Opinion: Experimenting with “Double Elim, Best Two of Three”

By on September 23, 2013 at 1:24 pm

“Double elimination, best two of three (or three out of five, for Ultimate Marvel vs. Capcom 3).”

That’s the standard tournament format for pretty much any mainstream tournament event in North America. As a community, we’ve come to accept that as the competitive standard and practice for those conditions. But while I was watching Daigo Umehara’s drastic range in results this weekend — contrasting his 10-0 and 10-2 victories over Evo 2013 champ Xian and Evo 2012 champ Infiltration in recent showmatches to his 7th place Evo 2013 finish and his out-in-four-straight-rounds set at the Tokyo Game Show Capcom Cup Qualifier tournament this past weekend — I am reminded that there’s more to explore in fighting games than just “double elim, best two of three.”


Borrowing Bans from MOBAs

The fighting game community has historically thought of banning a character as the ultimate in scrubbiness; if you decide to ban a character, you’re basically saying “Developers, we know better than you; this character is unfairly overpowered in a way that makes the game not fun, and we don’t think it’s worthwhile — or possible — to learn how to beat him.”

That’s why I was very interested to see Complexity Gaming’s Long “ShadyK” Tran’s recent post on the FGTV house blog about a suggested rules permutation for their upcoming “When’s Marvel?” weekly tournament. This week’s rule would allow each player in any given match to ban one character from the opponent’s roster before character select, and it’s inspired by the picks/bans mechanic in League of Legends, Dota 2, and other MOBAs. In these games, both teams take turns banning characters so that neither team can choose them (which is notably different from Tran’s suggested rule, mind you). Banning a champion isn’t scrubbiness, it’s an essential part of maximizing your chances of victory, and it adds a fascinating meta-game layer to your team composition strategy.

League of Legends, for example, has over 100 playable characters. Like fighting games, you’ll typically see a much smaller set of characters in regular rotation for competitive play (though, like fighting games, there’s nothing to stop a team from working an uncommon character into a team to surprise their opponents), but still, the range of potential teams is massive, especially when you factor in the difficulty of preparing for different permutations of five-man teams using that set of characters. Banning characters gives you a chance to cherry-pick your matchups somewhat, but give your opponents that same opportunity, meaning both teams need to be a little bit more versatile, with each player comfortable competing with a wider range of characters than just the one they may favor the most. Success goes to the team with the best knowledge of characters, matchups, and team composition strategy, not necessarily the most dangerous single composition.

As far as I can tell (I am, admittedly, a relative newcomer to MOBAs in general) this has a few effects; it means that you see a wider range of character played on the competitive stage (which makes it more entertaining to watch, certainly), and it forces players to cultivate both a broad base of character/matchup knowledge and a very solid base of execution fundamentals, since no one can rely on their favorite character making it past the ban phase. (Heck, your team might even decide to ban your favorite character if you’re going up against a team that plays that character even better than you do.)

It also gives players an instant Put Your Money Where Your Mouth Is card with regards to game balance issues. If you think one character is ridiculously overpowered and don’t want to play against him, you can ban him — and you may find over the duration of a competitive season that people who play that character get so good learning to play other characters that you want to ban them instead, meaning you’ll have to learn that original banned matchup after all.

Bans in UMVC3

Considering it doesn’t look like we’ll be getting any major patches for UMVC3 any time soon, adding a bans phase could be an interesting way to mix things up and give each player their own game balance tool — with the caveat that their opponent has one too, and they’ll have to learn how to play permutations of their team that don’t have their preferred point, second, or anchor characters. It would also force us to identify exactly which characters we depend on (and think others depend on for their teams) the most. Ban Chris G’s Morrigan, and you won’t have to deal with the bullet hell, but if you think Doom’s Hidden Missiles are what really makes her scary, you can ban Doom instead and save yourself from losing to TAC infinites. (Or maybe he’ll just pick his Ryu team.)

As a Zero player, I’m sure I’d immediately feel the sting of a bans-phase tournament. (What am I fighting for?) But I run Zero/Doom/Vergil, so banning Zero is probably the easiest pick for me to adjust to because anyone fits into the Doom/Vergil shell. Ban Doom, and you’re denying me a great assist and TAC infinite options; ban Vergil, and I’d have to find a new anchor and an assist I can use to help Zero with midscreen conversions. Basically, in order to be competitive at UMVC3, I’d have to be comfortable with combos, assists, and matchups for at least three different teams instead of just one. In my eyes, the player who can win with three teams is better than a player who can only win with one.

Who’s the Best?

We love tournaments because they tell us who the best player in the room is — whether than room is my living room, or the ballroom at the LVH for Evo 2014. And we’ve determined that “double elim, best of three” is the best available tournament format to determine the best player; double elimination allows players a chance to make mistakes, best of three allows players to counter-pick difficult matchups and learn more about their opponent, and both of those are doable without needing a whole month to run through the bracket.

But the way we structure our tournaments reflects the kinds of skills we value the most when we think of what it means to be “the best”. When we want to see players excel with their chosen character and show you how profoundly they understand their matchups on the big stage, let’s try running a single-character tournament like so many Japanese tournaments used to do, where each player has to lock in their chosen character at registration. When we want to see players display a broad understanding of character strengths and show off their alternate picks in competition, let’s run more character auction tournaments like the ones we see at UFGT. And when we want to see what a player like Daigo can do with several weeks of preparation time, let’s put on more showmatches. We can’t change the game itself, but we can change how we play and practice it — and it doesn’t just have to be “double elim, best of three.”