A number of commenters have astutely identified the change in shot distribution over Kobe's career, which is due to a combination of age (older wing players tend to drive less and shoot more as they get older or experience injuries) and league trends towards more 3-point shooting.
If you're interested in reading an excellent analysis of the 3-pointer in the modern NBA, I highly recommend Ben Morris's article on Stephen Curry, whose historic shooting ability is so absurd, one could argue he is underutilized despite his high field goal attempt numbers.
The 2016 Warriors have Curry who broke his own record for 3-pointers in a season (and now holds the #1, #2, #4, and #7 slots for that record), but they also have Klay Thompson whose season was good enough for #3 all-time in the books. Thompson is shooting .420 to Curry's .444 so it's not like Curry's being underutilized in favor of a poor outside shooter.
Great stat I saw today: the 95-96 Bulls made 544 3-pointers that year. The 2015-2016 Warriors, if you took away every 3-pointer that Curry made, would still have 675. The game has changed indeed!
I was brushing up on Zache Lowe yesterday and came across this (extremly HN-friendly) gem from a few years ago[0] about the Raptors' program for mapping ideal defensive positions, and I couldn't get over the part where the analytics team was begging the coaches to make players shoot more 3s. I guess all it took was a coach known for being a sniper in his playing days having two of the three best shooters in the league to make that analytical dream a reality.
The game has changed because defense is almost non-existent during the regular season. The NBA (outside of the playoffs) is basically street ball at this point. Quite frankly I'm surprised people are still willing to pay money to go to a game and watch it.
Yes, and not only that - Curry probably wouldn't be as useful without the defensive spacing and assignments necessary to deal with not 1 but 2 incredible 3-point shooters.
The author of the 538 piece makes an interesting case that he may actually be underutilized. The gist is that his efficiency has not gone down as his number of shots has gone up, and that a "bad shot" for Curry is still better than a "good shot" for most other players.
There's a famous Kobe quote on his philosophy on failing. During a conversation at some game:
"Deron Williams went like 0-9. I was like, 'Can you believe Deron Williams went 0-9?' Kobe was like, 'I would go 0-30 before I would go 0-9. 0-9 means you beat yourself, you psyched yourself out of the game, because Deron Williams can get more shots in the game. The only reason is because you've just now lost confidence in yourself.'"
A similar quote by Michael Jordan: “I've missed more than 9000 shots in my career. I've lost almost 300 games. 26 times, I've been trusted to take the game winning shot and missed. I've failed over and over and over again in my life. And that is why I succeed.”
Similarly, I saw/read an interview with Hall-of-Fame point guard Isiah Thomas where he was asked why he always took the big shots in the critical moments.
(IIRC He didn't mention the obvious, that he was the best player on the team.) He said it was because he practiced the most, which was plausible, and because he could miss the shot. He said for if some players miss that shot, it crushes them; they don't recover. He could take it, miss it, and take it again the next day without it affecting his focus.
He made a few of those shots and led his team to two NBA titles. The first one was the year after he threw away a pass and the game in the playoffs, an unforced error that cost his team their first trip to the NBA Championship:
Reminds me of the Jordan commercial [1] where he mentions he lost more than 9000 shots in his career, lost almost 300 games and missed 26 game winning shots. Amazing commercial (like most Nike commercials).
If something as banal as a shoe commercial has a legitimately inspirational message about failure and success, you could argue society has reached a new high.
Hah! Perhaps. I don't think you can call it banal considering the investment, however—it's straight-up propaganda. I highly doubt Kobe's three-point shot has anything to do with his using Nikes.
shot percentage is a pretty arbitrary stat. a .350 batting average is also very good. a .900 hockey goalsave percentage is abysmal. to keep with the sports metaphor, you can always move your goalposts.
My biggest complaint with Kobe is that he scored a lot of points because he took a lot of shots. Last night for example, he took ~60% of shots from the field for his team which accounted for ~55% (ignoring FTs). In other words, the expected return on his shot was lower than the rest of his team. Now, FTs change that story a bit, but even in the best case scenario he breaks even with his team.
Kobe was a great leader on the court, which is what allowed him to get away with shooting more shots than the rest of his team combined on a regular basis. He also had this "The Lakers are built around Kobe" team structure that made that the expectation.
EDIT: There's roughly a finite number of shots a team can take in a game. If one player takes most of the shots, then sure they will score a lot of points. But are they maximizing team points? If other members of the team have a higher shooting percentage then taking less shots and passing the ball around will be advantageous.
This doesn't account for the possibility that his teammates have a higher percentage because Kobe is the biggest threat to take a shot and the opposing team sets up to defend accordingly.
