Kobe on Advanced Analytics: "The Game is about Momentum and Pace"

[IgorK]

Analytics are great for analysis, which comes after the game has concluded.

I'm pretty sure you want all kinds of analytics in your scouting report. That plays a KEY role in how you defend, how you attack, your offense should be designed around those metrics as well. In game sure, someone gets hot and you have to adapt. Someone goes cold, same thing.

I think people are getting it twisted and thinking I mean analytics has no place in the game. Surely they help to point out some helpful trends. But they won't win you a game in the future alone. You need the human eye, knowledge about health, as well as hopefully insights into which motivational mix tape the opposing locker room is playing

[GeorgeMarcus]

Pretty sure in the end we decided who wins the game based on numbers. Scoreboard baby.

Momentum is very important...because it has the potential to affect the scoreboard.

Re: numbers without proper context. Absolutely agree that that's what you want to avoid. Disagree if you're implying those who making use of numbers in the +/- family aren't also considering context.

He didn’t like that I posted the on/offs of the Nets last year (with a minutes qualifier) even after I stated outright that bench players having positive on/offs does not mean they are better than starters with negative on/offs. What I did deduce however is that, for a team with an average point differential of -0.1, negative on/offs imply the starters were worse than an average starting unit. Also that depth was a big part of their success. There was plenty of context in the thread.

Saying that Rondae Hollis-Jefferson and Shabazz Napier had more of a floor impact over Joe Harris on the Nets last season is a textbook example of using on/off with very little context.

Then it's a good thing I never said that... Nothing that could even be misinterpreted as that

[dhsilv2]

Analytics are great for analysis, which comes after the game has concluded.

I'm pretty sure you want all kinds of analytics in your scouting report. That plays a KEY role in how you defend, how you attack, your offense should be designed around those metrics as well. In game sure, someone gets hot and you have to adapt. Someone goes cold, same thing.

I think people are getting it twisted and thinking I mean analytics has no place in the game. Surely they help to point out some helpful trends. But they won't win you a game in the future alone. You need the human eye, knowledge about health, as well as hopefully insights into which motivational mix tape the opposing locker room is playing

I mean there's no counter for "Eye of the Tiger". That's an auto win!

Analytics are a lot more than finding trends. Knowing your guy (who lets be honest you play 3 times a year and forget about) goes left 80% of the time or is a weak catch and shoot guy are things that analytics bring out and play key roles in the decisions players make in real time in game. When NBA players make comments like "I had no idea so and so allstar was that tall" and that was said by Jefferson about Manu when Jefferson joined the spurs...well it's pretty clear players know less about each other than fans generally assume. That's why having analytics available before games is so critical.

[Ryoga Hibiki]

Analytics point to past performance trends. They're not great at predicting true winning, and they never will be since they don't take into account intangible factors like motivation and health.

The goal of analytics is to make intangibles tangible. How can you say they never will?

Because it's like Science trying to prove God. The goal of analytics isn't to prove an intangible thing. It's goal is to provide data, isolate trends, and possible future performance. But it doesn't predict health, emotional states, or even fluke outliers that can affect the whole (think regression to the mean). Grayson f***ing Allen dropped 40 in a game last year. Even if it's a scrub game with neither side having anything to play for, no one predicted that.

Simply put, having the data available helps and is wickedly valuable. But if it's the end-all-be-all, why haven't the Rockets won a chip lately?

Are you serious?
Anyway, it would be super easy to prove God's existence

Sent from my Nokia 3210 using RealGM mobile app

[G35]

He has a point that analytics and advanced stats are not the end all be all. But check this list of MVP players. The common theme I listed is they are all around .550 TS% or better except Iverson. TS% listed is for the year they won it. I don't think it is a accident that the year Kobe won his, was also his 2nd most efficient season.

The ones who won multiple MVP's were all above the .550 thresh hold. Another thing that stood out is how many of these MVPs have never won a Championship. Quite a few of course are still playing.

TS% .644 2018-19 Giannis Antetokounmpo
TS% .619 2017-18 James Harden
TS% .554 2016-17 Russell Westbrook
TS% .669 2015-16 Stephen Curry
TS% .638 2014-15 Stephen Curry
TS% .635 2013-14 Kevin Durant
TS% .640 2012-13 LeBron James
TS% .605 2011-12 LeBron James
TS% .550 2010-11 Derrick Rose
TS% .604 2009-10 LeBron James
TS% .591 2008-09 LeBron James
TS% .576 2007-08 Kobe Bryant
TS% .589 2006-07 Dirk Nowitzki
TS% .606 2005-06 Steve Nash
TS% .632 2004-05 Steve Nash
TS% .547 2003-04 Kevin Garnett
TS% .564 2002-03 Tim Duncan
TS% .576 2001-02 Tim Duncan
TS% .518 2000-01 Allen Iverson
TS% .578 1999-00 Shaquille O'Neal

The threshold is actually 60% really or close to. The early 2000’s was lower but outside of that and Westbrook, almost everyone is approaching 60TS%.

