Stephen Curry and the duration of the great stagnation

Stephen Curry set a record In May of this year:

It took Reggie Miller 22 games to set an NBA playoff record of 58 three-pointers for the Indiana Pacers in the 2000 playoffs. Now, Stephen Curry has broken that mark in just 13 games.

He is now up in the 80s I believe.  Curry, by the way, is NBA MVP and his team is probably on the verge of winning the Finals.  The three-point strategy seems to be working: for Curry, for the Golden State Warriors, and also for last year’s champions, the San Antonio Spurs.

Yet the three-point shot has been in the NBA since 1979 (!), and for most of those years it was not a dominant weapon.

What took so long?  At first the shot was thought to be a cheesy gimmick.  Players had to master the longer shot, preferably from their earliest training.  Coaches had to figure out three-point strategies, which include rethinking the fast break and different methods of floor spacing and passing; players had to learn those techniques too.  The NBA had to change its rules to encourage more three-pointers (e.g., allowing zone defenses, discouraging isolation plays).  General managers had to realize that Rick Pitino, though perhaps a bad NBA coach, was not a total fool, and that the Phoenix Suns were not a fluke.  People had to ponder the expected value concept a little more carefully.  Line-ups had to be smaller.  And so on.  Most of all, coaches and general managers needed the vision to see how all these pieces could fit together — Arnold Kling’s patterns of sustainable trade and specialization.

In other words, this “technology” has been legal since 1979, yet only recently has it started to come into its own.  (Some teams still haven’t figured out how to use it properly.)  And what a simple technology it is: it involves only placing your feet on a different spot on the floor and then moving your arms and legs in a coordinated (one hopes) motion.  The incentives of money, fame, and sex to get this right have been high from the beginning, and there are plenty of different players and teams in the NBA, not to mention college or even high school ball, to figure it out.  There is plenty of objective data in basketball, most of all when it comes to scoring.

Dell Curry, Stephen’s father, was in his time also known as a three-point shooter in the NBA.  But he didn’t come close to his son’s later three-point performance.

So how long do ordinary scientific inventions need to serve up their fruits?  I am a big fan of Stephen Curry, but in fact his family tale is ultimately a sobering one.

Addendum: Tom Haberstroh fills in the history.


Here's a piece on the three-point line in college:

sports, like the military, are strongly conservative and resistant to change - something about the mindset of coaches and officers I suppose. I'm not sure you can draw overarching generalizations from them; or, if you do, those generalizations may skew pessimistic.

I would say that it's not really sports or the military so much as it is humans, human culture, and human organizations more generally. Status quo bias is quite strong everywhere.

I was talking with a professor who is also a mentor to me, and he said, "If you look at the faculty in most of the arts/humanities or social science departments here, they'll vote Democrat or Green. So yes, very liberal and progressive and all that. But you bring up some minor change to departmental governance or the curriculum in faculty meetings, and they HATE it! Organizations are inherently conservative, and I mean that with a small-c." (the professor was quite liberal himself)

A lot of the articles on the rise of the 3-pointer talk about NBA culture, and about how the 3-pointer was seen as less masculine and manly (another big factor in human affairs - gender norms). Even Poppovich, one of the most analytical of coaches, "hates" the 3-pointer - this is an emotional, gut reaction to how he feels the game "should be played", which come from his formative experiences and socialization into and of basketball. My point is that the story of the 3-pointer's rise is definitely a sociological story in which sociological factors basically slowed down the adoption of an obviously optimal tactic

I agree with most of what you said, but Pop's antipathy for the 3 point shot could as easily be attributable to his opinion on the aesthetic appeal of how the game is played in a 3 point environment vice a 2 point environment.

in that same vein, i think that it's especially difficult when the innovation has to take place under immediate and intense public scrutiny, and when the people executing the innovation (the players) are subordinate to the innovator (the coach). nobody wants to look a fool, and nobody wants to be told to do something that they worry will make them look a fool.

Sports are the opposite of conservative. It used to however that they didn't routinely change the rules to make the game more attractive TV audiences. When the rules stay consistent the received wisdom which is really just the intergenerational scientific method is for the most part correct. That's why innovative Sid Gillman never even made a Super Bowl. Defense and running the ball did win championships.

Now the rules are changed with alacrity and so it is more beneficial to innovate. But of course this article fails to grasp that 90 percent of this effect just like the QB effect in football is rule changes.

