The Math Behind Basketball's Wildest Moves | Rajiv Maheswaran | TED Talks 

"This game is not about spaciotemporal patterns in kinesiology or any of that. It's about feel. And buckets. It will always be about buckets." - Uncle Drew

Very interesting. I think the description buries the most fascinating points: most NBA playoff teams were using this software, and the Ray Allen shot in game 6 of the 2013 NBA finals only had a 9 percent chance of happening.

As someone that works in the sports analytics field, videos like these are always interesting! It still seems like many teams are slow to integrate findings like these into practice and play. If your'e interested in learning about how to get into the sports analytics field, I have a few videos on my channel that complement this one nicely!

He keeps coming very close to teaching me something, and then backs off like he's afraid that we'd actually learn something.

Imaging an android Basketball coach with the AI mind of Gregg Poppovich...yes it's RoboPop. (Cue groan and eye roll)

we already have a machine in the NBA, its called Greg Popovitch

This is not a TED-talk, this is a sales pitch.

The thing I noticed was that because the info can tell you players who can shoot well but take bad shots they are worth more as training people to take good shots is much easier than training people to shoot better

As a fan and sports statistics maniac, this is great. As a coach, this is a bit threatening. As a human being, it's frightening.

The points guy @ 4:24

It's always nice to see a Ted Talk about the science and mathematics in sports.

I want a link to that chart

I didn’t realize this was uploaded 3 years ago and the first thing I thought of when I saw the bubble chart was that orange bubble at the bottom was Ben Simmons 😂😂😂

Good talk and presenter. Thanks for posting.

give this to charles barkley. he reject it in a heartbeat

Thats a good point in his closing statement about how this could help design buildings and cities with better traffic flow

8:59 I saw that game and that play

go easy on the haterode people! granted, the talk doesn't actually delve into the math. rather it presents a high level description of machine learning. don't blame him for the fact that you already know something about this stuff, i bet you many people who watched this video got something out of it, i certainly enjoyed it :) of course, it really is a really fancy presentation, but that does not in any way take away from the fact, that for example a key problems of machine learning (prediction) is explained lucidly and illustrated vividly by example of spatio temporal pattern recognition in basketball (to find 'good' features to predict the correct 'class' of pick and roll). if you look for something deeper and less lofty, take a MOOC on this topic, read papers or textbooks but don't watch a TED talk!!

Football... Or the other football. So, football and handegg. Gotcha.

So what does one do with the data involved in this? Predictive algorithms. I'm sure some element of this program is involved in the auto-driving cars. Unfortunately I see this program getting military application. Not only does this apply to missile defense, it could be a "tracking" technology that changes how wars can be fought on every level of battle (infantry, navy, air, space?).

Excellent video.

This can also (or might be already) used to track movement of terrorists and hopefully prevent future attacks. Great TED

That last part about tracking all kinds of movement sounds like a double-edged sword, very similar to how facebook data on its users were used

This is amazing, yet frightening. Amazingly Freightful

Questionable recreation, Ray Allen was fading away with a defender in his face

Nice 👌🏾👌🏾

It would be interesting to learn how exactly they calculate the probability of a shot. It strongly depends on player's movements and skills besides defenders positions and their angles. Obviously, LeBron James and myself have different probability of making a good shot in a fixed game configuration.

really interesting. the thing to me is that it doesn't go into detail enough for my level of basketball knowledge but i understand he has to appeal to non-basketball enthusiasts. That bubble chart is really interesting and I want a copy!

Who cares about basketball I want the math behind women's wildest emotions pls

Really want to work in this lab, much more interesting than my own research lol

The only quote that synced in my mind: "so here's a bubble chart, what's TED without a bubble chart''

I would like to know about the differences in movement between men and women.

Good talk. There are things machines still can’t quantify or predict and may never be able to do so but this helps give you a ball park.

Some one send this stuff to take two! Maybe they could learn how to create pick and rolls in 2k!

Great Ted Talk. As usual

A great solution looking for a real problem...

but the program doesnt calculate that the player has an argument with his girlfriend earlier in the day and he's in a bad mood. He takes it out on the screener by giving him elbows when they contact which in turn makes the screener screen differently, which then leads to the ball handler reacting differently to the defense. too many variables that aren't being factored in, just enjoy the game.