TS% is great to see how efficient a player is and how well they can get to the line. But efficiency is sometimes overblown. James Harden might be a top 3 efficiency player ever, but I don't think he will ever win a title.

I think offensive creation is more powerful than efficiency...I realize that "offensive creation" is a vague term but that's what is, you can't quite describe it in a word or phrase. The players that can create offense in all situations i.e. playoffs, strong defenses, zones, double/triple teams...those players may not always be the most efficient but they propel their teams offense to a higher degree. Few players can do that.

Garnett surprised me; he was a decent FT shooter also. Duncan surprised me being under .600. I was not expecting that, but FT shooting hurt him.

https://www.basketball-reference.com/leaders/ts_pct_career.html


Not surprising, KG has a wealth of skills but he was not very efficient for an elite big man...I think he might be the least efficient elite big of all time. In career TS% during the RS, Garnett is at the very bottom of this list at 250. Below luminaries like Jeremy Lin, Dennis Rodman, Theo Ratliff, and Channing Frye.

I think pre-analytics NBA thought is really interesting and is undervalued by blog boys all the time. Being dismissive and smart-alecy about the mental aspects of the game when you've never played NBA ball isn't more empirical; it's actually less responsible because you're actively dropping qualitative information that can either be quantified, be used to interpret the numbers, or both.

Kobe's approach to the game was deception. He would start off by imposing his will and getting the defense to react a certain way, and then do the opposite to throw them off. He would do this in something as small scale as a single move, or across possessions, or across games within a playoff series. He would be very informed by film and scouting reports, focusing on tendencies and less on local optima.

It seems to me the dominant school of thought now is to do the most locally optimal thing per possession. It's definitely less of a mental/emotional calculation and more of just referring to data and basically executing a simple min-max. But I think winning is largely a separate experiential skillset, based on very minute details. That's something Kobe loves talking about.

+100

I played a lot of videogames growing up, and I was at that time when you had to go to the mall and play them at an arcade. When Street Fight II came out and then Mortal Kombat came out it was a great time for fighting games. The thing is when you play against the AI at home is far different than playing against a human opponent. The programmers can optimize the character until its nearly unbeatable, but I always felt there was some way to adapt and beat the AI, even on the hardest difficulties.

But going to the arcade...having to put your quarter up on the screen, basically saying you got "next"...(quarters were round pieces of metal we used for currency for cheaper items) and everyone would look at you and you give you that look. The same look you get when you got next on a basketball court. You have to prove yourself. The difference with playing in the arcade is your emotions change, you get excited, you may start to feel hot...sweat...and then you had people who were trash talking behind you, taking sides, saying which matchup was better and who could beat who.

In fighting games, a huge key to winning is adaptation to your opponent and momentum. Knowing what your opponent wants to do and either taking it away or using it against them. Also, there are times when you are down big, its looking like a sure L, and then you have that "real master comeback" and one thing can change the momentum of the game. You never give up, there was always a chance you could win, but you could always tell the guys would give up when things weren't going their way and then the guys who would/could adapt.

You see the same thing in the NBA...when things are going their way some players are invincible (cough James Harden cough) and then when things aren't going their way they look like garbage and you can see they do not adapt to the situation or to their opponent.....

[ShotCreator]

I think pre-analytics NBA thought is really interesting and is undervalued by blog boys all the time. Being dismissive and smart-alecy about the mental aspects of the game when you've never played NBA ball isn't more empirical; it's actually less responsible because you're actively dropping qualitative information that can either be quantified, be used to interpret the numbers, or both.

Kobe's approach to the game was deception. He would start off by imposing his will and getting the defense to react a certain way, and then do the opposite to throw them off. He would do this in something as small scale as a single move, or across possessions, or across games within a playoff series. He would be very informed by film and scouting reports, focusing on tendencies and less on local optima.

It seems to me the dominant school of thought now is to do the most locally optimal thing per possession. It's definitely less of a mental/emotional calculation and more of just referring to data and basically executing a simple min-max. But I think winning is largely a separate experiential skillset, based on very minute details. That's something Kobe loves talking about.