Kobe Bryant being integral to the Laker's success is absolutely correct. The Laker's made the strategic decision to structure everything around him and it worked well. No doubt that would have failed with a less talented individual.
However, it is important to view his accomplishments in that light. He scored 40+ points in games because the team basically said "Let's have Kobe score 40+ points today". Because of that, his absolute point totals aren't great markers for individual talent.
not all shots are created equal - the "long 2" has a lower expected value than taking a few steps back for a 3-pointer, or finishing at the rim for an easier 2
we can't really hold kobe's last game against him imo - he's definitely been a ballhog but if it's his last game he should have the opportunity to shoot until his arms fall off
It's only arbitrary when using it against non-comparable stats. For example, comparing shot percentage to goal save percentage is pointless, because they measure completely different stats.
A quick search shows that an NHL team will generally get around 30 shots per game. If your goal save percentage is about 45%, then you're not going to have a career much longer when the rest of the league hovers around 94%.
How curious that there's a band of no shots that's right on the three point line. But then there are plenty of shots a foot closer to the hoop.
The explanation is probably psychological, you don't want to take shots that are as about as difficult as 3 point shots, but worth only two. But the situation isn't really any better if you take a single step forwards.
Of course, the positioning of the defensive players matters, they're usually just inside the 3 point line.
Isn't it just that if you're half a step away from 3 points, you might as well take the half a step back and try for 3 instead of 2 with a shot that's pretty much the same in difficulty?
I think it's more like they are going for a three point shot, and want to make sure it's actually in the three point area. Pretty wasteful being just over the line, and only getting two for the effort.
You two are saying the same thing (and are both right).
He's (presumably) trying to maximize points per possession. His shooting percentage is a decreasing function with respect to distance from the basket. And there's a jump in the "shot made reward function" at the 3 point line.
A step forward doesn't take you that much closer to make the shot significantly easier. Of course, you're a little closer but not that much.
A step backwards moves your balance and you have momentum backwards so getting an immediate upright jump shot without losing power/accuracy is more difficult.
Two different things at play here. At least this is my experience as a former competitive basketball player...
I think it's situational and individual. For a big man catching the ball at the top of the key, looking for an opportunity to drive is the logical thought, but if you're a point/shooting guard, you're much more likely (through a combination of training, psychology and physical attributes) to move laterally or take a step back. And beyond the individual level, there's the near-universal truth that a spread floor is better than cramping everyone in the lane.
Wing players that are trying to represent a consistent dual threat of slashing or shooting shoot a high volume from that range. Often times you only have half a second to get the shot off with good balance.
It's called practice. That area for a shooter is no-mans land. They either will setup outside the 3pt line or significantly inside the 3pt line. It is simply a spot no shooter, and particularly one like Kobe, would ever setup and wait for a pass or call for the ball from.
Not only practice, but also training. Every shooting coach in the league will tell you that the deep two is the worst shot in basketball. Kobe had it drilled into his head since he was seven years old that you never shoot a deep two.
I wonder if we'll ever see a maverick player perform well because he's taking these shots, trains for them, specialises in them, positions for them, in a world where defenders train supposing you'll position differently, and reposition rather than take the shot if you happen to receive a ball in a deep two position.
If you could shoot 100% deep twos, you'd only need to make 67% of 3-pointers to make it better to shoot 3's. The typical FG% is about 60% for layups, and 40% for 3's. You'd have to get better at these deep 2's than layups to make this worthwhile!
You won't because you're making it even easier on the defense. Trying to optimize for that spot would screw up the positioning for everyone else on your team. The defending team would have it easy because you've essentially killed all chance for high post plays, PnRs, and a good number of cut/backdoor type plays. It would take far more time than I have to breakdown the why, but the main factors off the top of my head are:
1. Remember the NBA has legal zone defense.
2. Additionally, help defense is a lot easier in today's NBA. Forced isolations would be very difficult and even if you do it, there is less ground b/w help defense and the attacker.
3. You have many more quick and/or lengthy 3&D type wings in the NBA today. Help and close-outs are easier.
4. Long shots mean long rebounds. Putting wings one step in means the offensive team would lose out on one of the main disadvantages of zones - inability to secure defensive rebounds.
5. That one player will also be a liability on transition D. He will always be one step behind in case the opposing team forces a TO and runs.
That's just scratching the surface, but there is a reason it's called no-man's land for shooters.
A long two is the worst shot in basketball from an analytics perspective. I know that sounds incredibly obvious, but it's one biggest changes to game theory that the basketball analytics wave brought.
