[Memphis MadnessParticipant]

 

The NBA regular season is almost here and fantasy basketball season is heating up.  I just finished reading Moneyball and I thought I would use some of that inspiration to take a closer, metrics-based look at BASKETBALL.

So, I am writing this now because of the Dante Exum thread.  Someone had a minimal upside comparison for Exum as Fat Lever.  Well, I took a look at Fat Lever, and he had some really, really good years.  He was a guy I never watched play but if you looked at his stats, he could score, and rebound and pass with anyone.

Stats wise, Fat Lever had an all-time great year one year for a point guard.  Can’t remember which one it was but it was superior to any year by Jason Kidd or even John Stockton.

Ok, it was the 1986-1987 season, and taking the totals of points, rebounds, assists, steals, and blocks minus turnovers, Fat Lever’s ’87 campaign was rivaled by very few point guards.  Allen Iverson had a better year in ’05 but Mr. Fat exceeded Tim Hardaway’s best season.

As far as point guards go, Magic has the record, ironically, also in ’87.

Since steals, blocks, AND turnovers were kept as official records, Michael Jordan has the highest rating for any season, of course.  Actually, the three highest I have found.  His ’89 season ranks first, his ’88 season ranks second, and his ’87 campaign ranks third, although there were only 7 composite points difference between his ’87 season and his ’89 high water mark.

The next best player is ’99 Shaq but a decent bit away from MJ.  Third is ’94 David Robinson, but only 3 composite points behind Shaq’s best season.  The fourth player on the list is ’82 Moses Malone.  Slightly better than Hakeem Olajuwon’s best effort.

As far as small forwards go, ’85 Larry Bird tops ’10 LeBron James by a fairly decent amount.

’04 KG finishes behind Larry Bird but just ahead of vintage LeBron in the total productivity score.  So, that year’s version of KG is the best power forward of the modern stats era.

Starting five: ’89 Magic, ’89 Jordan, ’85 Bird, ’04 KG, and ’99 Shaq.  

I ranked these guys with a metric I am calling the Total Productivity Index:

Points + Rebounds + Assists + Steals + Blocks – Turnovers

And, for my money, I would use this formula to determine the optimal Big Three on any team.  For role players, I want efficiency, 3 point shooting, and defensive stats.  So, I looked at 3 pointers made plus steals plus blocks minus turnovers.  According to that, Manute Bol had the two highest years by far with that measure due to his insane blocks totals and minimal turnovers.  

Rasheed Wallace had some great years in terms of 3’s plus blocks plus steals minus turnovers including some years where he had 100 or more in each of the asset columns.

Shane Battier and Robert Horry show well, too.  

So, you have a perfect stretch four, third option kind of guy in Rasheed Wallace along with two of the ideal role player types in Battier and Horry.  I also think that Manute Bol was criminally misused although I never really watched him play.  To me, a guy who can block a ton of shots, hit 3’s, AND has relatively few turnovers should have a ton of value, even if he has an unconventional game.

It appears that a front court rotation of Rasheed, Battier, Robert Horry, and Manute Bol would be the IDEAL Money Ball type rotation where you cover all the bases at the 4 and 5 spots as well as small forward.  

For non-Big Three guys who also aren’t pure role players, I would go with another index for what I would call either a 4th option or a 5th option/6th man.  That would be (total free throws x FT%)/total field goal attempts.  Basically, I want a guy who can efficiently get to the line without taking a ton of shots.  

Outside the typical 8 man rotation, I would get a 3 point specialist.  Find a guy who hits a decent number of 3’s at a high rate.  Take 3 point field goals made x 3PT%.

Then from a Money Ball perspective, you can’t waste draft picks.  I also would really use the 2nd round to find a specialist or two or just to fill out the roster (third string point guard and third string center).  Probably not enough 2nd round picks on rosters.

Use the lottery picks for TALENT and Best Player Available.  In the second round, go after guys who fill needs.  In Money Ball, Billy Beane always wanted to go with college guys over high school kids.  In this case, you can probably get a more NBA-ready 2nd round pick if you use your selection on a 4 year guy.  Maybe not the upside, but you are less likely to whiff.  Gotta use cheap picks effectively.

From a draft point of view, I think you either want to be in the lottery to go after the BPA guys, OR you trade down into the second round.  Not only do you get a shot at experienced specialists you can also get a few UPSIDE flyer guys.  Or, use an extra second rounder to stash a guy overseas.  

For an MLE or a late season pick up go with a veteran with a solid +/-.  Basically, you want to go after a guy who makes good things happen when he is on the court.  Don’t overpay these guys, but you need at least one of these guys on the roster.

D League guys might be an even safer bet than four year college guys to fill out the back end of your bench.  So if you want a 14th man I would look at D League stats depending on what you need.  Also take a look at the college stats and even the preason/summer league stats if these guys were on an NBA roster at any point.  Ideally, go after a guy with the high (3PTx%) + STL + BLK – TO ranking.  Or, a 3 point shooter where you take the raw 3 pointers made stat multiplied by the 3pt percentage.  You look at D League stats as well as college stats to see if there is an improvement, or at least not a fall off.  Not sure you can get good plus/minus stats from the D League.  You probably get a distorted look at things anyway.

You need to fill out a roster, BUT you have to do so effectively, and you cannot overspend.  Pay your top guys but know what you are looking for.  Then use your money wisely to fill out your rotation.  Don’t overpay for a one dimensional 3 point shooter.  Then you obviously don’t want to overpay for your generic locker room veteran guy, or a third string center, or a third string point guard.  You need these things, but you don’t need to spend lots of money on them.  Then, you cannot waste draft picks.  Picking well in the second round should give you good value over the long run.  I would even think about either trading up into the lottery or trading down into the second round.  Not sure picks in the 25-30 range have a ton of value.  

I haven’t looked at what the ideal 6th man should do but I think it would be a guy who can bring a different dynamic to the team, the ability to change the pace, and mix things up.  The main question is actually, can this guy work with the 1st unit AND the second unit?  You can’t just look at a 6th man candidate in a vaccuum.  

I would lean towards using an 11 man rotation.

Starting five.  Then your 6th man who can ideally replace at least one or two guys on the starting unit.  Then a full second unit with a good mix of role players who can make plays (especially on defense) without turning the ball over.  At least one of these guys (ideally) can effiecently get to the line and at least one above average 3 point shooter.  

If your 6th man is hurt I wouldn’t change the second team, I would use a guy from the deep end of the bench (12-14) to play the 6th man role since the back end of the bench is where you stash your specialists. You also don’t want to mess up your second unit.  

 

[King CaluchaParticipant]

I’d suggest you go into APBR’s forum. Things have moved really fast since Moneyball.

[King CaluchaParticipant]

I’d suggest you go into APBR’s forum. Things have moved really fast since Moneyball.

[NBAjunkie81Participant]

 Stud young pitchers – Mulder, Zito & Hudson – all of whom pitched well over 200+ innings – this fact was conveinently omitted from the Brad Pitt movie – which implied that Billy Beane took a bunch of grocery store clerks, beer truck drivers & brick layers all of whom had one very special talent & he masterfully stitched then together like a modern day Dr. Frankenstein to produce Two 100+ win season in the early part of the new 21st Century… The truth is much less exciting… Organized Professional Baseball has been played in the United States for close to 150 years & great Pitching combined w/ great Defense wins all day, every day in this sport…

Professional Basketball also boasts an Inescapable truth. Since 1947, when the NBA was born the team that wins the Championship has a Superstar player who competes for League MVP & is a future Hall of Famer & his presence elevates the game of Everyone else on the team – (This includes the ABA)… In 68 years of the NBA (&ABA) a team w/ hopes of a Championship Must have this player…

 You created a new metric but it yielded the Same result – meaning we ALL know who the top 3 to 5 players are at Any given time & Only those teams that employ these men have ANY chance to win a Ring

  – Where does a team acquire these "Super" stars?  – invairably in the Draft Lottery – In the History of the NBA Every future MVP was drafted in the Top 10 of the Lottery – Only 3 men have escaped the top 10 but even that small # was Still taken by #15 at the latest & two of them Still did not win a Ring & the one who did was not really a person who should have Ever fallen below #5 – they are: Karl Malone, Steve Nash & Kobe Bryant… Modern Analytics yields these same conclusions… By hook or by crook, for a team to win a Championship, they must acquire one of these Top Super Stars & then everything else looks will fall into place w/ a good GM….