The only thing that would allow Kobe to be effective in diversifying his approach to the game would be talent and ability, which analytics has tracked pretty well for years now.


I really don’t think people understand just how good the right analytics can be at forecasting outcomes. There are guys who watch zero sports and literally farm millions through daily fantasy using algorithms similar to RAPM for forecasting box scores.

Things as violatile as PTS/TRB/AST/STL/BLK every night.

Kobe makes it sound like(or even says directly), the most important skills to have in basketball are untrackable and mental. Absolute ****.

It’s a game of substance, and talent. The ability to play it out a high level(which is always trackable) is BY FAR the most important aspect.

Quote is nothing more than arrogance and superstition.

[snaquille oatmeal]

Kind of rich coming from Kobe, a guy who went for numbers instead of winning for awhile.

Nah you are missing something in your criticism. Kobe wasn’t that interested in numbers as in “let me pad my stats” he was more interested in getting his rhythm and momentum back. He was heaving half court shots at the end of quarters and 24 seconds before it was trendy. He was doing those things because he was not afraid to ruin his numbers he was doing it to get back in rhythm.

[KyRo23]

Kind of rich coming from Kobe, a guy who went for numbers instead of winning for awhile.

Nah you are missing something in your criticism. Kobe wasn’t that interested in numbers as in “let me pad my stats” he was more interested in getting his rhythm and momentum back. He was heaving half court shots at the end of quarters and 24 seconds before it was trendy. He was doing those things because he was not afraid to ruin his numbers he was doing it to get back in rhythm.

That sounds a lot like personal momentum rather than team momentum...

[snaquille oatmeal]

Kind of rich coming from Kobe, a guy who went for numbers instead of winning for awhile.

Nah you are missing something in your criticism. Kobe wasn’t that interested in numbers as in “let me pad my stats” he was more interested in getting his rhythm and momentum back. He was heaving half court shots at the end of quarters and 24 seconds before it was trendy. He was doing those things because he was not afraid to ruin his numbers he was doing it to get back in rhythm.

That sounds a lot like personal momentum rather than team momentum...

Rhythm and momentum affects both, the individual players and the team as a whole. One carries the other and vise a verse

[KyRo23]

Nah you are missing something in your criticism. Kobe wasn’t that interested in numbers as in “let me pad my stats” he was more interested in getting his rhythm and momentum back. He was heaving half court shots at the end of quarters and 24 seconds before it was trendy. He was doing those things because he was not afraid to ruin his numbers he was doing it to get back in rhythm.

That sounds a lot like personal momentum rather than team momentum...

Rhythm and momentum affects both, the individual players and the team as a whole. One carries the other and vise a verse

Seems pretty detrimental to force shots like that. Does more for killing momentum than gaining it.

[snaquille oatmeal]

That sounds a lot like personal momentum rather than team momentum...

Rhythm and momentum affects both, the individual players and the team as a whole. One carries the other and vise a verse

Seems pretty detrimental to force shots like that. Does more for killing momentum than gaining it.

That you are right, but you have to consider that we are talking about Kobe under Phil Jackson and Tex Winter who strongly believe in those little things that change the game when rhythm and momentum swing your way.

[John Murdoch]

This is where its interesting to comp Magic and Curry in terms of all time point guards. Curry builds momentum by succession of 3's that build up a sense of overwhelming for the oppoment . Magic would literally out run guys and see over them

[Lalouie]

Not sure if there was ever a topic on this, the interview is not new, but Kobe was asked about advanced analytics and the modern game about 8:45 of this interview:

Watch on YouTube

The quote about advanced analytics is:

I think it has a place in the game. I don't think it gives the full answer of why things happen, and how to create momentum. The game really is about momentum, and who can control momentum, more so than numbers. There's the emotion of the game. Who can dictate tempo and pace. Those things are the most important, because those things actually win games.

Numbers look good, and can kind of tell you a story. But the true story lies in the tempo and pacing of the game.

Of course it's not possible to quantify what he is talking about (and that's the point he is making), but intuitively I feel like what he is saying makes sense. Numbers tell part of the story, but they will never tell the full story. It's why you can't say one player is greater than another purely by looking at the numbers (whether simple box score numbers or advanced analytics). I like his point that the numbers can tell a story, but not the whole story. I think some people lose track of this sometimes.

