Best of 5 isn't objectively better

by AJ "Tyron" Martinez @ • February 7 2020

A summary of arguments I hear for Bo5 over Bo3: “reduces randomness, rewards gimmicks less, gives players more time to warm up and adjust to each other.” TL;DR “it makes it more likely that the better player wins.” Most of the counter-arguments are practical stuff about the amount of time a bracket takes.

I think “I like it more” is a sufficient argument for Bo5. I also think attempting to defend it as objectively better should probably set off alarm bells in your head.

Tournament rulesets are arbitrary, even in games that were actively designed with a competitive focus. Like, the game assigns winners,1 and sticking close to that leaves your ruleset relatively “pure”.2 There’s a lot of choices about game-breaking bugs that are similarly straightforward. Nobody really complains about mitigating softlocks, console crashes, or advantages that aren’t available to both players (port priority, P1-specific combos3).

But as soon as you decide “yeah, we’re gonna play multiple games in a prescribed format” you have officially stepped beyond the boundaries of Designed Gameplay. That’s not a part of the game; the rules, from here on out, are your responsibility.

Every parameter of a tournament ruleset is important, and none of them are really objective. Longer sets reward players with better ability to “solve” their opponents and adjust to a player’s style; shorter sets reward psychological guerilla warfare, stable nerves, and the ability to play solid with very little information. Long events4 test mental stamina and the ability to remain focused in a loud-ass convention hall while people make chimpanzee noises watching Guilty Gear pools; shorter events, or events broken into multiple days, reward the ability to warm up and get focused fast. Hell, something as simple as the position of your opponent (next to you, looking at the same monitor, or on the opposite side of the table?) affects the effectiveness of fake mashing and controller watching.

Change the format and you’re effectively spinning dials, nerfing certain players and buffing others; if this seems ridiculous, imagine an event split over 3 weeks where every set is played Bo30, and tell me that everyone would place the same as a two-day event run Bo5.5 Without even touching the “big” decisions, like character bans or controller restrictions, you’re creating side effects—side effects that might look trivial from a high level, but will definitely not feel trivial to the person who gets reverse 3-0’d6 just before surviving their pool.

what the fuck are you even talking about

This isn’t going to segue into some brain-max shit where I argue for Bo1 or Galactic Hyper-Swiss in every event.7 It isn’t even really shouting at people who want Bo5, even though it’s kind of disappointing to see people complaining about that as if $THEIR_FAVORITE_GAME is somehow being run in a fraudulent or invalid way. It’s mostly just…look, tournaments are fuckin arbitrary.

Even double elimination itself, even though it’s usually the most practical way of running events, still gives players brackets of wildly varying difficulty and requires potentially contentious decisions re: seeding. Try to dodge that by running large events round-robin, and you get the aforementioned three-week tournament—which, I’d like to add, would be sick as hell if it took place on a cruise ship.

The purpose of the tournament is not to determine who is the Platonic “Objectively Best Player”, because nobody can agree on what that means, and it’s impossible to construct an event that would test it anyway. The purpose of the tournament is to determine who is the best at winning the tournament. If that means DPing on every throw tech, going high 9 times in a row, and scumfucking people in Bo3 before they even know what’s happening, that’s what it means. Have fun with it.

  1. This is the cool thing about video games, you don’t need the WORD OF GOD (in the form of referees or officials) to negotiate basic interactions between game mechanics. Take the NFL’s rules on pass completion, which are a nightmare hellfuck that no reasonable human can be expected to enforce consistently within the split-second windows required to keep play moving. Imagine a world where NFL referees have to determine whether or not you did, in fact, tech that throw. ↩︎

  2. All hilarious Bowsercide and/or Ganoncide rulings aside. These are some of my favorite rulings because 50% of discussion about them is “the game behavior often isn’t consistent, is it worth the relative ‘impurity’ of imposing an out-of-game rule to make sure that the winner of a match isn’t decided by a random boundary fluke?”, and 50% “WHY WOULD YOU NOT BUFF BOWSER, HE’S FUCKING TRASH” ↩︎

  3. I have played at least one game where charge moves are only available to P1. Have fun with that ruleset. I think you just have to run it like chess at that point, where players play both P1 and P2 in equal measure. ↩︎

  4. length of bracket * length of sets * length of games ↩︎

  5. You don’t even have to step into concept space for this one, honestly. Punk had an electric run through EVO 2017’s SFV bracket, but after a bracket reset versus Tokido, he completely crumbles—even though he 2-0’d Tokido pretty convincingly in top 32! A difference in mental stamina and nerves turns the rematch completely on its head. This stuff happens, and the structure of a tournament is a significant factor in when and how badly. ↩︎

  6. Winning the first two games, then losing the next three in a row. F ↩︎

  7. I will argue for 1-stock Bo9 Brawl every time, though. ↩︎