 

[NBAjunkie81Participant]

 Stud young pitchers – Mulder, Zito & Hudson – all of whom pitched well over 200+ innings – this fact was conveinently omitted from the Brad Pitt movie – which implied that Billy Beane took a bunch of grocery store clerks, beer truck drivers & brick layers all of whom had one very special talent & he masterfully stitched then together like a modern day Dr. Frankenstein to produce Two 100+ win season in the early part of the new 21st Century… The truth is much less exciting… Organized Professional Baseball has been played in the United States for close to 150 years & great Pitching combined w/ great Defense wins all day, every day in this sport…

Professional Basketball also boasts an Inescapable truth. Since 1947, when the NBA was born the team that wins the Championship has a Superstar player who competes for League MVP & is a future Hall of Famer & his presence elevates the game of Everyone else on the team – (This includes the ABA)… In 68 years of the NBA (&ABA) a team w/ hopes of a Championship Must have this player…

 You created a new metric but it yielded the Same result – meaning we ALL know who the top 3 to 5 players are at Any given time & Only those teams that employ these men have ANY chance to win a Ring

  – Where does a team acquire these "Super" stars?  – invairably in the Draft Lottery – In the History of the NBA Every future MVP was drafted in the Top 10 of the Lottery – Only 3 men have escaped the top 10 but even that small # was Still taken by #15 at the latest & two of them Still did not win a Ring & the one who did was not really a person who should have Ever fallen below #5 – they are: Karl Malone, Steve Nash & Kobe Bryant… Modern Analytics yields these same conclusions… By hook or by crook, for a team to win a Championship, they must acquire one of these Top Super Stars & then everything else looks will fall into place w/ a good GM….

 

[Magic JordanParticipant]

 Moneyball is already for the dinosaurs.  Well, not exactly… but the future of player and team prediction revolves around machine learning.  Machine learning is something I focus on a bunch in my job and it will blow your mind.  A properly aggregated set of data (or de-aggregation when needed) is needed, but with that you can control for almost anything you want.  A deep learning neural net is what I have seen used most, to date, but that is a black box.  Nobody knows why neural nets work, they just know that they work, and that is fine for most things.  However Genetic Programming, (Artificial intelligence that basically emulates evoultion) is the next step.  

In short, you give it a ton of x’s… and it derives a Y.  They are a mofo to code.  We are getting close though, in which case we will turn it loose on forecasting individual games, prospects, but most importantly to me, sports betting.

I have obviously omitted a ton of details, but suffice it to say that machine learning is the future for… well just about everything.  Sports included.  In fact, in the fields of Economics and Computer Science… shortly it will no longer be good enough to be one or the other.  Due to machine learning leaning heavily on economic principles, with the technical aspects of computer science… people will have to start having a firm knowledge in both.  Once quantum computers come around (IF), this stuff will really get crazy due to their ability to crunch an unthinkable amount of numbers is no time at all.  Even in something as seemingly "random" as sports, this type of computing will astonish you at what it can predict.

 

 

[Magic JordanParticipant]

 Moneyball is already for the dinosaurs.  Well, not exactly… but the future of player and team prediction revolves around machine learning.  Machine learning is something I focus on a bunch in my job and it will blow your mind.  A properly aggregated set of data (or de-aggregation when needed) is needed, but with that you can control for almost anything you want.  A deep learning neural net is what I have seen used most, to date, but that is a black box.  Nobody knows why neural nets work, they just know that they work, and that is fine for most things.  However Genetic Programming, (Artificial intelligence that basically emulates evoultion) is the next step.  

In short, you give it a ton of x’s… and it derives a Y.  They are a mofo to code.  We are getting close though, in which case we will turn it loose on forecasting individual games, prospects, but most importantly to me, sports betting.

I have obviously omitted a ton of details, but suffice it to say that machine learning is the future for… well just about everything.  Sports included.  In fact, in the fields of Economics and Computer Science… shortly it will no longer be good enough to be one or the other.  Due to machine learning leaning heavily on economic principles, with the technical aspects of computer science… people will have to start having a firm knowledge in both.  Once quantum computers come around (IF), this stuff will really get crazy due to their ability to crunch an unthinkable amount of numbers is no time at all.  Even in something as seemingly "random" as sports, this type of computing will astonish you at what it can predict.

 

 

[Memphis MadnessParticipant]

 In the Moneyball book after another post season 5 game flameout, Billy Beane admitted that his stuff only worked in the regular season.  Beane said that in the playoffs you need 3 pitchers and luck.  

Ok. In basketball, the only major difference between a Dirk and a Carmelo is the ring.  That’s really about it.a superstar isn’t even enough. Ask Mailman, Stockton, Barkley, Elgin Baylor, and some others — you still need an efficient, well constructed team.  