I read this as: if he's outside of the 3-point line and open, it's in his best interest to get as close to the line as possible and take the shot. He'll never intentionally step just over that line and degrade down to a two pointer. This fact should really reduce the number of shots in the "just over the line" zone.
As you go in a bit, I suspect it's much less likely that he would have started open outside the arc. If a three isn't an option he would take the easiest shot available wherever it was, and "easiest" usually means "closer (if open)".
And finally, if he's just inside the arc and wide open, he's going to have time to move a step back to try the three.
So I think there are few situations where it would make sense to take the shot just inside the arc.
Well the defensive player guarding Kobe better be positioned next to Kobe!
Really, all the positioning is dictated by the offense.
Defense ends up just inside the 3-point line because the offensive player sets up just outside the 3-point line, not the other way around.
Kobe (and pretty much all basketball players) are continuously planning how they will get open and where they will shoot from.
Part of that plan is the knowledge that an almost-but-not-quite-3-pointer is a really dumb shot.
So they'll usually catch the ball at the 3-point line and either shoot or dribble substantially closer to the basket.
This is true. But I've noticed that when playing basketball I'm more willing to give a shooter an open shot in this area than behind the arc or a few steps inside of it.
It made me realize why some players are willing to take a shot with one foot behind the line: because they're open!
Those players, with the exception of Michael Jordan who is an exception to almost all NBA players ever, are forwards or shooting guards. Their job is not to assist, it is to shoot.
That said, basketball is more mobile now than it every was and passing plays such as the pick and roll are an essential part of any team's arsenal. A good player will assist his teammates as well as shoot. On the other hand, it's really not fair to pick out these players as good or bad since their circumstances have all varied. LeBron James played on teams with Dwayne Wade, Chris Bosh, Ron Artest, Kevin Love and other excellent players. Kobe Bryant played with Shaquille O'Neil; Michael Jordan with Scott Pippen and so on. It's easy to pass when the players around you are also capable of scoring. I honestly can't think of a great shooter who's played with Carmelo Anthony. That could be because he's a ball hog and they never get to shoot and are thus forgotten, it could also be that he's the only good shooter and a centerpiece of their team.
I think Jordan played as both a shooting guard and a point guard in his career. Sometimes be played them both in the same game. When he played for the wizards he played more of a small forward. During his comeback (and second championship dynasty) he was shooting over 40% from three point range, which is basically PG material. As a younger player he was a solid rebounder as well. He also won defensive player of the year once. That's a pretty versatile player.
Sure, he was frequently the primary ball-handler, but that's also been the case for Kobe, LeBron, and Kevin Durant. That does not make any of them a point guard. In fact, the elevated three-point percentage is indicative of a shooting guard rather than a point guard. Think Reggie Miller versus John Stockton. Point guards typically have a defender playing them close, which makes it harder to get a clean and reasonable shot. Unless, of course, you're Steph Curry, who can shoot from anywhere.
Perhaps the better argument would be that these players (i.e. generally the best players on the court at any given time, whomever they were playing), don't really fit the normal positions. In fact, as I pointed out above, the modern NBA has really broken down positional play. At one time you would have a point guard at the arc, distributing to a shooting guard for the outside shot, a strong forward or center to drive the basket or a small forward to move it to one of those players. Michael Jordan broke that mold in a time where that mold was still dominating tactical thinking.
In the past 5-10 years, however, the game has come to rely on a mobile, smaller player. Draymond Green and Kawhi Leonard are good examples. The screen play, one of the most popular plays in modern basketball, requires both players be able to shoot, since who shoots is dependent on which player the defender pivots to. In addition, these players become important defensively important because of their ability to quickly react to fast movement of the ball. It's not coincidence that both of the players I listed are considered top defenders, despite playing forward. Kawahi Leonard has won defensive player of the year twice, and that award is dominated by centers and, to a lesser extent, guards.
You're right -- although of all the great players, Kobe seemed the worst at making the players around him better. Of course, I have not watched the NBA lately, so don't take me too seriously.
While this is true, there is a lot a negativity around the fact that Kobe was the second coming of Michael Jordan in nearly every sense.
What I mean by that is he copied his style, his shot, his drive to perform, but you're spot on in pointing out it really was all about him. Jordan took pay cuts to get better guys around him, he relentlessly made guys play up to his standards.
Kobe took huge contracts, drove out other players who made the team better and refused to let others share his spotlight. What I remember most is in several games after Shaq left, watching come down the court and repeatedly looking guys off who were wide open, then either take a wild three point shot, or drive the lane and lose the ball.
I've never seen Jordan accused of taking a pay cut before. The lockout was because of his huge salary. I don't even think there was a salary cap then...
