the chances of winning are higher. When opponents have 10 assists or more than the Lakers, the probability of winning the game is 60% or less whereas if the Lakers have 10 or more assists than the opponent, the probability of winning is 99% or better. When assists are low and turnovers are high, the probability of winning is under one percent (regardless of how well the other stats may be). Likewise, when assists are high and turnovers low it’s very close to 100%.

Some of the other results were somewhat surprising. According to the model, the number of 3-point shots attempted or the number of offensive rebounds is irrelevant and has no significant effect on the outcome of a game. We would have assumed that as the number of 3-point shots attempted increased, less high percentage shots would be taken and it would have a significant affect on the outcome of the game (especially considering how much emphasis Coach Phil Jackson puts on it). However, apparently the number of 3-point shots doesn’t matter so long as they’re good shots.

On the rebounding side, offensive rebounds proved to be a far more insignificant stat than defensive rebounds. During the first round playoff series against Utah, many Lakers fans (us included) were screaming at the Lakers bigs to box out and bemoaning Boozer’s forearm shivers to Pau Gasol’s lower back. However, at the end of the day, as long as the Lakers have an advantage on the defensive rebounds, what happens with the offensive rebounds is not very relevant. When the Lakers have a double digit advantage on the defensive glass, the probability of victory is 99% whereas when they have a double digit deficit on the defensive glass, it’s only 64% or less.

Finally, free throws, free throws, free throws. Looking at differences of free throws is usually the common excuse most fans use as a reason for a loss. According to the results of the regression model, free throws attempts are only significant at the 90% level, not at the 95% level that the previously mentioned stats were. In laymen terms, getting to the free throw line is important, but not crucial for victory. For example, even when an opponent has 10 more free throw attempts, the Lakers still have an 80% chance of winning the game. Only when the Laker’s opponents shoot 17 additional freethrows, does the probability for winning go under 50%. This may be cause for some alarm should the Lakers meet up with LBJ in the Finals.

In conclusion fellow Lakers fans, be more upset at one-on-one play and turnovers than not getting a free throw from the ref. The stats for the 2008-2009 Lakers show that team ball, reducing the turnovers, and securing defensive rebounds are the main keys to victory.

  • shane bien

    this was unbelievable. Thanks!!

  • kb24bestever

    Great stuff.

    GO LAKESHOW.

    damn am mad the bulls lost thats ma team in the east.
    and dont get it twisted because i bleed purple and gold but you know i just wanted them to win and specially cuz they were playing the lakers.

  • kb24bestever

    i ment the celtics ma fault.

  • Deus_Ex

    I WANT THE CELTICS!!!!

    I want the Celtics to come out of the east so we can finish them off once and for all. Granted I’d love to see Kobe mop the floor with LBJ as we win our 15th Championship but it won’t be as sweet as Kobe sticking it to Paul Pierce.

  • http://powapps.com/ilakers/ iLAKERS

    This is great!

  • chad

    In summation, the Lakers are the best team in the NBA and will win the championship this year. how yah like them apples…….Booyah bringing back the trophy.

  • xfellerx

    spss?

  • xfellerx

    also: title should be lakertistics or laker-tit-sticks.

  • Lakers 24 7

    I’m guessing Arsine Khayoyan is Armenian? hehe

  • Potorik

    WOW!!! This is by far the greatest, most fantastic article I have ever read in my life!! I am definitely going to completely alter my life and career path to become a Lakers statistician!!!! GREAT article!!!

    GO LAKERS!!!!!!!!!!!

  • Potorik

    Probability of me loving this article: 100%
    I am going to print this out and show it to my kids

  • http://www.visionofthefans.com Chris Manning

    Arsine & David are great friends of TLN! In fact, we all watched Game 3 together; good times!

    Fantastic article you guys!!! Wow!

  • PauLAsol

    what about fgm% ?? everytime i see lakers having a bad night shooting i feel its going to be a 51% chance of winning

  • imfasterthanur

    [Comment ID #69717 Will Be Quoted Here]

    I was thinking the same thing.

    Studying Kinesiology, I hated that class lol. But this article made me hate it less ;)

  • Arsine Khayoyan

    We used STATA, instead of SPSS, but they’re basically the same thing. Also, field goal percentage was significant, but it was highly correlated with Assists, which makes sense.

  • kb24

    Great article. However, I feel like defensive rebounding and assists are obviously 2 of the most important stat categories. A) The more defensive rebounds you get, the more shots are missed, and the number of defensive rebounds directly correlates with opp FG%. B) The more assists you have, the higher your FG% and FGM total will be. I enjoyed reading the scientific explanation and trying to figure out all the math involved, but it seems to me it also just takes a little common sense.

  • LEADERFISH

    [Comment ID #69717 Will Be Quoted Here]

    how do you feel being 7th? lol

  • Armen

    Arsine are you from glendale? lol

  • ja

    No offense but this has a very John Hollinger-esque feel to it. And I abhor him.

    It seems you’ve done your best to avoid that classification–scrapping margin-of-victory, etc.–but it must be noted that a statistical analysis of anything with so many variables is an inherently difficult thing to do.

    There are some issues I have (maybe they can be resolved if you walk me through it): you quickly gloss over offensive rebounding and mention that it is seemingly inconsequential (“irrelevant”) to the outcome of a game. How can this be true? Maybe for the Lakers who have an extremely high ranking FGM% and generate a lot of points. But I feel like this type of stat, when abstracted, could possibly be erroneous. Would this model of examining data stand up to other teams in the league? And, are you simply considering the victory being the end result (stay with me)? Would a 1 point victory, in which the Lakers get out-rebounded 10+, be considered the same as a 10+ point blowout victory?

    As well, after reading the article, it states some no-brainers: don’t turn over the ball, create more assists. But I guess it is Lakernomics 101 ;)

  • pauu

    love it. I often think how cool it would be if you could have this kind of statistical report for each team instead of the stupid combined ‘statistics’ that someone like hollinger uses.

    thanks

  • Mitch4Pres

    good job. F U JOHN HOLLINGER

  • xfellerx

    [Comment ID #69741 Will Be Quoted Here]

    you know it kinda sucks. dont worry. it isnt the last you will hear from me. ill be first for years to come.

  • http://www.aol.com LoveArsine&Dave

    Unbelievable – now its time to take it to the next step…ya’ll gotta bring FG% into the equation, and it’s there! I cannot believe how close y’all are. Are you gonna apply for jobs with the org.? you know you should cause we need more of this. You make a dynamic team and I know that OUR team can benefit from what you’ve done – now, break it down sommore. Oh yeah!

  • http://myspace.com/sonicussj Whatsa

    [Comment ID #69741 Will Be Quoted Here]

    How do you like knowing that you did not know what SPSS was and was totally relevant towards the article?

  • WifelovesLuke

    “Deus_Ex
    May 2nd, 2009 at 9:58 pm
    I WANT THE CELTICS!!!!

    I want the Celtics to come out of the east so we can finish them off once and for all. Granted I’d love to see Kobe mop the floor with LBJ as we win our 15th Championship but it won’t be as sweet as Kobe sticking it to Paul Pierce.”

    I couldn’t help myself for rooting for Boston in Round 1. I too, want a rematch with them. Sticking it to the thugs that they’ve become would put such joy in my heart. lol