So is Marc Gasol a superstar? Yeah, if the Grizzlies win a title.  If we crater this year then he won’t be regarded as one.  So you at least need a borderline superstar, then a title turns that guy into a superstar.

[Memphis MadnessParticipant]

 In the Moneyball book after another post season 5 game flameout, Billy Beane admitted that his stuff only worked in the regular season.  Beane said that in the playoffs you need 3 pitchers and luck.  

Ok. In basketball, the only major difference between a Dirk and a Carmelo is the ring.  That’s really about it.a superstar isn’t even enough. Ask Mailman, Stockton, Barkley, Elgin Baylor, and some others — you still need an efficient, well constructed team.  

So is Marc Gasol a superstar? Yeah, if the Grizzlies win a title.  If we crater this year then he won’t be regarded as one.  So you at least need a borderline superstar, then a title turns that guy into a superstar.

[Memphis MadnessParticipant]

 You can also say that if you don’t have a superstar then go after a superstar just past his peak.

The ’04 Pistons lacked a surefire superstar of course, but you can argue that the 2008 Celtics, 2011 Mavs, and the 2014 Spurs lacked a superstar at their peak, either they were towards the end of their prime or just past it. 

The Shaq-Wade Heat team featured a superstar on the way up and another on the way down.

[Memphis MadnessParticipant]

 You can also say that if you don’t have a superstar then go after a superstar just past his peak.

The ’04 Pistons lacked a surefire superstar of course, but you can argue that the 2008 Celtics, 2011 Mavs, and the 2014 Spurs lacked a superstar at their peak, either they were towards the end of their prime or just past it. 

The Shaq-Wade Heat team featured a superstar on the way up and another on the way down.

[Memphis MadnessParticipant]

 Basically, don’t trade Dominque Wilkins for Danny Manning. 

Or, on a smaller scale don’t trade Rudy Gay for Tayshaun Prince’s corpse.  

The ’95 Rockets augmented this rule by trading former All Star but a still solid Otis Thorpe for declining superstar Clyde Drexler.

It seems as if declining superstars have value, possibly as much as two top shelf starters — possibly more.

However, if all you needed were a superstar, even the best player in the league perhaps, then the Cleveland Cavaliers would have a title by now.  

[Memphis MadnessParticipant]

 Basically, don’t trade Dominque Wilkins for Danny Manning. 

Or, on a smaller scale don’t trade Rudy Gay for Tayshaun Prince’s corpse.  

The ’95 Rockets augmented this rule by trading former All Star but a still solid Otis Thorpe for declining superstar Clyde Drexler.

It seems as if declining superstars have value, possibly as much as two top shelf starters — possibly more.

However, if all you needed were a superstar, even the best player in the league perhaps, then the Cleveland Cavaliers would have a title by now.  

[T Rex]

 I could see the Cavs really struggling in the regular season. By the time Kyrie comes back, it will almost be time to start resting LeBron and Love and Varajao and etc.

And the Tristan Thompson thing looks like a huge distraction, like a ticking time bomb.

There’s too many decent teams, even in the East, for Cleveland to waltz through the first half of the season without Irving. If/when the Cavs struggle, the TThompson stale-mate just becomes more of a distraction, and LeBron isn’t above pouting and holding the franchise hostage. Then you have a mediocre coach in over his head, and a team that doesn’t really like him a whole lot.

Normally I wouldn’t hold the first half of a new team’s first season against them, but the team almost fell apart last year, and there’s plenty of challenges and distractions that are going to test them again this season.

Everybody expects them to dominate the East, and there’s going to be a circus of media and other distractions/challenges if they start out slow.

And they very well could start out slow. LeBron started slow last year, and he’s a year older. And Team LeBron is in an off-court war trying to get Tristan the money he doesn’t deserve.

 

 

 

 

 

[T Rex]

 I could see the Cavs really struggling in the regular season. By the time Kyrie comes back, it will almost be time to start resting LeBron and Love and Varajao and etc.

And the Tristan Thompson thing looks like a huge distraction, like a ticking time bomb.

There’s too many decent teams, even in the East, for Cleveland to waltz through the first half of the season without Irving. If/when the Cavs struggle, the TThompson stale-mate just becomes more of a distraction, and LeBron isn’t above pouting and holding the franchise hostage. Then you have a mediocre coach in over his head, and a team that doesn’t really like him a whole lot.