Which is completely different than both Shaq and Kobe's multi-year $70 - $100 million dollar contracts. By doing so, it freed the Bulls up to sign better players and deepen their bench a LOT more than other teams.
One could use the data to assist another NBA player to play an extremely long game of HORSE with him. I am certainly not an NBA player, so I expect my game with him would be fairly short.
It will be interesting to compare this shot chart, of the greatest scorer of his generation, to what Steph Curry's, who may be the greatest scorer of his generation, career shot chart ends up being.
Where Curry's game is extremely efficient, with most points coming from behind the line or near the basket, Kobe was shooting all over the place. Kobe's game was also much more reliant on getting to the foul line.
They were/are both prolific scorers but do it in completely ways.
There are a lot of players that can shoot as well as Kobe... the difference is that the group of players that can shoot as well as he did AND are his size AND share his athleticism AND are willing to sacrifice as much as him is a very, very short list.
It's a nice contrast with Steph, who is objectively already the best shot-maker ever at the NBA level. The group that shares his size and athleticism is relatively large, but his shot-making sets him apart.
At one time, Kobe's career arc looked like it might topple MJ's. I think he cloned his game after MJ to a fault. He copied many of his moves and his intensity, but he also copied his shot selection. It turns out he wasn't quite as good at making shots as MJ, so this is what you end up with. If he had been a more willing passer and a better teammate, he might have been the GOAT.
Their height and weight differences justify the different style, though. Curry is 1.91 / 86kg, hard to penetrate inside with force since he lacks the mass.
Jordan and Kobe are both 1.98, you can compare them better.
The differences in their gifts also account for a lot. If Curry continues through his prime like he's played the past two seasons, he'll be the greatest shooter of all time. Kobe was never a great jump shooter but was remarkable at finding ways to get shots, as well as getting fouled.
Typically, in the NBA, the best scorers fit closer to the Kobe mold than Curry mold, Jordan and Lebron, for example. It will be interesting to see what influence Curry has on future generations and if others will follow his path like Kobe and Lebron followed Jordan's.
From what I hear, kids are now shooting long shots like Curry when they goof around, where my generation (I'm 34) used to like playing on lowered rims so we could dunk.
Curry is a great shooter that also benefits from the rule changes to give shooters more space outside the paint. I never liked the constant hand checking allowed, and am glad they stopped it. I wonder how Curry would fare in the 90s NBA given his smaller frame. I think he would still be a great shooter, but could he get his shot as easily? Debates like this are what makes sports fun.
It seems like a fairly common sentiment when comparing the '96 Bulls to '16 Warriors is that the Bulls would win under '96 rules and the Warriors would win under '16 rules.
There were also players who shot the ball extremely well before the three-point line took effect. Steph is averaging around 30 points per game with 5 three's a game. Without the additional point, it drops him to 25 points per game. Still good but probably not an MVP shoe-in anymore.
I think historical comparisons like this are just generally nostalgia. In general, players get better over time. There is more competition for a limited amount of slots, so the competition is more fierce. Players start training with better intelligence from an earlier age. They push harder. All things being equal (they seem to be on paper), the more recent team would have the advantage.
Removing the three point line would also detract from Durant, James (his shooting has been a big part of his game the second half of the season), and Westbrook. I think Curry would still be a major contender.
They all utilize the three, but none of them feature it quite like Curry. KD has a lot of shooting in his game, but Westbrook and Lebron are both relatively poor three-point shooters and don't shoot close to KD's volume, let alone Curry's.
Lebron and Westbrook's effectiveness of offense are largely based on getting to the rim. In fact, San Antonio has faced Lebron in three NBA finals and each time they deployed a strategy designed to tease Lebron into taking jumpers rather than drive or play inside. Doing that against Curry would be a death wish.
I've noticed this trend as well. I play pickup games with college students and noticed a major increase in the volume of three pointers. The game is definitely changing.
Out of curiousity, how old are you? Those were probably the biggest names in basketball in the 80s. It's also possible that they are known in Europe due to being co-captains of the first US Olypmic basketball team to allow professional players (Though they were past their prime at that point, and the remaining players were near the top of their NBA careers).
Sure, now that you mention him. And I'm sure I'd recognize some other old names upon hearing/googling, too. Just not in my active memory as basketball is a bit fringe sport around here.
Well, I know who he is, but from a sample of not-very-many in the UK, my wife had no idea and my colleague sat next to me (male, early 30s, loves cricket) also had no idea.
Manually. Companies like Stats(that also now running the optical tracking in the nba) has rooms full of people watching videos of games and tagging events.
I know a lot of guys that work at stats. Some of it is manual but over the last couple years they have actually build some pretty cool stuff to track all this. The company has done three rounds of massive layoffs over the last year though so who knows if advancements will continue.
