Every NBA roster contains players with nearly identical physical gifts and wildly different careers. One becomes an MVP. One becomes a rotation piece. One is out of the league in four years. The difference is almost never the thing that got them drafted, and it is rarely effort, because at this level everyone works. What actually separates a leap from a plateau is a combination of biology, opportunity, organizational competence, and one specific measurable skill that predicts shooting growth better than shooting itself does. Here is what the evidence says about how NBA players improve, when they stop, and why some franchises produce so many more leaps than others.
The Curve Everyone Is Fighting
Career trajectory research splits an NBA career into three phases: development from roughly 19 to 24, prime from 25 to 29, and decline from 30 onward.
Read that carefully. The development window is about five years long and it closes at an age when most people in other professions are still considered junior. Everything a franchise does with a young player is a race against a deadline, and the deadline arrives at 24.
This is the hidden logic behind an enormous amount of NBA behavior that otherwise looks irrational. It is why teams give minutes to a struggling 21-year-old over a competent 29-year-old on a team going nowhere. It is why a fourth-year breakout is treated as a revelation and a ninth-year breakout is treated as a fluke. It is why front offices pay steep prices for young players who have not yet been good. They are not buying production. They are buying remaining development window.
The curve is an average, not a law. Late bloomers exist, and elite shooters in particular tend to show a delayed peak, arriving later and holding their level longer than the general pattern predicts. But betting against the curve is betting against most of the evidence.
What a Leap Actually Looks Like
The word "leap" suggests something sudden. In the data it almost never is.
Shai Gilgeous-Alexander averaged 10.8 points, 3.3 assists, and 2.8 rebounds across 82 games as a Clippers rookie in 2018-19. He was a useful young starter and nobody's idea of a franchise player. He would go on to win MVP averaging 32.7 points a game. That is not a leap; it is six or seven consecutive years of compounding, each one adding a few points of scoring and a bit more efficiency, until the accumulated distance became enormous.
Jalen Brunson is the same story in a lower key. Dallas took him 33rd overall in 2018, in the second round. He averaged 9.3 points and 3.2 assists as a rookie on 34.8% three-point shooting, then 8.2 points the following year, which is to say he got slightly worse. In 2020-21 he jumped to 12.8 points on 53.4% from the field and 42% from three. Then he left for New York, became an All-NBA guard, and was Finals MVP of a championship team in 2026.
The pattern in both cases is that the third or fourth year is where the shape of the career becomes visible, and that the two seasons before it looked unremarkable or worse. Franchises that give up on players after two years are giving up before the evidence arrives.
The Skill That Predicts the Leap
If you want one number that forecasts whether a young player's shooting will improve, it is not his three-point percentage. It is his free throw percentage.
Research on shooting development has found college free throw shooting to be predictive of three-point improvement in the NBA, and the reason is that free throws isolate mechanics. Same distance, no defender, no movement, no clock. A player who shoots 85% from the line has a repeatable stroke that can be extended behind the arc with reps. A player at 60% has a mechanical flaw, and distance makes mechanical flaws worse rather than better.
Three-point percentage on low volume, by contrast, is mostly noise. A rookie who takes 40 threes and makes 15 has told you almost nothing. His 180 free throws have told you a lot.
The G League Is the Actual Development Engine
Practice does not develop players nearly as well as games do, and there are only so many NBA minutes. The G League exists to solve that arithmetic.
Pascal Siakam is the canonical case. Toronto took him 27th overall in 2016, a pick most mock drafts had going in the second round, and he arrived as a genuinely raw prospect with athletic tools and little polish. He was assigned to Raptors 905, led them to a G League championship, and won Finals MVP there in 2017. Two years later he won NBA Most Improved Player as Toronto won the championship.
That sequence is the model: real competitive minutes, in a real professional league, against real opposition, working on specific weaknesses under game pressure rather than in a gym. A prospect who plays 34 minutes a night in the G League gets more development reps in a month than he would get in a season at the end of an NBA bench.
Two-Way Contracts and the Middle Class
The mechanism that makes this work at scale is the two-way contract. Each NBA team may carry up to three, separate from the standard 15-man roster, and only players with fewer than four years of NBA service are eligible.
A two-way player moves between the NBA roster and the G League affiliate, earning considerably more than a standard G League salary and less than a full NBA contract. He can appear in up to 50 NBA regular-season games; beyond that, or to play in the postseason, the team must convert him to a standard contract, which typically comes with a raise to the league minimum.