Normally I wouldn’t hold the first half of a new team’s first season against them, but the team almost fell apart last year, and there’s plenty of challenges and distractions that are going to test them again this season.

Everybody expects them to dominate the East, and there’s going to be a circus of media and other distractions/challenges if they start out slow.

And they very well could start out slow. LeBron started slow last year, and he’s a year older. And Team LeBron is in an off-court war trying to get Tristan the money he doesn’t deserve.

 

 

 

 

 

[TRC1991Participant]

The 6ers implemented the money ball approach years ago and guess what it’s gotten them?

Dead last 3 years in a row with another horrible year looming 

 

 

 

[TRC1991Participant]

The 6ers implemented the money ball approach years ago and guess what it’s gotten them?

Dead last 3 years in a row with another horrible year looming 

 

 

 

[I am an idiotParticipant]

In a way, everybody is right. You need a superstar but a superstar can’t win single handed. Look at the Cavs last year and Bulls before Pippen and the rest supported MJ. The closest to a superstar winning a ring single handed is Walton in ’77 but even he had Lucas.

As far as I can tell, only two teams have ever gotten a ring without an MVP or MVP candidate, the Supersonics in ’79 (though Sikma and DJ were pretty good) and the Pistons in ’04.

Since most future MVP candidates are recognized early and are drafted low (usually top-5, almost always top-10) the ’76er’s strategy makes a certain sense. Stink up the place, draft low and if the draftees are not superstars, get rid of them and try again. It is tough on the fans but is the most likely way to get struck by lightning.

 

 

 

 

[I am an idiotParticipant]

In a way, everybody is right. You need a superstar but a superstar can’t win single handed. Look at the Cavs last year and Bulls before Pippen and the rest supported MJ. The closest to a superstar winning a ring single handed is Walton in ’77 but even he had Lucas.

As far as I can tell, only two teams have ever gotten a ring without an MVP or MVP candidate, the Supersonics in ’79 (though Sikma and DJ were pretty good) and the Pistons in ’04.

Since most future MVP candidates are recognized early and are drafted low (usually top-5, almost always top-10) the ’76er’s strategy makes a certain sense. Stink up the place, draft low and if the draftees are not superstars, get rid of them and try again. It is tough on the fans but is the most likely way to get struck by lightning.

 

 

 

 

[King CaluchaParticipant]

"Nobody knows why neural nets work"… False

Although behavior can be modelled in different ways through past information, it’s not a guarantee that it will be a good predictor, but that’s another discussion. The efforts in the sports science field have been greatly improved by the new player tracking system.

Also, there are some minor details that should be analyzed on case-by-case basis. For example: Let’s assume we want to develop a rating for defensive plays where you reward evaluate each situation and reward the player’s execution. Example: Roy Hibbert intervenes in a pick and roll situation. His coach asks him to back up to contain the roll, but the opposing player hits a 3 immediately after the pick. Is this his fault? The plan could have been executed with perfection but the result was still unsuccessful for his team. This is a very simple player… how about help defense and rotation? So at the moment we only have approximations were we estimate a player’s impact as objectively as we can but one could argue these results are far from perfect.

 

 

 

[King CaluchaParticipant]

"Nobody knows why neural nets work"… False

Although behavior can be modelled in different ways through past information, it’s not a guarantee that it will be a good predictor, but that’s another discussion. The efforts in the sports science field have been greatly improved by the new player tracking system.

Also, there are some minor details that should be analyzed on case-by-case basis. For example: Let’s assume we want to develop a rating for defensive plays where you reward evaluate each situation and reward the player’s execution. Example: Roy Hibbert intervenes in a pick and roll situation. His coach asks him to back up to contain the roll, but the opposing player hits a 3 immediately after the pick. Is this his fault? The plan could have been executed with perfection but the result was still unsuccessful for his team. This is a very simple player… how about help defense and rotation? So at the moment we only have approximations were we estimate a player’s impact as objectively as we can but one could argue these results are far from perfect.

 

 

 

[Magic JordanParticipant]

 Okay, the phrase "nobody knows why they work" isn’t entirely accurate,  but for the sake of a basketball crowd it works.   We know how they work, but lookong into them to see what exaclty has been leanred is very much a black box, this has not been proved otherwise.  A neural net isn’t a black box in the sense that you can’t examine its code or make modifications.  However it is a black box in the sense that a mathematical formula cannot be derived from a function that is continually evolving.  

