I wrote the article below in 2004. Almost 6 1/2 years ago. I went back to read it because I wanted to know if the science has changed at all or if the building of an NBA team has changed due to the science. The answer is no and no. The science is the same and in fact, the input of numbers into building a team is diminishing and being displaced by non traditional qualitative factors
Sure many of us across the NBA have spent a boatload of money on “sabremetrics”. It has not been valueless. We are better able to coach the teams we have because of all the information. But the reality is, we can not put together teams based on stats. You can take all the PER, WS48, WP48, Adv plus/minus and the rest and when you add them together the day before the season starts you still know nothing more than the minute before you added them all together. They are meaningless when it comes to putting together a team.
Why is this the case ? Two reasons: coaching and chemistry.
Each coach has a different style and each player contributes differently depending on the players around them and the style of play and coaching. How a player on another team will fit into the coaching environment and system of your team can not be answered by stats.
Add to that the material impact chemistry has on a team, where one nucklehead hurts, two nuckleheads kill and bad chemistry ruins a season and the numbers and analysis quickly become meaningless. You can try to understand both coaching and chemistry, and we continue to experiment with new ways to do so, but you can’t quantify either.
Until you can quantify coaching and chemistry, you can not use the numbers to build a team. Period end of story. You can use them as partial input along with scouting and other elements, but there ain’t no Moneyball solution or the NBA and I don’t see Bill James walking through that door with a solution. Stats will continue to play a role in lineups, matchups and trends, but teambuilding, not so much.
Of course there are other elements that we are rapidly expanding at the Mavs that go into our team-building methodology, but I’m keeping all that to myself 🙂
But for shits and grins, here is my article from April 2004
The question comes up all the time. What is the NBA equivalent of MLB’s Moneyball approach? Are there stats that can be used to come up with a better model for building an NBA team? The answers are Yes and No.
Yes, there are stats that are out there that could be used to better build an NBA team, but No, they can’t be used for building an NBA team, because the stats that most likely most correlate to a player and team’s success are not being collected.
I’m not here to say I know exactly which variables independently, or collectively equate to getting a competitive and financial advantage. Only actual testing will determine what works. I will say which stats I think are most important, and let you know that its more than just a little bit of a logistics challenge to try to accumulate accurate data.
Here are the stats I think the Mavs will need to figure out how to collect as a first effort towards determining which have the greatest impact on success:
- Deflections, Deflections for turnovers
- Defensive Penetrations Allowed, Defensive Penetrations stopped (did you stop your man from penetrating in the paint before he shot or gave up his dribble)
- Assists in paint, Assists outside of paint, within each, assists leading to jumpshots, assists leading to layups, assists leading to foulshots, and within each of these, are they part of fastbreak or not.
- Rebounds in traffic, Rebounds from free throws
- Shot percentages – location zone of shot, and within each, whether guarded or open
- Turnovers – forced, unforced, rule violation turnovers by type (traveling, palming, etc.)
- Touches – How often, where, duration in seconds, conclusion (pass, assist, shot, turnover as a percentage of total touches)
- Charges taken, blocks given
- Blocks above head, blocks that were strips, fouls on block attempts by each
- Turnovers caused – steals, forced by defense (i.e., preventing your man from crossing 8 sec line, or guarding your man for more than x seconds leading to 24 sec clock violation)
There are obviously more, but these are a good starting point to see what works. Unfortunately, this data is not always available just from tape, so it would be necessary to have multiple peopleat the scorers table at the game to see it, confirm it and get help on it. That’s what makes all this so difficult to collect. It may well be that we need to add cameras to each gym that can cover all the action and then go back and determine the information. Either way is expensive, which means it could be along time in coming to the NBA.
32 thoughts on “Building Teams Using Quantitative Analysis in the NBA – The Last Rites”
Pingback: Is Quantitative Analysis of the NBA (APBRmetrics) Dead?
hopefully, the Mavs QA is looking at Rasheed Wallace. This one’s a no brainer. Not only can Sheed still contribute. He’ll help develop the skills of your other big men.