Wow his commentary on the park game was taken straight out of game 6 of the 2013 finals.

If anyone has a link or some advice how to get into this I would like to know. I am relatively new to the world of machine learning and programming and would like to learn more about this work. Thanks in advance

Thank you very much too for sharing!

As a spurs fan this talk made me sad

How can something like that be so interesting?

This can be implemented in a different department like in the Police. All the Police cars patrolling around the city could be tracked as dots and as soon as they find out something is going on, they could contact the police patrolling near to that spot so that the police could get there even faster.

Very insightful

how do we know that the clip he showed isnt edited

the coolest TED talk ive seen yet! get this man hired on the NBA 2K19 team! jaja

He actually doesn't ever say anything. He almost does a bunch of times.

Amazing use of technology on moving dots. However, I feel like there is more to the technology that I want to know more about... this video is too short.

"football, or the other football" this guy... I like him

Well that got really weird really quick

Great video!

How was the presentation done? Looks more difficult than developing the machine learning algorithms ;)

I would love to see this data available on the web.

Unusually clickbaity title for TED ... “... wildest moves”??!!

The way pressure effects individuals is impossible to model as it varies on a day to day basis even within the individual. Some coaches who can read this and see who is "feeling it" on a given play will still be necessary.. So integrate this stuff cyborg style and we have the best teams ever

@4:22 It's The Points Guy!

Probability prediction is one of the scariest things to me because as machines get smarter (by this i mean better able to process complex data) the more our movements are being monitored (gathering complex data), so the more data they have and the more likely they are able to predict what we are doing outside of the "monitored areas". Some would say that if your not doing anything wrong then why does it matter and to that i make the same argument back and would go so far as to say that if best guess in the human experience is replaced with perfect guess we trade off a "mistake learning" experience (the process in which we develop common sense) for one that creates a dependency on computers that could send the entire human race down a unrecoverable path.

Interesing technology, but wonder how useful it is at present. Like most new technologies of this sort, guessing it will probably take a while to be very useful & efficient.

First of all, great concept. I know that some don't like analytics in basketball, but honestly, it's here to stay. Look at the coaching difference between Steve Kerr and Byron Scott. But focusing on the technology here ... can we extrapolate it beyond movement? What about tendencies? Since we are talking about an algorithm that allows a machine to learn. Aren't technology companies like google and facebook already using stuff like this with personalized ads? I'm pretty sure U.S. Health insurance companies would like to invest in something like this. IBM already come out with commercials basically saying they want to track your medical history no matter which doctor you see. And something like that isn't something I'm for or against per se. While the technology could lend itself to making a more accurate diagnosis, it takes away a bit of privacy. While I have nothing to hide, I've also been programmed to fear things that even hint at being told I have a "preexisting condition". With my new obamacare insurance, I'm less scared of that now. I know this technology has been in existence in many various forms, but this talk reminded me once again of how fascinating and terrifying this could be.

very interesting!

Great !!

Great viedo

This talk is to advertise his software and let everyone know its being used by the top team. Sorry, no preview included. You have to buy it in order to try it.

Skip would be proud of this guy, bringing up the Ray Allen's shot, lol. The most clutch shot in basketball.

is anywhere I can get this on git?

So that's how Christian Laetner hit that shot to beat Kentucky in 1992 -- he out-spatiotemporal patterned the Wildcats!

Right now I feel 20% enlightened, 30% hopeful/optimistic, and 50% fearful... Why is that? Is this what big brother looks like.. The answer: Yes. My solution: We need to make as big advancements in Morality, honest brotherhood, and goodness as we do in science, art, and technology.

Where is the wild moves part?

4:24 - is that @thepointsguy?

What is the name of the app he's using?

is there anyone who can explain the meaning of the chart at 8:25 in Chinese? Plz~

Where can we get at this data? I want to know some more pattern's and data around this.

