How NBA GMs Actually Build a Contender

There is no formula for building an NBA champion, which is inconvenient for everyone who writes about the NBA. What there are instead are four repeatable paths, each with a different risk profile, a different time horizon, and a different way of failing. One team hoards draft picks for half a decade. Another skips the wait entirely and buys a finished core. A third wins a lottery and builds around a generational prospect. A fourth simply sees a player nobody else sees. The last three champions were built three different ways. Here is what each path actually requires, using teams that just proved it works.

Path One: Accumulate Assets and Wait

On July 6, 2019, Oklahoma City traded Paul George to the Clippers. The return was Shai Gilgeous-Alexander, Danilo Gallinari, four unprotected first-round picks, one protected first-rounder, and two pick swaps. At the time it was widely read as a rebuilding team taking the best available haul for a star who wanted out.

It became the foundation of a championship. Sam Presti spent the following years converting every tradable veteran into more picks, eventually controlling a stockpile of future first-rounders large enough that the league's other 29 teams had to plan around it. Gilgeous-Alexander developed into an MVP, a scoring champion, and a Finals MVP in the same season, joining a very short list of players to do all three at once. The picks turned into Jalen Williams, into other rotation players, and into the flexibility to move up and down draft boards at will.

Oklahoma City won the 2025 title with a roster where nearly every important player was either acquired in that trade or drafted with a pick that came from a trade like it.

The trade kept paying out for years after everyone stopped watching it. Its final pieces did not convey until the 2026 draft, by which point the full return on Paul George read as Gilgeous-Alexander, Jalen Williams, Thomas Sorber, Brooks Barnhizer, Aday Mara, and a protected 2027 first from Sacramento. One veteran, traded once, produced a championship core and was still generating assets seven years later.

The path's requirement is patience of a kind most organizations cannot survive. It means several genuinely bad seasons, an owner who will not fire the executive running them, and a fan base willing to watch. The failure mode is straightforward: accumulate the picks, then miss on them.

Path Two: Buy a Finished Core

New York won the 2026 championship without developing a single member of its starting five.

Jalen Brunson arrived as an unrestricted free agent from Dallas in 2022 on a four-year, $104 million contract that looked generous at the time and became the best value in the sport. Mikal Bridges came by trade in the 2024 offseason. Karl-Anthony Towns arrived weeks later in a three-team deal that sent Julius Randle, Donte DiVincenzo, Keita Bates-Diop, and a protected first-round pick to Minnesota, with Charlotte facilitating. OG Anunoby and Josh Hart both came by trade as well.

Every starter was acquired. None was drafted and raised in-house. Leon Rose's front office assembled the roster the way you would assemble a fantasy team, and it ended a title drought that stretched back to 1973, closing out San Antonio in five games with Brunson taking Finals MVP.

The requirement here is capital, both financial and in assets, plus the willingness to spend it before you are certain. Every trade in that sequence cost real value. The failure mode is expensive and public: pay full price for several very good players, discover they do not fit, and have neither the picks nor the cap room to correct it.

The Discount That Made It Possible

New York's title has a specific enabling condition, and it is worth isolating because it is the rarest ingredient in team building.

In 2024 Brunson was eligible to sign a four-year, $156.5 million extension immediately, or wait a year and, if he earned another All-NBA selection, become eligible for a deal worth roughly $269 million. He signed immediately. He left approximately $113 million on the table so the franchise could afford the roster around him.

There is no cap mechanism that produces that. No exception, no clever structuring, no front-office skill. A player simply decided that winning was worth nine figures. Towns and others also restructured or accepted terms that helped. Championship rosters are frequently built on a star's willingness to be underpaid, and it is the one variable a general manager cannot manufacture.

Editorial illustration: How NBA GMs Actually Build a Contender

Path Three: Win the Lottery, Then Draft Well Around It

San Antonio took Victor Wembanyama first overall in 2023 and reached the Finals three seasons later.

What made it fast was not Wembanyama alone. The Spurs took Stephon Castle fourth overall in 2024, and he won Rookie of the Year. They took Dylan Harper second overall in 2025. They traded for De'Aaron Fox to give the young core a veteran lead guard. By 2025-26 they finished 62-20, second-best in the league, beat the defending champion Thunder 111-103 in Game 7 of the Western Conference finals, and fielded the second-youngest roster ever to reach the Finals.

Wembanyama was Western Conference finals MVP, averaging 27.3 points, 10.9 rebounds, 3.1 assists, 1.4 steals, and 2.7 blocks across those seven games.