Teams use the slots three ways: developing genuine prospects, holding situational depth for injuries, and scouting a player at close range before committing a roster spot. For second-round picks and undrafted players, the two-way is very often the only door into the league that is actually open.
Why Organization Matters More Than Talent
The reason certain franchises produce disproportionate numbers of leaps is unglamorous: their affiliate runs the same system they do.
Organizations known for development align the two levels deliberately. Shared terminology, mirrored offensive concepts, matching defensive coverages, coordinated strength programs. When a player is called up, he is not learning a new language under playoff pressure. He is running actions he already knows against better opponents.
Where the affiliate runs its own system with its own staff and its own priorities, a call-up starts from scratch every time, and the G League minutes convert into far less NBA value. Two franchises can invest identical resources in development and get completely different returns based on whether the two halves are speaking the same language.
Skills That Age Well, and Skills That Do Not
Development is not only about adding. It is about adding the right things, because skills decay at very different rates.
Explosiveness goes first. The burst that creates separation off the dribble is the most athleticism-dependent skill in the sport and the first to erode. Shooting, feel, passing, and positional defense hold up far longer.
This is why the most durable career move in basketball is the shift toward catch-and-shoot three-point shooting. Players who make that transition extend their productive careers by an estimated two to four seasons, because they are trading a skill that is disappearing for one that is not. The players who refuse the transition, who keep trying to be the same creator at 33 that they were at 27, tend to fall off abruptly.
A young player developing a reliable catch-and-shoot three is therefore doing two things at once: making himself useful now, and buying years at the end of his career he has not reached yet.
Why Most Players Never Develop a Jump Shot
Given that shooting is the most valuable and most durable skill, and given that every player knows it, the obvious question is why so few actually add one.
Part of the answer is mechanical, and it is the free throw signal again: a player whose stroke is fundamentally sound can refine it, and a player whose stroke is not is attempting a rebuild rather than a refinement. Rebuilding a shot means getting worse in public for a year or more, at exactly the age when a fringe roster player is fighting to stay employed.
Part of it is opportunity. Improving a shot requires taking it in games, and a coach fighting for his job is not enthusiastic about a 22-year-old taking developmental threes in the fourth quarter. The players who develop shooting are usually the ones on teams with nothing to lose or in the G League where losing is acceptable.
The anecdotes are real and the base rate is low. Most drafted players do not meaningfully improve their shooting, which is why teams have shifted toward drafting players who can already shoot rather than betting they will learn.
The Deadline Problem
There is a structural cruelty in how development and contracts interact.
A first-round pick signs a four-year rookie-scale deal, and his team has to decide whether to extend him near the end of it. That decision lands right around ages 22 to 23, which is to say near the middle of the development window rather than the end of it. Teams are forced to price a player before they know what he is.
Look again at the two careers above. Brunson averaged 8.2 points in his second season. Gilgeous-Alexander was a solid young starter with no MVP case anywhere in sight through year two. Both were still years away from the versions that would define them, and both would have looked like reasonable players to let go at the moment a team had to commit.
That is the trap. Extend early and you may be paying market rate for a player who plateaus. Wait, and a leap in the final rookie year means you now bid against everyone else at his new, much higher price. There is no option that lets a front office simply wait for certainty, because certainty arrives after the deadline does.
It also explains a pattern that looks like incompetence from the outside: good players changing teams for very little in their fourth or fifth year. Frequently the original team was not wrong about the talent. It was forced to choose before the evidence existed, chose conservatively, and watched the development happen somewhere else. Brunson's entire Knicks era began exactly that way.
The lesson for anyone evaluating young players, in a front office or from a couch, is that two mediocre seasons are not a verdict. They are usually just the part of the curve that happens before the interesting part.
What Development Actually Requires
Strip it down and four things have to line up.
Time inside the window. Roughly ages 19 to 24, and no amount of work extends it much.
Reps that count. Games, not drills, which usually means the G League for anyone not already in an NBA rotation.
A mechanically sound foundation to build on. Refinement works, reconstruction rarely does.
An organization that teaches the same thing at both levels. Otherwise the reps do not transfer.
Most players who fail to develop are not missing effort or talent. They are missing one of those four, usually the one they had no control over.
Related Reading
- How to Build a Basketball Player That Wins
- NBA Rookies Who Changed the Game Immediately
- The Best Undrafted Players in NBA History
- The Biggest Busts in NBA History
In Career Mode your player's development runs on the same principle: the reps you put in are what he becomes. Keep his form up with our daily NBA Bingo board.