Comment by darryl3 -
Pingback: Mark Cuban - NBA Owner
The problem with your analysis is that you are assuming that the existing methodologies that you are familiar with are the only formulas that can properly use these type of techniques. There are benefits to be had from the type of formulas you mentioned and yes, of coarse, everything is limited by chemistry and coaching…but only so far. There are also may other ways to use number analysis if you are a clever enough and a true creator of process and not just a “moneyball” disciple.
I don’t have a stats PhD. but I am a math genius (or so they say) and if you understand things like lowest common denominator algorithms, how to use probability formulas (my specialty) and basic math properties like sample size, 80/20 theories, cross referencable data tremendous advantage can be had.
Just because an attempt at gaining advantage has not been overly successful doesn’t mean it is not possible.
BTW I’m working on a coaching analysis. This has been a fascination I’ve had since the I was obsessed with college ball in the 80’s. Carlisle’s best asset is his half time adjustments…he’s one of my favorites.
Comment by veggamattic -
Maybe the analysts should apply deeper and more sophisticated statistical models and re-do all the computation. For example, players’ stats should always be correlated with their coaches and the teams, which should all be considered as variables. Also, make bosses variables, too. 🙂
Insights are always hidden in the network. Simple averages do not reveal them.
Comment by notinhell -
Comment by 16shoes -
Pingback: View From Above (10.1.10 – 10.7.10) « Northwest Roots
Dear Mr. Cuban,
My name is Keith Langdon I am a 38 year old paraplegic man who has been coaching basketball for the past 20 years. I aspire to be working for an NBA organization such as yours whether it be assisting or otherwise. Can you tell me how I can make this happen? It would be greatly appreciated. Thank you in advance.
Comment by alpha7232 -
Can’t you define coaching/ playing style and evaluate players to those objectives?
Chemistry is a tough one, but I wouldn’t its complexities cause any nails to go into the coffin of statistical analysis.
Comment by jonathanjoyce37 -
@bryanb426 Hi Brian. Amen to linear mathematics. I could never understand why Americans love baseball because you’re right: it is a linear sport with maybe the exception of a triple play. But what do I know…I never played baseball.
You can’t use linear mathematics when you need complexity science (and non-linear math) to help in the analysis of complex, adaptive, self-organized systems in nature, business as well as sports(including basketball teams).
Having been an athlete myself, I can say that the psychological factor is much more important than trying to figure out how to track stats that do not have any significance.
Look at France in World Cup 20101. They have world class players and they got knocked out because the “chemistry” between the players and the coach was terrible — which came down to psychology instead of statistics. I predicted 3 out of 4 teams to make it to the World Cup Final because when I analyzed the 32 team — I looked at the psychological factors involved.
Spain won because of team unity and the psychological factors involved. They did not play the best in the tournament. They did not score that many goals — but they managed to win.
The Germans were actually a better team when it came to team play. The Dutch did not win (I was in Netherlands watching the final) because they did not play “the Dutch style” of football. They wanted to win no matter how which threw them off psychologically as a team. Had the Dutch focused more on the psychological factors of each player, it would have been a different game. (If you recall, Dutch lost twice in the Final already.)
Comment by entrepreneurdex -
Sport geeks love stats.
Comment by darryl3 -
Pingback: Pure Sabermetrics» Is NBA Sabermetrics (APBRmetrics) Dead?
The reason sabrmetrics doesn’t truly work for other sports (outside of baseball) is the nature of the statistical method. Moneyball is predicated on using linear regression to predict competitive (market) efficiencies. The key word is linear, baseball is linear, everything happens in a linear fashion, where the sum is the total of the parts, therefore predictable in a regression equation. I.E. 4 total bases in succession by one player always equates to a run. Sports like soccer, basketball and hockey have too many “intangibles” related to player performance and nothing always equates to a basket or goal.
Comment by ironman33 -
See Mark. That’s why ur in charge up there. It’s cuz u get all scientific with them. You’re on another level. haha. Love it.
Comment by justapodcastmaxgreen -
I know this was quoted from your post in 2004
“That’s what makes all this so difficult to collect. It may well be that we need to add cameras to each gym that can cover all the action and then go back and determine the information. Either way is expensive, which means it could be along time in coming to the NBA.”