You can control, modify and predict its behavior.  However I challenge you to explain why a neural network could be given two different sets of data, yet it is still possible to achieve the same output.

Yes we know they use back propagation,  however deriving a mathematical model that is consistent for a neaural network is simply not possible in a general sense.  This is why we call it a black box.  

[Magic JordanParticipant]

 Okay, the phrase "nobody knows why they work" isn’t entirely accurate,  but for the sake of a basketball crowd it works.   We know how they work, but lookong into them to see what exaclty has been leanred is very much a black box, this has not been proved otherwise.  A neural net isn’t a black box in the sense that you can’t examine its code or make modifications.  However it is a black box in the sense that a mathematical formula cannot be derived from a function that is continually evolving.  

You can control, modify and predict its behavior.  However I challenge you to explain why a neural network could be given two different sets of data, yet it is still possible to achieve the same output.

Yes we know they use back propagation,  however deriving a mathematical model that is consistent for a neaural network is simply not possible in a general sense.  This is why we call it a black box.  

[King CaluchaParticipant]

It’s so darn interesting, isn’t it?

A few months ago I read an article about a new system that rated decision making in the NBA. Basically the player is awarded points if he makes the right plays within the strengths of the team. It’s rather simple in concept, but I assume it requires a lot of processing.

Example: If player A is in the elbow and his shooting % is low from that area, then passing to player B who is a 3pt specialist who is open in the corner would improve the chance of scoring (or at least the expected return for the play). Therefore player A would be awarded points for passing the ball to player B.

Now, player B can shoot from 3 or pass to player C, who is a big man open under the rim. If he decides to shoot, his rating takes a hit, because he didn’t take the best available option, regardless of the outcome.

It doesn’t necessarily rate the best players, but it could be good for finding good role players, system players… if you know what I mean. You always want those players in your roster. You still need a star in order to get a ring.

 

[King CaluchaParticipant]

It’s so darn interesting, isn’t it?

A few months ago I read an article about a new system that rated decision making in the NBA. Basically the player is awarded points if he makes the right plays within the strengths of the team. It’s rather simple in concept, but I assume it requires a lot of processing.

Example: If player A is in the elbow and his shooting % is low from that area, then passing to player B who is a 3pt specialist who is open in the corner would improve the chance of scoring (or at least the expected return for the play). Therefore player A would be awarded points for passing the ball to player B.

Now, player B can shoot from 3 or pass to player C, who is a big man open under the rim. If he decides to shoot, his rating takes a hit, because he didn’t take the best available option, regardless of the outcome.

It doesn’t necessarily rate the best players, but it could be good for finding good role players, system players… if you know what I mean. You always want those players in your roster. You still need a star in order to get a ring.

 

[Magic JordanParticipant]

Yes it is interesting, I absolutely love it.  

It would certainly take a lot of processing, and a ton of different data sets.  I’ve been trying to scrape all of the player movement data I can get my hands on, it’s just so damn tedious though.  

I am going to try to find that article because out of all the things advanced anayltics are currently used for, I think the use of finding role players or system players is currently it’s most beneficial purpose.  If you get a chance and remember where that article is link me.

 

 

[Magic JordanParticipant]

Yes it is interesting, I absolutely love it.  

It would certainly take a lot of processing, and a ton of different data sets.  I’ve been trying to scrape all of the player movement data I can get my hands on, it’s just so damn tedious though.  

I am going to try to find that article because out of all the things advanced anayltics are currently used for, I think the use of finding role players or system players is currently it’s most beneficial purpose.  If you get a chance and remember where that article is link me.

 

 

[King CaluchaParticipant]

I’m pretty sure it was presented on MIT Sloan Sports analytic conference.

[King CaluchaParticipant]

I’m pretty sure it was presented on MIT Sloan Sports analytic conference.

[Magic JordanParticipant]

 I wish they would make the player tracking data available aside from me having to scrape NBA.com or pay a ridiculous amount for a SportsVU subscription.  There are no good basketball apis.  At least that are documented.  It’s a pain the arse having the compile all this stuff beute force.  

[Magic JordanParticipant]

 I wish they would make the player tracking data available aside from me having to scrape NBA.com or pay a ridiculous amount for a SportsVU subscription.  There are no good basketball apis.  At least that are documented.  It’s a pain the arse having the compile all this stuff beute force.  

[Author]

Posts