> It began use in the NBA at the start of the 2013–14 season, and it is now in use in all 30 arenas, following trials during the previous season in 15 arenas.[
Which I guess means before that they had some guy who just recorded shots on a board. Would be interesting to see how it was collected prior to this technology
Knowing almost nothing about Kobe Bryant (I haven't followed the NBA for years), it looks like there's an interesting trend in 3 point attempts. At any given year of his career, the distribution of 3 point shots made compared to other shots made seems fairly balanced, as one might expect.
However, the number of attempted 3 point shots (or 3 pointers missed) seems to rise significantly in the latter half of his career. I wonder if there's something to be said about an increase of confidence (deserved or not) behind the 3-point line, and whether that extra point outweighs the increased likelihood of missing as an overall statistic (obviously it could make the difference between winning and losing in a single game). Or perhaps it's indicative of a trend for the NBA overall with more 3 point attempts in recent years. Or I could be seeing something in the data that isn't there.
Part of the reason may be that three point shooting has become much more popular in the NBA. Consider it: In many cases, if you shoot from a few steps further back then your reward increases 50%. Current MVP Stephan Curry is the epitome of this trend.
Also, driving to the basket, probably his leading alternative, takes speed, strength, and the ability to take a beating. Probably all those abilities declined as he aged.
Great catch. What you're observing is the effect of a new trend (or a nonstationary distribution, if you prefer) wrt. shooting tendencies, as it affects the coaching and strategy of one (star) player. Your suggestion that
> [the] extra point outweighs the increased likelihood of missing as an overall statistic
is a relatively-recent discovery for people in the NBA: the expected point value of a three-point shot is much higher than the expected point value of a very very long two-point shot.
I think it's primarily health & age. It's less taxing to take shots outside than to try to drive up to much bigger, younger players constantly.
I'd also say that as he got older, his athletic ability declined (injuries and age), so in order to continue getting points, he needed to attempt more shots.
Say what you will about the NBA (like the fact that they're about to vote to have corporate logos on jerseys in the near future), but what they're doing with data science, and especially the open sourcing of it, is a gold standard not just for sports but for many other industries.
Section I-Player The player is out-of-bounds when he touches the floor or any object on or outside a boundary. For location of a player in the air, his position is that from which he last touched the floor. [0]
It depends on how they picked the points where he 'took the shot from'. You generally don't take a shot from a standing position in normal plat, you jump. You might not always land in the same spot from which you launch.
I suspect they picked the point where he jumped from, rather than too, therefore you can run parallel to the backboard, jump under it (from behind, and land in front) and lay it up so it goes in.
The colors are bad for this visualization, but were chosen because the Los Angeles Lakers wear those colors, and Bryant spent his entire career with the team.
It's amazing that there's a noticeably higher hit rate on a very precise line right down the middle and a relative shot gap 10 degrees on either side before reaching his "sweet zone". Anyone know why being only a few degrees away from center makes it harder to score?
partly, it's that totally open shots (where there wasn't a defender in the area) will tend to be charted as coming from dead-center. So if Kobe gets a steal and has a clear path to the basket, or a defender gets out of position, he's going to get an easy shot which is marked as being taken from the center unless it's obviously taken from waaaaay off to the side.
Relatedly, shots that are marked as being taken from slightly to the side are more likely to be contested, often by a much bigger defender.
When you shoot from directly in front of the basket, if the shot hits the rim it can bounce off the rim, hit the backboard, then bounce back into the rim. This is actually pretty common.
When you are shooting from the side and miss, the ball will be much more likely to bounce off the backboard to the side.
When I saw that I just assumed that when playing he has practised shooting from the centre line and prefers to take that extra step on the line before taking a shot.
Anyone know how the data is collected? Is it computer generated by some software watching the game or is there someone faithfully capturing this into some database? Does the data exist for other players.
I'm a bit amazed that this kind of data is captured and wonder how accurate it is.
See my comments below on how the data is actually collected.
But yes, there is some opencv in this project(and probably in almost every other commercial machine vision project out there), but the smart parts are hand tailored, opencv have useful general purpose algorithms, but to get the best results you need to dig deeper.
You could probably get better AI listening to the sounds of the ball hitting the ring/back board/etc rather than studying images with different angles and players an shooting styles.
If you're interested in reading an excellent analysis of the 3-pointer in the modern NBA, I highly recommend Ben Morris's article on Stephen Curry, whose historic shooting ability is so absurd, one could argue he is underutilized despite his high field goal attempt numbers.
http://fivethirtyeight.com/features/stephen-curry-is-the-rev...