That's why Moneyball was something of a red herring. Baseball was the one sport where the received wisdom on scouting might have been incorrect. Also like Billy Beane argued it's really hard to scout high school kids. But the take away from Moneyball wasn't baseball scouting is off kilter, but instead that these stupid jocks don't know math and statistics. Obscuring the fact that the only reason that advance stats pay dividends is because rule changes made the three pointer a lot more attractive and the only reason that defense no longer wins championships is because you can't even breathe on a receiver anymore or you get called for PI.

Regarding the bizarre mix of conservatism and experimentalism in sports, see this Klosterman piece, which is the best article ever written about college football.

I don't think it is all rule changes that upset the recieved wisdom.

For instance I can't think of a rule change that lead to the adoption of the T-Formation. And today's modern spread offenses would have given older defenses fits under any set of rules.

That's true. But keep in mind but of those strategies bore their best results in college not the NFL. A lot of college formations evolved out of the kind of players each university had access too. That isn't as pressing an issue in the NFL which doesn't have recruiting. That's part of the reason why college sports have remained as popular as they are despite the vastly inferior level of play.

Innovation also often involves risk, and most coaches are risk averse for structural reasons: everyone sees the magnificent experiment that blows up in disaster. No one really notices the conservative coach that does an ok job for a few years. Why do so many football coaches fail to go for the two point conversion given it's high expected value or go for it on fourth and one within their own 20 yard line?

Although with the average coaching tenure in most sports getting shorter and shorter, we might see more innovation as the payoff to being a conservative coach is reduces; Chip Kelly might be ground zero for this in the NFL, although even he's reverted to conventional football wisdom in his second year quite a bit compared to his first year in the NFL.

Of course, I'm not sure how much this applies to other sports. The NFL is the only thing I follow closely.

That doesn't make the NFL conservative though. Once an innovation proves seemingly successful every coach rushes to copy it. That's the opposite of conservatism. Risk aversion is an element of conservatism but it's hardly conservatism. Sports is far too winner takes all and results oriented to be conservative in any sense.

Also, the expected financial rewards of conservatism are still the highest of high. The three point shot is not disruptive in the same way as other technologies. Otherwise there would be no 7 footers left.

The fact that Andrew Bogut is now a bench player for the Warriors might be the harbinger of exactly that....the demise of the big man in the NBA

Andrew Bogut isn't a very good basketball player. Not to mention the contract he signs next year will be quite a deal larger than an excellent wing player like Trevor Arzia's.

Hmmm...having trouble reconciling those two sentences.

Big men are so in demand that even bad ones get excessively paid. This would be the opposite of the conditions that would prevail if the age of the big man was in eclipse. Hmmm... not so hard.

OK fair enough, but he's at least a good enough basketball player to command big $ even if it's just for his size. And be honest he's not a bad player. He's very good defensively and a very good passer, just no good on offense.

Somehow, big men still dominate top 10 draft picks, 2015 draft is going to be headed by two big men (Towns and Okafor). There’s some kind of obsession with getting a dominant big men.

As long as the rules allow for fouling inside the paint, players will take more 3's. Imagine applying the same criteria on 3 pts shots that they apply inside the paint.

This is a surprisingly depressing post about three-pointers.

What, that we can figure out the earth's climate in its entirety faster than we can figure out a basic strategy in an infinitely smaller, closed system?

"we can figure out the earth’s climate in its entirety"


He's up to 95 through game 5 of the Finals!

you missed the "and.. part of the article.

Basketball people didn't even figure out the right technique for shooting a ball until the 1950's.

When the three-pointer was instituted, the NBA (and other levels of basketball) was a much more uptempo, fast-breaking game than it is now. It wasn't until the pace slowed down that teams were able to see how crucial the three was in a halfcourt offense.

Is the set shot less accurate than the jump shot or one handed shot? From what I understand, the one hand jump shot was developed to evade defenders while shooting.

The set shot is significantly more accurate. That's why you don't jump when you shoot FTs. The jump shot was invented in reaction to players' increased athleticism and range, since set shots were getting blocked.

An unguarded free throw-style set shot is plenty accurate, but the shots people were taking from open play prior to modern shooting techniques were not. Shooting percentages in both the NBA and NCAA didn't even crack .300 until about 1950. Watch footage from the 1940's and 50's and you see an impressive array of shooting techniques, many of which would have gotten you laughed out of the gym a few decades later.

