That's not obvious to me. If you play single elimination you can make each match roughly twice as long for the same tournament length.
I made my statement under the assumption of keeping all other variables fixed.
I agree with you that race length and the number of eliminations can be offset to leave the level of skill more of less constant.
I think an interesting theoretical question is the following: what tournament format (defined by match length, number of eliminations, seeding, alt/winner break) is the most efficient? That is, for a given likelihood of having the best player win the tournament, what format would result in the lowest number of expected games played?
Last edited: