NBA All-Star Game Selection, Format and History: How Rosters Form and Games Change

Published 22 September 2026Features

This layered structure governs a weekend that looks little like its origins. Over decades, the league has repeatedly altered how players reach the roster, how those players are divided into teams, and how the games are played.

Fifty Years of Shifting Roster Ballots

The current voting formula represents the latest compromise in a long-running argument over who deserves authority over the roster. During the first 24 editions of the event, fans had no direct say in the outcome. Sportswriters and broadcasters selected each conference’s five starters along with three bench players, leaving NBA coaches to fill out the remaining spots on the bench.

Team composition carried rigid constraints during that early era. Until 1973, the league mandated that every single NBA franchise had to have at least one representative in the game, while placing a strict cap of three players from any single club. Total roster sizes expanded as the league developed. In 1961, the NBA enlarged each conference squad from 10 players to 12.

The balance of power shifted before the 1974-75 season, when the league first granted fans the right to elect the starting lineups. For four decades, fan balloting stood as the sole mechanism for choosing starters, sometimes elevating popular veterans over players with stronger regular-season production.

The 2016-17 revision curtailed that unilateral fan control. By introducing active player ballots at 25 percent and media ballots at 25 percent, the league added peer assessment and professional analysis to counter pure fan sentiment, while preserving the public's 50 percent weighting.

Positional rules remain strictly defined within this framework. Voters in each conference select two backcourt guards and three frontcourt players. Once those 10 starters are locked in across both conferences, head coaches select the 14 reserves, adding seven players from the Eastern Conference and seven from the Western Conference.

``` ALL-STAR SELECTION BREAKDOWN (SINCE 2016-17)

• Starters: 2 guards, 3 frontcourt players per conference
• Fan Vote Weight: 50%
• Player Vote Weight: 25%
• Media Vote Weight: 25%
• Reserves: 7 players per conference, chosen by NBA head coaches
• Total Roster: 24 active players ```

Dismantling the Conference Matchup

For most of league history, roster selection doubled as team assignment. The Eastern Conference played the Western Conference, and conference affiliation determined who shared the court.

That geographic separation ended in 2018, when the league abandoned the traditional East-versus-West game design. Under the revamped 2018 system, the entire 24-player pool was selected through the existing conference voting system, but the game itself became a draft. Two captains were named based on fan voting: the starter who received the highest overall fan vote total in the East, and the starter who received the highest overall fan vote total in the West.

The draft took place in two distinct phases:

1. Captains drafted from the eight remaining starters, selecting players regardless of conference affiliation or position.
2. Captains drafted from the pool of 14 coach-selected reserves, again ignoring conference lines.

A Western Conference captain could select an Eastern Conference guard with the first overall pick, pairing rivals on the same court. The draft system showed that while conference standing still determined who qualified as an All-Star, the game itself had become an exhibition.

Target Scores and Round Robins

The league did not stop with player drafts. In recent seasons, the format of the game itself has shifted toward shorter, tournament-style play.

The 2025 All-Star Game abandoned the standard four-quarter structure entirely. Instead, the league staged a four-team, three-game mini-tournament. Rather than playing against a traditional countdown clock, every game in the 2025 tournament was decided by a target score: the first team to reach or exceed 40 points won the contest.

The redesign continues to shift. According to the NBA’s 2026 All-Star documentation, the event is set to adopt a U.S. versus World round-robin tournament structure featuring four distinct 12-minute games.

These changes affect the tactical nature of the event. A 12-minute mini-game or a sprint to 40 points creates completely different substitution rotations, defensive incentives, and possession counts than a standard 48-minute professional fixture.

Exhibition Numbers and Historical Record Limits

Because the event operates under shifting tournament rules and draft pools, its statistical output stands apart from the rest of the NBA calendar.

The league produces full box scores for every All-Star Game, detailing minutes played, field goal percentages, assists, rebounds, and scoring totals. However, these figures are generated under competitive conditions that share little with official league play. The games feature no regular-season standings implications, no playoff seeding consequences, and, in recent years, target scores or shortened periods rather than regulation 12-minute quarters.

While the league publishes comprehensive game logs for the weekend, official league publications and statistical databases track All-Star performances separately from career regular-season and playoff totals. A 40-point performance in an All-Star mini-tournament is not included in a player's official regular-season scoring average or career scoring total.

The event serves as an individual accolade. Selection to the 24-player pool counts on a player's career resume, but the actual box score from Sunday night functions strictly as exhibition data.

With the 2026 format moving to a four-game U.S. versus World round-robin, the NBA All-Star Game continues to distance itself from traditional conference play, testing whether international tournament rules can generate the competitive pressure that standard exhibition games lost long ago.