That's true, but most free throws are done one handed, jump shot style, just without the jumping.

The three point shot has in recent years transformed the game into a dull shooting contest. Turn on an NBA game and you'll regularly see players on fast breaks halt before the line to take a three instead of go for an uncontested layup. I agree with Larry Brown and Pop (who nevertheless makes excellent use of them) that the 3 is ruining the game.

I've literally never seen this happen. Perhaps a poorly-contested layup, but not uncontested.

I'll probably screw this up: But, I find it weird that, on an economics blog, we find people bashing individuals for exercising their comparative advantage.

I get more utility out of an Allen Iverson dual threat player than I do a pure three point shooter.

What are his dual threats, assault and battery

This calls for 4-point or even 5-point options at the NBA level. There is no great stagnation, NBA rules edition.

Especially considering that a four or five point shot would merely be a simple technology, involving 'only placing your feet on a different spot on the floor and then moving your arms and legs in a coordinated (one hopes) motion.'

Imagine the excitement of watching that simple technology be used to propel the ball for 3/4 of the court to earn 5 points. After all, the 'incentives of money, fame, and sex to get this right' would make it easy to see the results of such a rule change, right?

Just divide the court into point ranges, 1 - 10 (or higher), where the average expected value across all ranges is equal given league average shooting.

I've often wondered about the lost Borges story of the young boy with the magic gift -- every three pointer he made would go in. And yet, his gift is unrealized and unknown. Then I realize that poser Shyamalan already made that movie and I go back and make another 2000 on e-minis.

Wasn't the three point shot start with the ABA in the 1960s and the NBA started in 1979? So it has been around longer. Also the NCAA implemented it in the mid-1980s when college basketball was still the main NBA training ground.

It is important to remember that most new innovations take longer to actually implement. At this point or even five years ago, drones could have been delivering pizzas to your house but it is going to 10 - 20 years before it happens.

This is another labor that robots can free us from with superior results. (Not my idea: see Vonnegut's Galapagos. The robots will be on cereal boxes.)

I bet you still get good odds on a final four of MIT, Caltech, CMU, and Duke. Always Duke, alas.

And by the way drones will not be delivering pizzas. Average pizza is over. Perhaps a driverless wood-fired oven will cruise your neighborhood and cook one for you while you wait, it takes 3 minutes.

Boring. How 'bout them Blackhawks?

They they will compete with the Bowman/Babcock Red Wings as a dynasty?

In a way, the Blackhawks are just another Bowman dynasty. Scotty Bowman is an advisor. Son Stan is the General Manager.

Really? Blackhawks? As a team name? In 2015? UGH

Urso, we need to remember the name of this great tribe of people. Stop trying to expunge their memory from the earth.

Is this a brilliantly understated joke? I'm terrified to comment and otherwise destroy comedic gold.

Urso prefers the Chicago Redmen.

The Chicago Noticers

Has Tyler ever had a post about hockey?

Proficient shooting likely requires work from a very early age. Before the three point shot, perimeter shooting was strongly disfavored. Thus, propsective players did not work on their shot as much, and the pipeline of NBA quality players who were also shooters was quite small. Even if everyone realized right away that the 3 point shot was important when the rule changed, I would expect a 25 or so year lag between that time and the time when enough players who could shoot were in the NBA to make a shooting-heavy offense viable for more than a few teams.

This is an important lesson for other spheres. Changes in human capital happen much more slowly than changes in the law. We can provide incentives for people do X, but many skills cannot be acquired quickly or at all by older workers.

"It's really funny," Bird says, "I never even practiced 3-pointers. We might have thrown up only a couple of them. The only time I practiced them was right before the 3-point contest in 1988. Danny [Ainge] would get the rack out and we'd rebound and throw the ball back out and shoot some 3s. But we didn't fire up 100 3s after every practice."

