Dynasties are built in the middle of the roster

Sustained contenders are separated less by their stars than by what they pay for their seventh through tenth players.

Everyone can name the two best players on a champion. Very few people can name the seventh, and the seventh is usually the reason the run lasted more than one season.

The structural problem with building around stars is that star salaries are broadly fixed at the top of the scale. If a maximum contract is a maximum, then every team that has a top-five player is paying roughly the same for him, and the difference between rosters comes entirely from the surplus generated elsewhere. Surplus is a wage below production. It comes from rookie deals, from players recovering from injury, from specialists whose skills are undervalued because they do not fill a box score.

So the interesting competitive question is not who has the best player. It is who has the best twenty million dollars of role players. That is a scouting and development question rather than a recruiting one, and it favours organisations with patience over organisations with money.

I would push this further than most. I think a well-run team should be reluctant to pay a third star, because the third star consumes the flexibility that keeps the depth alive. Two stars and six correctly priced contributors beats three stars and a roster of minimum-salary bodies over eighty-two games, mostly because injuries are not optional. Somebody will miss twenty games. The team whose ninth man is playable survives it.

The counterargument is real. Playoff basketball shortens rotations, and depth that mattered in February can be irrelevant in May, when you play seven men and the ninth never leaves the bench. Fair. But you have to reach May with your stars intact, and the way you do that is by having someone else to play in the games that do not matter.

There is a cultural component too, and I say that knowing how vague it sounds. A roster where the eighth man believes he might play is a different working environment from one where he does not. Practices are more competitive. Development is faster. That is not measurable and I am not going to pretend it is, but every long run I can think of had it.

The teams that fall apart quickest are usually the ones that got the top of the roster right and then stopped paying attention.