Do you still think this is true? I mean seriously, what costs a team more, signing a multi-million dollar bust or installing a couple of hundred thousand dollars worth or cameras/film equipment?
Several NFL teams I know have 22-camera systems that follow each player on each play of each game. Less than half of that and you could follow all ten on the court for each game, each possession (trip down the court) and really perform some good qualitative (NOT quantitative statistical noise like rebounds, etc) analysis not only on your team but on the opponents.
Combine THAT with coaches’ knowledge of the players and chemistry (ie. intangibles) and you’d really set yourself apart from the pack.
You’re dead on with one thing though – the tried and true (“tired and old”) statistical approach certainly has come nowhere near fulfilling its promise.
Once again, I can help you there.
Comment by bryanb426 -
Pingback: Top Posts — WordPress.com
I used to play football (soccer) seriously enough to get scholarship at SMU and play on the U.S National team — so I have been exposed to some good coaching. The problem with stats when it comes to athletes is that they don’t account for the personal drama factor. Soccer, basketball or football — it’s all relative. Drama in athlete’s lives will provide some indication of the upcoming performance. I know this from a personal standpoint as a player. When I had no drama, my mental state was in a much better position to give maximum output. When I was happy without any negative energy from my personal life, I performed significantly better.
Providing extra bonuses if the player has a better free throw percentage is irrelevant. As an athlete, despite what anyone might think, you play for the love of the game – not money. It’s not easy being a professional athlete – despite what they get paid. It’s a job when large money and pressure is involved — and the athlete has to love the game in order to perform at high levels.
I think professional sports needs to pay more attention to the “drama stats” of players. Pro clubs in soccer conduct a substantial amount of psychological testing on players before they acquire them — but this is not enough. They need to keep weekly drama stats of each player off the field as well. No, I’m not trying to act like “Big Brother” or Google here but pro athletes are a special breed. With so much money being earned and their inexperience in handling it, they are going to find trouble sooner or later: drugs, sex, rock’n’roll and more drugs, sex and rock’n’roll.
I’m sure that you already know all this but I’m gonna say it anyway – despite being so ovious. Peak performance of players is affected by their personal lives – not matter what anyone says. We’re all human – even the megabuck players.
Look at Wayne Rooney (Manchester United). He has been accused in the English tabloids of seeing a prostitute named Jennifer Thompson. It’s not the first time Rooney has been caught out paying for sex. The problem is that he is married – which creates personal drama – which creates poor performance. Last year, you saw signs a the ‘Old Trafford” stadium that called Rooney the ‘white’ Pele. He is no Pele (my favorite player) but he is damn good. This year, it will be interesting to see if he can put his personal drama aside and perform. My bets are that he’s going to struggle until the first of the year. Drama has to evaporate sooner or later.
Personal drama stats are as critical as other stats in my opinion. Even one player’s personal drama can affect the rest of the team — especially if they are the heart and soul of the team – like Rooney.
When I ran my first venture fund, I focused more on developing the psychological profile of the founder and the rest of the team – vs. financial projections. It was important to me to figure out the stress level of entrepreneurs before I funded them as well as their potential personal drama. I used some proprietary ingredients as well to predict drama outcomes. 🙂
Before I ran my venture fund, I angel funded an entrepreneur and later found out he was screwing around with the hot assistant. This created a lot of drama for all – once everyone found out. The guy couldn’t keep his zipper closed and the venture eventually failed due to so much drama.
I am sorry to use soccer analogies, but since I’m Serbian, it is all the same no matter what sport. (Serbs love both soccer and basketball and you should look for more players from there.)
Great athletes (the ones who perform consistently) have less personal drama in their lives. I worked with Emmitt Smith (on business) when he broke his NFL rushing record. One thing I observed about Emmitt is what a great person he is off the field. Sure, we all have personal dramas, but I’ve never personally seen Emmitt with personal drama – while I worked with him and he was playing for Phoenix then.