Three-point shooting was undervalued in the '80s and '90s but its current prominence is not a reflection of its intrinsic value to "basketball". As you allude to, the game is played by different rules today than it was 20 or 30 years ago. And those rule changes have made 3-point shooting more valuable. Permitting formerly illegal defenses reduced the cost to playing good three point shooters (who tend to be poorer one on one defenders and rebounders), reduced the value of slashing/driving and posting up (which is much more difficult against formerly illegal defenses) and increased the premium on players who can space the floor because of their outside shooting. Banning hand checks also created more space for jump shooters and reduced the premium on strength and physicality. These changes increased the incentive for teams to play players with a certain skill set (3-point shooting) over other skill sets (1-on-1 defense, post-up and rebounding) and gave all players incentive to prioritize three point shooting over other skills. Its just not the same game.

good point

Can we prove that this is truly a stagnation - a human failure in the inability to find an ideal strategy - versus an evolutionary system whereby different strategies predominate at different periods of time?

If the Cavs had Shaq in his prime, I don't think small-ball would work. How much does the rise of the 3-pointer reflect better understanding (it contributes, I'm sure) and how much does it represent the current lack of dominant big-men? We may have a natural experiment ahead when Towns and Okafor enter the league next year.

There's been a lack of dominant big men for a long time.

Here's something nuttier than how long it took to get good at the 3 pointer: the lack of imitation of Kareem's skyhook.

And Rick Barry's free throws

Barry's form was actually the common form in the 20s and 30s of shooting free throws.

Insane. Probably seen by guys as nearly unmanly as underhanded free throws.

Could someone clarify why Tyler is referring to this as an invention? I understand that the rules that permitted the 3-point shot are an invention, but is he referring to the use of the shot (and its increasing popularity) as an invention? Is this not more about learning?

I also understand that Curry is one of the shorter players and that player height peaked in the early 1980s and then again in the early 2000s. But I can find no data for median player height and I wonder if outliers influenced the averages of the peak years. I wonder if median height has increased and, putting Curry aside, this explains the increased popularity of the 3-pointer.

@RM The inventions are the techniques and strategies to leverage the 3-point rule. Someone had to learn to train people to be better shooters. Coaches must invent strategies and refine them against counter-strategies. The balances change as the shooters get better.

Technology is generally like that. For instance, for at least a decade it has been more economical (per unit of compute power) to make CPUs with 100 slower cores rather than 1 fast core. But software has to be written differently to preform well on these. Software developers are not ready, as a population, to productively exploit more than 1 core at a time. (Tools and techniques have to change, and then there is all the old code which is barely possible to rewrite.) And it does not pay to act until there are a lot of multi-core processors out there. (So today CPUs we have about 4 cores, which is a compromise.) The "Invention" of the multi-core processor is not merely something that chip companies do in isolation.

The defensive rules changed fairly recently (before the 2001-02 season).

Before then, you had to guard your own man (who may have been standing on the 3pt line) or you could double-team the guy with the ball.

After the rule change you could go wherever you want provided you don't stay in the paint 3 seconds without guarding anybody.

For a while, teams didn't drastically change how they played defense. Until Tom Thibodeaux came along and installed the idea of packing the paint to stop the most efficient shot in the NBA- layups and dunks. This most prominently happened when he was a Celtics assistant I'm 2007-2010. So about 5 years after the rule change.

Of course now offenses have evolved to beat the pack-the-paint strategy by shooting more 3's, about 5 years later.

Anyways, I would argue that changes do indeed happen relatively fast.

Im the Finals, the Cavs have responded to the Warriors 3pt shooting by slowing the game to a crawl, to some success considering their injuries.

Here's a good Grantland article on the subject.

The NBA is a monopoly, which limits the count of people in it, which limits the count of innovators.

Firms subject to regulations that restrict their earnings engage in inefficient behavior. NBA has a salary cap, which suggests that players engage in inefficient behavior, possibly including less innovation. Since the salary cap works at the team level, the inefficient behavior should show up at the team level, but not necessarily the individual level. Cowan says that three point shooting required changes at the team level, so the delay in adopting three point shooting could be team-level inefficiency.

The NBA does revenue sharing. Until a few years ago, it "taxed" teams that exceeded their salary cap. Now it just redistributes revenue. Both are limits on team owners' earnings, which could lead to inefficiency at the owner level. Owners might be more inclined to hire managers they like rather than managers who manage efficiently. Same with coaches.

Also, for owners, winning is not necessarily the most lucrative earning strategy.

Another thought: Local governments subsidize major league sports teams. How much do the subsidies distort the incentive to innovate?