In making humans dots we can begin to lose sight of emotion , motivation, bonding , self control limits. Spatial temporal pattern recognition is fine even laudatory but there are blind spots as these persons dont have eyes in the back of their head though they have stereo and peripheral vision to a more or less extent. Sound ie voice recognition fills in these blind spots with team talk. For example the back screen - you're blind- a switch has likely occurred and a mismatch created or a backdoor lane opened. Sensory operation feel and especially sound has to be paired with the technical mechanics of motion . Similarly another example is the voice call of slide. switch , or stay this is critical in screen defense in horizontal diagonal or vertical screens to avoid initiated collision static and the problem of being picked off. Ok sensei what do we do now ?

I think athletes know the sport better than the software.... and sometimes better than their coach

its let me know a lot good speaker good talking

Where can I get this software?

Hmm... I am a complete and total basketball nut... But my eyes glazed over after about 4 minutes.

this is very cool!

Great

wow awesome graphics

8:30 I lost.

What's the product called?

To be perfect at some thing you have to deepen into it, this way will make you know the concepts at anytime and makes you realise thing from the mind point of view, so your actions can comprehend the knowledge. There are superficial concepts, and there are deeper concepts, just like math. normal math is explained by another deeper math knowledge that came from another concepts, and so on... every thing explain others, till it is called a miracle which is not according to the law of human mind to comprehend, but it is rather a mysterious concepts beyond human mind that might confuse the mind and might even make human lose their minds...

ray allen was doing a step back with tip toes at the corner three, so did the machine account the physical state of the play hmmmm

you are awesome

Really liked his talk except his wide scale applicability at the end which sounded dubious

USC research faculty doing a sale presentation

That’s kinda scary; we could actually be able to predict the future and even tho it would be an amazing and useful innovation, it would take something else out of our lives and I’m not sure if I want to lose it 🤔

"Rajiv Maheswaran and his colleagues are analyzing the movements behind the key plays of the game, to help coaches and players combine intuition with new data." Basically, nerds who could never play sports, want in on the multi-billion-dollar business of sports, that's generally getting less competitive and seeing more predictable results, thanks to a growing overuse of sabermetrics and biomechanical analysis. Not against "math for sports." It has its place, even in sports. Just don't like how there's this constant push (advertisement) of such concepts that sports somehow better need something that sorta defies the whole point of human competition sometimes. "Sabermetrical" and analysis-heavy sports teams like the Boston Red Sox are becoming almost TOO predictably the winners of everything. To the point where you almost ask, "Why even bother? They math away the human guesswork."

I have questions that I'm wondering if anyone will see and hopefully answer: 1) Isn't a good shooter BY DEFINITION a shooter who takes good shots (and likewise for bad shooters)? 2) How are things in your algorithms weighted? As in, what's more important: the angle of the defender of the distance between them and the shooter, and how did you decide how to quantify this? Is it effectively an educated opinion? The second question above is something I've wondered for so long, how do you choose the weighting of variables in an algorithm?? (I also don't know how to implement that but it's just syntax and not OPINION).

This game has always been, and will always be, about buckets.

Can these guys make the next 2k game please.

It’s 9% chance for the shot to happen and go in? A computer can’t detect a persons instincts or whether or not and how much someone has practiced a certain shot for these moments.

Moving dots aka Championship Manager 2001.

This is not TED talk. It's a fucking advertisement for his software. I haven't learn a thing...

fun in basketball ... scary across society

I still didn't learn much. I wish he would have gone deeper and given his take on how sports movement software could potentially affect everyday human movement.

Wow introduce one of the most clutch shots in the history of the game and all you can tell us about it was that it was a 37% shot

The title is a bit misleading, involving something like "data analysis" or "machine learning" would be better

Hindsight is always 20-20. Easy to 'predict' the game after the game is over. the machine gives Lebron 33% chance to make the shot which means 67% chance of missing it, and 88% chance of not getting an offensive rebound and 63% for Ray Allen not to make the tying 3. If you were to ask the machine before the play if Miami would win the game (a yes or no answer), there's 100% chance that the machine's prediction would be wrong. How's that for machine learning? A machine can't learn, it can't predict, a machine can process an extremely large amount of data, that's all.