Emmitt is a perfect example of a great athlete minimizing his personal drama to provide maximum performance for his team. Emmitt was very careful in how he conducted his personal life. Everyone (entrepreneur or not) can learn much from Emmitt in reducing the personal drama from their lives to generate maximum performance in whatever they do.
P.S. You may want to talk to Emmitt about it. Also, you may want to study how Nadal creates maximum performance as well. They (his coaches) are keeping it a secret. Maybe if you gave Nadal season tickets to Mavericks, he might provide some insight.
Comment by entrepreneurdex -
I would look into what stats John Wooden and Phil Jackson paid/pay attention to. While at it, I would look into Coach K’s playbook to see what stats he cares about. That would be a good starting point for relevant statistical analysis.
Comment by josephwesley -
I find myself interested in your movement in thought but in agreement with some things more than others, or least the phrasing of some things.
“one nucklehead hurts, two nuckleheads kill”
I agree with that conceptually and based on some lineup research of my own.
“the input of numbers into building a team is diminishing and being displaced by non traditional qualitative factors”
I can see the input of “numbers” as a share of total “input” diminishing with the rise of “non traditional qualitative factors” but I’d be surprised if the actual volume of numbers input broadly defined went down. I wouldn’t let it go down in real terms.
“Until you can quantify coaching and chemistry, you can not use the numbers to build a team. Period end of story. You can use them as partial input along with scouting and other elements, but there ain’t no Moneyball solution… Stats will continue to play a role in lineups, matchups and trends, but teambuilding, not so much.”
I am not seeing teambuilding quite as a separate activity as much as you. The roster and the rotation feed into systems, lineups, matchups, plays,
lineups stats, trends and adjustments.
Comment by justbeyondthearc -
Nevermind that the Thunder, led by GM Sam Presti, are another fine example of quantitative analysis leading to efficient, cost-effective results.
You may say they got lucky in landing Kevin Durant, but then again didn’t you get lucky in landing Dirk Nowitzki? Isn’t there always some element of luck to land a superstar?
Comment by adamcmadison -
I can’t help but feel a potential reason for this post is because Cuban tried to build the Mavs through (in part) his own use of “quantitative” analysis yet it hasn’t been working well the past few seasons.
Cuban, you’re an owner. That means you have a chance to SHOW us how to build a team, not TELL us.
I would be happy to make a little friendly wager that the Rockets, maybe the most notable team in the league with regards to sabermetrically-inclined decisions, will be better than the Mavericks (judged by Pythagorean record) this season.
Comment by adamcmadison -
It’s interesting how much of what you say above translates to filmmaking, as many of your other concepts as well. The “stats” would necessarily be different, however. It all boils downto, are you working with, or against someone?
Comment by Matches Malone -
maybe this sounds weird i dunno ?
was Joaquin Phoenix’s im still here
inspired by me, i worked with him in 2006
was suppose to be a russian bad guy for multiple days
pa’s didn’t like me or joaquin himself maybe i dunno really
his beard is reminiscent of mine tho i am a better skilled rapper mc
Comment by ajew -
this is why the brain typing field is so fascinating to me. it is interesting, though, that a team with a full time brain typing analyst, the celtics, even had chemistry troubles involving Rondo and the other guys. it is arguable however, if those chemistry issues ever even derailed a playoff run, because their run in 08-09 was pretty impressive for no Garnett (taking Orl to 7) and they came ever so close to the mountaintop last season.
i’m sure Dirk (infp) didn’t mind kidd (istp) taking over that vocal leadership role
Comment by jdpapa3 -
seriously see all nba champions years ago
rockets Hakeem “the dream” was shorter and smaller
tim duncan was doofier and more uncool than Shaq & KG
jordan was gifted and also blessed with a unattractive pippen
scottie pippen is very underrated he was a huge piece of mj’s wins
everyone is human guys that get far are usually picked on when young
Comment by ajew -
your wrong marky baby as mans wong
best nba teams have at-least 2 or 3 pieces now
people with something to prove artest, kobe and gasol
like a man who is ugly and wants to be knows as pretty !
or a man with a small wang who wants to show he has a big wong
like a actor named lebroz from brooklyn who’s website is lebroz.com
Comment by ajew -
Pingback: Analysis of TMK – Mortal Kombat Conquest Episode # 2 – Part II Eternal Warrior
It isn’t easy for anyone-as I just demonstrated. (wink)
Comment by nickmortensen -
You cannot hope to rid your organization of knucklheads until you learn to spell knuckleheads correctly.