How to test this; do innovations come from the NBA, college, high school, or pick up games? Where did three point scoring first become common? What are the pay-offs for winning at each of those levels and what are the restrictions on the pay-offs?

Science resembles sports in a number of ways (subsidies, high barriers to entry, high status research organizations pick the star researchers from lower status poorly funded organizations, etc.), but many of the incentives are different. Can we compare the rate of innovation in one with the other with out knowing how incentives are structured?

The NBA is a monopoly but most Major Sport Leagues seem long term exist with a monopoly. There was an alternative league the ABA which lasted a while until was merged into the NBA right before the 1980s glory.

Anyway, the strange thing of the Major Leagues is when first created the leagues economically marginally profitable. At some point the major league has competitors most of which disappear after a years. Then as the league gains popularity, a successful alternative starts existing (American League, AFL, ABA) but the situation is not stable as the successful alternative league is merge into a monopoly league and the league enters a new glory days. (NBA - 1980s, NFL- 1970s and Baseball-Turn of the century)

I think the primary reason for single Major Leagues is because the league needs economically viable losing teams to survive. Otherwise the league has bankrupt losing teams which will disappear in the middle of season. Just read 1890s baseball leagues to see the instability of competing leagues. (That it makes Championship & Playoffs better.) I do wonder if Professional Teams can go global?

Back when the NHL had six teams, the owner of the Red Wings also (basically) owned ~1.5 of the other teams, in order to give the Red Wings someone to play

Caveats: I forget the source (probably Grantland) and my recollection isn't perfect.

It looks like it took a little while for NBA players to shoot the 3 pointer well. Until 1986 the entire league shot less than 30%. Then it generally increased to around 35% in 1995, about where it's remained since.

Part of this may be that teams undervalued the 3, so didn't recruit as many shooters as they should have. But it may also be that it took a while for there to be a generation that grew up with the 3 point shot.

Effective field goal percentage, which is value per shot, has remained remarkably steady since 1979.

Yet, for all these 3 pointers, without the injuries to Love and Irving, the Warriors would have already lost.

The Warriors would be the weakest NBA champs in a decade.

The Cavs are much much much better on defense without Irving and Love on the floor.

"The Cavs are much much much better on defense without Irving and Love on the floor."

True, but the additional offense makes up for it. GS could never have so successfully "gone small" with Love and Irving around. The Ca

Plus, the exhaustion factor would not be so great with a bigger rotation.

The Cavs would have been the 7 seed if they were in the West.

They were 33-3 with mozgov, love, lebron, and kyrie. That is the best team in the league, not a seven seed.

Contrary to the claim made in the post and by many commenters, becoming a proficient 3-point shooter does not require a man to have been trained in the art from infancy. Common sense should be enough to tell you that pre-pubescent boys generally do not have the upper body strength to shoot from three point distance using a correct, adult form. Since he became semi-famous Kyle Korver has said in a number of interviews that he only became physically strong enough to shoot with his current form when he was 17.

I think the example of Larry Bird best gets to the heart of the matter here: he was only one of two players in the inaugural three point year of 1980 to shoot greater than 40% from three while taking at least 100 attempts. In this case at least, we know he had the potential to use this shot very, very effectively his entire career. But in the next four seasons Bird averaged only 0.9, 0.7, 1.0, and 0.9 attempts a game. His 3-point percentages in these seasons were poor, because at this low frequency "real" 3-point attempts are mixed up with last minute heaves from half-court at the end of a quarter. When he once again reached the 100 attempt threshold in 1985 and beyond his percentages were an extremely good 43, 42, 40, and 41. The real problem in the 80's NBA was not that it necessarily had NO good three point shooters, but that coaches insisted (as many high school coaches still do today) that the three pointer is a "low percentage" and therefore bad shot. Ultimately, this failure to exploit three point attempts is attributable to the incredible failure of coaches to understand the concept of expected value, which is usually taught to students in around the 10th grade. It may sound impossible that so many people could be so dumb for so long, but it's the truth.

In basketball, illogical playing styles seem to persist longer than other sports, for reasons that are mysterious to me. As a great example, look at the way the game used to be played in the 1930's:

Respectfully, upper body strength doesn't have much to do with good shooting form. Distance and arc come mostly from the legs, with the arms and hands mainly just for creating a higher launch point and for creating backspin on the shot for a soft landing in the case it's not a direct make.