The requirement is luck you cannot schedule, followed by competence you can. Plenty of teams have won a lottery and wasted it. What San Antonio did was treat the generational prospect as the beginning of the work rather than the end of it, hitting on premium picks in consecutive years while the cheap years of Wembanyama's rookie contract were still running.

Path Four: See What Nobody Else Sees

In 2014 Denver drafted Nikola Jokic with the 41st overall pick. The selection was announced during a Taco Bell commercial, because a second-round pick was not worth interrupting the ad break for.

Jokic won MVP in 2021, 2022, and 2024, and was the unanimous Finals MVP in 2023 when Denver won its first championship, averaging 30.2 points, 14.0 rebounds, and 7.2 assists in that series.

This is the cheapest path and the least reproducible. A superstar found in the second round costs nothing, occupies almost no cap space for years, and hands a front office the single greatest asset in the sport: a max-level player at a minimum-level price. It is also the path nobody can plan. Every team scouts the second round. Almost every team is wrong about it.

What the Four Paths Share

Strip away the differences and the same idea sits underneath all four: acquire production that costs less than it is worth, and hold it for as long as the rules allow.

Oklahoma City did it by collecting rookie-scale contracts in bulk. San Antonio did it with premium picks in consecutive drafts. Denver did it by finding an MVP in the 41st slot. New York did it by paying market price for good players and then having its best player voluntarily accept below-market pay.

Every one is the same trade in different clothing. The gap between a player's value and his cap number is the only durable competitive advantage in a league with a salary cap, because talent alone can always be matched by someone willing to spend.

The Fifth Path, Which Does Not Work

There is a fifth thing teams do, and it is the only one with a perfect record of failure: staying in the middle.

A team wins 38 to 44 games, misses the playoffs or exits in the first round, and picks somewhere between 11th and 18th. It has no cap space, because its payroll is committed to good-not-great players on market-rate contracts. It has no premium picks, because it was not bad enough. It has no tradable stars, because it does not have any. Every asset it holds is fairly priced, which means none of them can be turned into anything better.

This is the treadmill, and the cap structure makes it worse than it looks. The mid-lottery is where the flattened odds are least generous relative to expectation, market-rate veterans are the least tradable contracts in the sport, and there is no mechanism anywhere in the CBA that rewards being adequate.

The uncomfortable implication is that the middle is a choice, not a fate. Every team on the treadmill got there by making a series of individually defensible decisions to stay competitive. The paths that work all require a period of deliberately not being competitive, or a very large amount of money, or a piece of luck. What none of them tolerates is drifting.

How to Read a Rebuild From the Outside

If you want to know whether a rebuilding team is actually building something, the win-loss record is close to useless. Four things tell you more.

Who owns its picks. A team with its own first-rounders plus somebody else's is accumulating. A team whose next three firsts are pledged elsewhere is paying off a previous regime, whatever the standings say.

How many cheap years it controls. Count the rookie-scale contracts with real players attached. That number is the width of the window.

Whether the veterans are tradable. Expiring contracts and reasonable salaries are ammunition. Long deals at market rate for 28-year-olds are not.

Where the payroll sits relative to the aprons. A team approaching the second apron without a genuine contender is in the worst position in the sport: expensive, capped, and legally unable to improve.

Read those four together and a rebuild's real timeline becomes visible, usually well before the results do.

The Constraint Every Path Now Faces

The 2023 collective bargaining agreement changed the endgame for all four routes. A team above the second apron cannot combine salaries in a trade, cannot use any mid-level exception, cannot send cash, and has a future first-round pick frozen and potentially pushed to the 30th selection.

The practical effect is that the assembled-core path has a shorter shelf life than it used to. A team that trades for three expensive stars will cross the second apron the moment those contracts mature, and at that point it loses nearly every tool it would need to fix a weakness. The asset-accumulation path survives longer, because rookie-scale contracts keep payroll down while the team is good.

This is why the smartest front offices now think in overlapping windows rather than single peaks. The goal is not to build one great roster. It is to have the next cheap star arriving before the last expensive one has to be paid.

The Job in One Sentence

A general manager's actual job is to keep as much value as possible under contract for as little as possible, for as long as the rules allow, and to know precisely when the window is closing.

Everything else, the draft strategy, the trade construction, the free-agent pitch, is a tactic in service of that. The four paths are just four ways of finding underpriced production. The teams that sustain success are the ones that never stop looking for the next one while enjoying the current one.

Closing illustration for How NBA GMs Actually Build a Contender