Thoughts?

there is more momentum in basketball than any team sport(and i don't watch hockey) that in such circumstances analytics becomes less and less important. you cannot quantify momentum. that's like trying to quantify luck. a lot of "momentum" comes from upstairs in between the ears. what the minds thinks or maybe even INTUITIVELY the mind actually thinks LESS, controls how the body behaves. ie if you are mentally checked out you have no momentum. there's a dynamic shift in team thinking when curry pops two 40'ers in a row as concomitantly the other team deflates.

my guess is analytics is more important to the defense, however when a team is hot,,,,,IT'S HOT

defense calculates more. offense not so much. bill russell said the celts only ran a half dozen plays but there were options off of each......and that was 55 years ago

[G35]

No one tries to "discredit" Kobe. It only comes up when people try and say he's top 10 all time. It's an insane debate where someone goes "Man, I think Kobe's the 12th best human in history at basketball", someone else goes "FU hater. Kobe >>> LAMEbron!" and then we get into a debate about why Kobe can't break into the top 10.

He is TOP 10 ALL TIME. No question.

He's not. And he simply can't break into. His peak is not good enough and his longevity doesn't get him there. All he has are his # of titles, and 3 of them come when he played with an actual top 10 guy. Anyone else he's up against as a potential top 10 guy would have won in his place.

See? It's this nonsense that starts it all.

Was Shaq a top 10 player in 2000?

You only get ranked higher AFTER you win titles.

Like Dwyane Wade would not be ranked as high as he is if he only won the one title in 2006. Or does that mean Lebron was playing with a top 20 guy for two of his titles?

See this is they hypocrisy you see all the time.

The problem with Lebron is (and you'll see it in the responses) all the players he played with got WORSE playing next to Lebron.

All the players that played next to Kobe got BETTER. All the players who were better next to Kobe:

Shaq won his only MVP next to Kobe
Pau became more efficient and a better rebounder next to Kobe
Lamar Odom
Bynum
Ariza


Then you look at Lebron

Wade was worse
Bosh was worse
Love was worse
Kyrie bailed

There is a lot of pressure on Lebron to prove he can enable talent like Kobe did multiple times. Its easy to be ball dominant and stack stats...its much harder to enable your teammates to be great at the same time, which Lebron has not shown he can do......

[XxIronChainzxX]

He is TOP 10 ALL TIME. No question.

He's not. And he simply can't break into. His peak is not good enough and his longevity doesn't get him there. All he has are his # of titles, and 3 of them come when he played with an actual top 10 guy. Anyone else he's up against as a potential top 10 guy would have won in his place.

See? It's this nonsense that starts it all.

Was Shaq a top 10 player in 2000?

You only get ranked higher AFTER you win titles.

Like Dwyane Wade would not be ranked as high as he is if he only won the one title in 2006. Or does that mean Lebron was playing with a top 20 guy for two of his titles?

See this is they hypocrisy you see all the time.

The problem with Lebron is (and you'll see it in the responses) all the players he played with got WORSE playing next to Lebron.

All the players that played next to Kobe got BETTER. All the players who were better next to Kobe:

Shaq won his only MVP next to Kobe
Pau became more efficient and a better rebounder next to Kobe
Lamar Odom
Bynum
Ariza


Then you look at Lebron

Wade was worse
Bosh was worse
Love was worse
Kyrie bailed

There is a lot of pressure on Lebron to prove he can enable talent like Kobe did multiple times. Its easy to be ball dominant and stack stats...its much harder to enable your teammates to be great at the same time, which Lebron has not shown he can do......

Players get ranked when their careers end. MJ wasn't the GOAT in 1988.

But, again, this just proves my point. It's almost prophetic.

[madmaxmedia]

I like his general point that the numbers are the outcome of successfully dictating momentum and pace of the game. But if that's all you focus on, then you miss out on how to actually achieve the end that results in those numbers.

Recognition, awareness, decision-making, poise, creativity, intelligence- those are all factors but you cannot actually directly measure them. Except perhaps the ultimate test that is NBA2K season mode.

[ccameron]

I think pre-analytics NBA thought is really interesting and is undervalued by blog boys all the time. Being dismissive and smart-alecy about the mental aspects of the game when you've never played NBA ball isn't more empirical; it's actually less responsible because you're actively dropping qualitative information that can either be quantified, be used to interpret the numbers, or both.

Kobe's approach to the game was deception. He would start off by imposing his will and getting the defense to react a certain way, and then do the opposite to throw them off. He would do this in something as small scale as a single move, or across possessions, or across games within a playoff series. He would be very informed by film and scouting reports, focusing on tendencies and less on local optima.