Look at some of LeBron and Dellavedova's shots toward the end of game 4 when they got tired. They couldn't stop hitting the front of the rim, and I promise you their legs were feeling soft, not their arms.

Your point about expected value is right on, though. Similarly, I'm consistently amazed by how often coaches misvalue timeouts and fouls. I'll stretch the point here but how many times do you see an NFL coach call one of only three timeouts early in a half to prevent a 5-yard delay of game penalty? That timeout will be worth WAY more than 5 yards late in a close game.

Okay rant over, I have to go back to work.

There is a bare minimum of upper body strength needed to have good 3 point shooting form. Young men usually don't have it until well into puberty. Derek's point stands.

Lou Boudreau throws a sick behind-the-back pass in that video.

Maybe the real story in the finals is the decline of NBA centers, and the rise of small ball. The Warriors remove their center and the Cavs can't take advantage. We are seeing that more and more. Is it a rules change, a defensive strategy change, or what? Why can't centers score over 6-6 players in the low post? Why are centers often picked at the top of the draft, and then in the last 5 minutes of NBA games, when everything is on the line, teams take them out and go small ball?

It's also worth noting that apart from LeBron, the Warriors are facing an almost unbelievably weak finals opponent. When you have two castoffs from the Knicks, a power forward that can't shoot, Matthew Dellavadova as point guard, and virtually no bench, how is this series even 3-2? (Yes, I know.)

And the playoffs this year have seen more than one player who can hit threes but not free throws. That would have been inconceivable in 1980.

They are down 2 of their 3 best players. The Cavs were the stronger team at the end of the regular season.

"how is this series even 3-2?"

Because the Warriors are good but not great. Their talent after Curry and Thompson is not overpowering either.

Tim Duncan still does pretty good, but he's a hybrid PF / C. If Okafor doesn't turn into a scoring machine in the next few years, then the era of the scoring center will truly be over. Unless he's drafted by the Knicks (likely), in which case a lack of performance won't be his fault.

Oh wait, the Knicks are drafting 4th, not 2nd. There may be some hope for Okafor yet.

Free Throws are undervalued. Every player should be shooting 90%+.

Really low FT percentages do baffle me. Sixty year old men can shoot 90%+ with enough practice. Given how much post players get fouled, you'd think that'd be a major focus.

Meant as a reply to anon above.

I think the most likely explanation is that incentive to improve really isn't that great. If you do the following calculation, I don't think you'll find that getting better at free throws (for most players) will actually benefit them very much (and I'm sure someone on the internet has):
marginal point gain from improved free throw shooting --> marginal increase in win shares --> marginal increase in expected lifetime salary

Tyson Chandler is a notable exception, who vastly improved his usefulness on offense by getting better at free throw shooting - because that let him take advantage of an existing (I think) ability to roll to the rim, catch and dunk.

Getting better at free throws is probably not as easy for large humans. Some hypotheses why:

1) really big hands make it harder to shoot the ball
2) the free throw distance is too short for big guys - they have to be too gentle
3) bigs have to shoot with a higher arc, because the ball starts higher (it feels like Dirk shoots with a *really* high arc) - maybe they're not actually taught that?
4) big guys simply have fewer lifetime reps shooting, or are less comfortable doing it - they've been able to get by on their size

One last thing: being really tall means the talent/skill/work-ethic filter to the NBA is wayyy coarser than for other sports.

The 3 on 2 break used to be a thing of beauty. the 3 point shot is ruining the game. Too big of a change. We pay to see Russel Westbrook dunk not Stephan Curry toss up long shots.

Tyler writes, "I am a big fan of Stephen Curry, but in fact his family tale is ultimately a sobering one." Can someone help me here? What does Tyler mean or what makes Steph Curry's family tale sobering?

"it may take generations to properly value an innovation"

Is this really what Mr. Cowen was trying to say? If so, it seems willfully obscure. What's next, lots of math?

there may be a straussian reading I'm missing.

Stephen Curry one of the all time great NBA players of all time. Although he is of great nobility, this does this mean that he matches up to the standard of Michael Jordan and Kobe Bryant. Now, Curry at 80 years of age is Most valuable player of the NBA game that ever lived. Astounding achievement this is during the 1970s where there were no technology to compensate for challenges. When the three-pointer was instituted, the NBA was a much more up to speed and fast-breaking game than it is now. It wasn’t until the pace slowed down that teams were able to see how crucial the three was in a half court offense. Michael Jordan achieved much more, but consider the fact that he is still one of the all time greats, he deserves great credibility.

