Double Elimination vs Triple Elimination: How Many Lives Does Your Tournament Need?

Double elimination gives every team two lives. Everyone starts in the winners bracket; a first loss drops you into the losers bracket, and a second sends you home. The two bracket champions meet in a Grand Final, which may include a bracket reset if the losers-bracket finalist wins the first match — the undefeated finalist has to be beaten twice.

The structure costs 2N − 1 matches for N teams, or one more with a reset. It is the standard for esports majors and most serious amateur competition precisely because it absorbs one bad result without doubling the length of the event.

The longest individual path is about 2·log₂N + 1 matches — seven at eight teams, nine at sixteen. That number, more than the total, is what determines how long a competitor's day is. Learn more about double elimination.

Triple elimination adds a third bracket. Teams start in the Winners Bracket on zero losses, drop to Second Chance on one, drop again to Last Chance on two, and are eliminated on the third. Three champions — one from each bracket — meet in a Grand Final of three to five matches.

The total is 3(N − 1) matches plus the champion's own losses: 21 to 23 at eight teams, 45 to 47 at sixteen. Second Chance is laid out exactly like a double elimination losers bracket; Last Chance is the genuinely new part, absorbing Second Chance losers cohort by cohort in a strictly serial ladder.

The longest individual path is roughly 3·log₂N + 3 — twelve matches at eight teams, eighteen at sixteen. That growth, not the total match count, is what puts a ceiling on the format. Learn more about triple elimination.

Choose double elimination as the default. It absorbs the single upset that makes pure knockouts feel unjust, and it does so at a cost most events can actually absorb: 14 to 15 matches at eight teams, 30 to 31 at sixteen.

It is clearly right above sixteen teams, where triple elimination is not offered at all, and it remains the better trade at sixteen: 30 matches with a nine-match worst case against 45 with an eighteen-match worst case.

It is also the right call when your matches are long. If a single match runs 30 minutes, the difference between a seven-match and a twelve-match path is two and a half hours for one competitor.

Choose triple elimination on small fields where a single upset would genuinely distort the outcome and matches are short. At four teams the whole event is 9 to 11 matches — barely more than a double elimination — and no team can be knocked out by one bad game.

It suits best-of-one competition especially well. When each match is a single short game, results are noisy; a third life restores the reliability a longer series would have given you without lengthening every match.

At eight teams it is a genuine alternative to a round robin: 21 to 23 matches against 28, but with a real knockout final rather than a points table. Be honest with yourself about the day, though — seventeen rounds and a twelve-match worst case is a full programme.

Triple elimination is a niche format and pretending otherwise would not help you run a better event. It is justified on small brackets where the cost of a random early loss is high — four to eight teams, short matches, real stakes — and it becomes impractical quickly as the field grows.

The reason is in the closed forms. Double elimination's longest path is about 2·log₂N + 1; triple elimination's is about 3·log₂N + 3. The gap widens with every doubling of the field, and it is that gap, not the total number of matches in the bracket, that limits the format. If you are running sixteen teams or more, choose double elimination and spend the saved hours on better match quality instead.