It seems to me the dominant school of thought now is to do the most locally optimal thing per possession. It's definitely less of a mental/emotional calculation and more of just referring to data and basically executing a simple min-max. But I think winning is largely a separate experiential skillset, based on very minute details. That's something Kobe loves talking about.

The only thing that would allow Kobe to be effective in diversifying his approach to the game would be talent and ability, which analytics has tracked pretty well for years now.


I really don’t think people understand just how good the right analytics can be at forecasting outcomes. There are guys who watch zero sports and literally farm millions through daily fantasy using algorithms similar to RAPM for forecasting box scores.

Things as violatile as PTS/TRB/AST/STL/BLK every night.

Kobe makes it sound like(or even says directly), the most important skills to have in basketball are untrackable and mental. Absolute ****.

It’s a game of substance, and talent. The ability to play it out a high level(which is always trackable) is BY FAR the most important aspect.

Quote is nothing more than arrogance and superstition.

Why arrogance and superstition? He is't saying "throw all the advanced mumbo jumbo out," he is saying it has a place. As in, of course you are going to need to get good at all those things that are measured in advanced stats. But what's wrong with saying the stats don't tell you why things happen?

When he says controlling momentum of the game is more important than the numbers, the obvious and dull thing to say is the team that scores the most points wins. Yes it's about the numbers. But I think it's a valid point that the optimal thing per possession is not always the best strategy in the long run. Anyone who plays cards could tell you that. What's wrong with saying the stats don't capture everything, or even that the mental aspect is the most important? You'll hear that from fighters and martial artists all the time, how the most important aspect is mental -- except you know that they spend hours every day training their body and perfecting their technique. That's just a given. For THEM, the most important thing is mental. I think that is Kobe's stance here. It's clear from other interviews Kobe spent hours analyzing player tendencies, etc., and to the extent these things were available, he probably did use this data. But for HIM, the mental aspect is most important. I see nothing arrogant or superstitious about this.

[MrSparkle]

Well said by one of the best.

Good example is the 2011 Bulls vs. the Heat series. As far as analytics go, Mike Miller and Haslem were done that season. Bosh was struggling all year long and had been outplayed by Boozer in the regular season. Chalmers was a scrub compared to Rose. So while Lebron and Wade are 2 elite players, the question was whether they could withstand Thibs and the Bulls waves of defense.

But what happened? The Bulls had no chance. Each positional match-up, Miami had momentum advantages. Lebron and Wade caught fire from the behind the arc in G6 and put the Bulls out. Before that, you had Miller and Haslem throwing back the clock to their younger healthier years and playing spry basketball and making open shots. And Rose didn't know how to pace or create anymore against Lebron's man-to-man defense.

I think Hinrich, Korver, Boozer and Deng were good examples of players who had 2 gears - no grasp of momentum nor pace, and you can see they always had mediocre results in a deep playoff series. Good for an impressive 50-60 wins in a regular season, but come playoffs they were a far-cry from the more versatile, 5-gear players. Without Rose, those guys (sans Kirk) lost to the 8-seed Sixers in a flash.

Ironically, young Jimmy, Belinelli and Nate had that momentum/pace thing going on, IMO. And no surprise to me, that underdog 12/13 Bulls team actually upset the more talented Nets. You look at their advanced numbers and there was nothing special there, but they actually did more damage than the "better" prior year team.

The Spurs are another good example. Analytics would tell you that some of those guys were cooked and not good enough anymore to win a chip, but Manu/Parker played their vintage roles in crucial moments - when they needed that momentum. Duncan, Green and the spare parts steady rocks during the season. In many ways actually, Duncan wasn't that momentum type of player - he was a stable 1-gear threat who always did the "right" thing, but Parker and Manu were the unpredictable magic of that long-time winning Spurs team. Kawhi took them to another level.

[NO-KG-AI]

Analytics said that Spurs team was way better than they had any business being. ^ They don't fit the narrative you're spinning at all.

[Chanel Bomber]

momentum is definitely a huge factor that rarely gets discussed/measured in these conversations

Kobe's a smart dude, I can listen to him talk ball for hours

I like the fact that he didn't dismiss analytics.

Kobe's dominance from 2008 to 2010 was a lot about his ability to have complete control over momentum and pace. His advanced stats were not particularly great (also because he never actively tried to protect his shooting percentages and he never hesitated to take difficult shots at the end of the clock), but he was a puppet master on the court. He completely dictated pace and had complete control over what was happening on the court. At that time, LeBron didn't have that ability yet, although he would obviously develop it later in his career (with a better statistical output than Kobe). Kobe was a lot better than his advanced stats suggest.