Comment by nickmortensen -
Mark, great timing on the blog post. I agree with many of your points and would like to present an alternative way in determining a player’s value.
A player’s value is measured by:
– Total number of searches for the player. The stats should include the data from nba.com, mavs.com, google.com, dallasbasketball.com, streetball.com, facebook.com, Mavsfastbreak.com, MavsMoneyBall.com, Friends.mavs.com and other properties.
– Number of video views broken down by player. (same data source)
– Number of comments in response to videos (same data source)
– Number of Facebook Fans, Twitter followers etc.
– Downloads – if downloads are available on sites – segment the players based on the number of downloads.
– Uploads – We all know fans love to upload videos, pics and articles about their favorite player. Segment and rank the player based on the type of media uploaded. For example, player wallpapers really drive traffic and user interaction. I recommend putting a “Wallpaper” category on the homepage of mavs.com. that section alone will out perform many of the inner pages. streetball don’t lie. lol
Considering the trends in online advertising, the stats become more valuable over time. So valuable in fact, the data gets overlooked. Could you imagine Billy Hunter at Ad:Tech? lol
Comment by ryanmendezstreetball -
I couldn’t agree with you more. Some players can excel on one team/system or in one role, and then completely underperform after a trade (and vice versa). The differences are usually chemistry and coaching.
My colleagues and I have actually explored these nontraditional qualitative factors for every player in the league, and have built the information into an online web application called “The Team Chemist,” to help NBA front offices build better teams.
It was recently covered in The Huffington Post. You might be interested in taking a look for your team.
Company website/contact: http://www.theteamchemist.com
Thanks for shedding your experience on this oft-debated topic.
Comment by nabeelahmadtc -
Hi Mark and the entire Blog Maverick Community (This will be my last post on this, I promis, but just hoping to grab attention for a good cause),
With your history of being a proud rugby player with a loud voice that can be heard, I’m reaching out to you to ask you to join the fight to help Save Cal Rugby (Not looking for money, just your voice). I personally am a UC Berkeley grad and rugby alum, and I speak for the entire Cal rugby community when I say we would love to have you involved in some capacity.
I’m not sure if you are aware of everything that has happened with rugby being relegated to a “Varsity Club” (whatever that means) at the school, which would effectively kill the proud program that Cal Rugby has worked so hard to build since 1882. With 25 national championships since that title became official in 1980, the administration has made a mistake in doing this and did not calculate the repercussions this would bring. There are a lot of angry alums out there who are major donors who are planning to take action. We are all mobilizing as well to stage a large show at the upcoming Cal vs UCLA football game. We simply REFUSE to rollover on this and will not allow this to happen. The Rugby program is self-funded, brings in $300,000 in sponsorship money to the general athletic fund, and has offered to start up a Varsity Women’s program on our own dime.
With Rugby growing steadily in the US, and now an Olympic sport, this will affect the entire USA Rugby community and spits in the face of the progress that has been made with rugby in our land.
If you have a few minutes, please take a look at this video clip of an interview with our fearless coach, Jack Clark. I believe the two of you would get along smashingly and share many of the same values: http://www.csnbayarea.com/09/30/10/1-on-1-with-Cal-Rugby-Head-Coach-Jack-Cl/landing.html?blockID=322201&feedID=7093
Mark, again, we would be grateful for any verbal/written support you could provide in our fight to right a baseless wrong. We ask that you help spread the word about the Save Cal Rugby facebook Fan site (http://www.facebook.com/pages/Save-Cal-Varsity-Rugby/151937381513223?ref=ts), as we are hoping to get tens of thousands of supporters on the facebook site. Please also spread the word to follow us on twitter @SaveCalRugby.
Please feel free to reach out to me with any questions at all at email@example.com. Thank you for hearing us out.
Save Cal Rugby
Comment by savecalrugby -
Comments are closed.