Relevant SMBC comic about this:

The obvious explanation is that because the NBA is a much bigger concern today than in 1979, its talent pool is much larger, and so there are better shooters now. That's why putting 5 guys on the floor who can shoot the three has become almost commonplace, and players like Curry routine try threes from well behind the line.

There's plenty of knowledge that are simple and cheap but still not widely implemented. Look at the trans fatty acids in foods. We know it kills people but we still need 3 more years to ban it in the U.S. I can think of many more examples that still haven't been widely adapted.

The premise that there are no more low hanging fruits, or even the definition of low hanging fruit, should be reexamined.

There's something about this I disagree with. This is true - "The NBA had to change its rules" - but this isn't - "to encourage more three-pointers". The NBA changed it's rules just like most sports did - possibly due to concussion research, or maybe something else. More 3's is a result of the rule changes, not a reason for them. I still don't think the warriors won because of the 3 though. They've been awesome at 3's for several years now, but this year they had the #1 defense. AND they did not play a healthy starting point guard the whole playoffs. In other words, curry broke this record against almost exclusively backup point guards. I don't think there's been some resistance to change in the league.

Sorry if any of this repeat the above discussions.

(A) Thinking about the slow uptake in shooting threes in terms of organizational frictions:

Why didn't players themselves go further down the road to exploring three point shots earlier in the trend given the "incentives of money, fame, and sex to get this right"? Obviously, frictions others have listed from head coaches and front offices are up there, but--maybe I missed someone mentioning this--don't overlook the actual teammates as important frictions.

A coach's incentives for the team's performance clearly point to quicker uptake on 3 pointers, since a coach's fate almost always rises and falls with the team. Yet, the incentives of individual players can be related to team success in complicated ways. More threes come at the expense of the biggest players, those with a specialty in posting up, and ball dominant guards. The last group loses shot attempts and therefore points since three point shots overall and especially the best ones are assisted. The assist does not offset the lost points as assists are not valued as highly as points in the free agent market or on the dating scene. Examples of many and costly disputes over fewer shots are easy to find. Also, at the margin, additional threes typically come from secondary players who are spotting up waiting for a pass from ball dominant stars. These players don't have the power to agitate for more shots.

(B) Lack of evidence

Another crucial factor to the slow uptake is the over reliance on 1 stat -- field goal percentage -- and its replacements. Comparing 3pt% vs 2pt% will lead to the *obvious* conclusion that 3 pt shots are worse since the %age is lower. I mean, duh, come on. Two stats very quickly ended this argument. Firstly, expected value per shot does not equal the percentages of makes, and point per possession stats changed the conversation quickly. Secondly, %age stats by shot location slightly beyond "2 or 3?" (not even the advanced work by folks like Kirk Goldsberry) were extremely informative. The gap in % for midrange shots vs 3 pt shots is much smaller than the gap between 3 pt shots and all 2 pt shots since you don't contaminate have to compare 3s to layups. Additionally, people realized corner threes in particular were high value shots; they are literally closer to the rim, and less contested conditional on a shot attempt. These realizations ("follow on innovations"/"commercialization innovations") helped push through the frictions many have listed in the delayed uptake.

(C) Is this even a slow uptake?

One way to recast your argument about learning and discovering strategies related to 3s is as a story of evolution. If you take a team's strategy as dictated by coaches, then the uptake is driven by strategy changes at a coach level, which are typically relatively stagnant within tenures. Say coaches have 4 year tenures on average. Then we're only on generation 9!*** It's not clear that getting to the current levels at the 9th iteration is "slow." Relatedly - How many generations does it take a neural network to beat Super Mario, where the inputs, feedback, and objective function are unambiguous?

*** I would add that the "optimal level" of threes a coach could choose was lower than it is currently earlier on given the available level of 3 pt shooting talent. Lower optimal levels means that the feedback about the correct change (increase 3s by X) is lower if you take 200 threes in 1980 than if you take 200 threes in 2010. This would slow the jumps forward between iterations.

I missed an obvious tie in: "I would like to make three points: (A) (B) and (C)"

Comments for this post are closed