The World Cup teams of the US and Germany are in an odd position for their upcoming match on Thursday. If the game is a draw, then both teams advance to the round of 16, which means that the ideal strategy – as pointed out by Mark Heggen on Twitter – is for the two teams to agree to just sit on the ground and let the clock run out. Considering that the US coach Jürgen Klinsmann is from Germany and that West Germany infamously took part in the fixed Anschluss game, it’s hard not to wonder if there will be a little funny business during the match. At the very least, the match threatens to be a defensive snoozefest.
As a game designer, of course, this situation is appalling. Anytime a game’s rules rewards players for boring themselves (not to mention us, the viewers), the problem is with the game designer, not with the players. A similar match-throwing problem haunted the round-robin badminton tournament at the last Olympics, and game designer David Sirlin wrote a great article which placed the blame squarely on the shoulders of the tournament organizers. In the case of the US-Germany game, the problem is not quite so egregious – especially since the loser will have to play Belgium in the next round – but obviously the system could be improved.
I am a game designer, after all, so what would I do? The core of the problem is that whenever a game has more than two teams (and the World Cup is basically a game with 32 teams), diplomacy automatically becomes an important tool, regardless of the intentions of the game designer. Therefore, if winning is not required to advance, diplomacy can override the usual incentive for victory. The problem is that diplomacy is considered totally verboten (and a little unmanly) in competitive sports, even though it is considered totally normal in tabletop gaming. (As they love to play in Germany, of course!)
The solution is to focus the two teams on the one incentive left for them in the tournament – the seeding for the next round. The tournament’s rules should dictate that if two teams can advance with a draw, then the group’s other two teams are eliminated immediately. (The phenomenal card game Battle Line – by the German Reiner Knizia! – has a similar rule that allows capturing a flag early if a player can prove that no cards left in the deck could beat the ones on his or her own side.) However, instead of just cancelling the match between the US and Germany, which is problematic for many reasons, the tournament should let them play the game solely to determine who gets the top seed out of the group. Therefore, the two teams are rewarded for playing the game at full effort and have no incentive to make a mockery of the competition.
Now, context is important here as every rule always has a complexity cost. Adding this type of special casing to a video game or, especially, to a tabletop game might not be worth the extra overhead for the player to keep track of it. However, because the stakes for the World Cup are so high and because avoiding diplomacy is a design aesthetic for the tournament, this rule could be useful. Indeed, after the Anschluss game, FIFA changed the World Cup rules so that the final games of the round-robin phase would be played simultaneously to reduce the incentive for diplomacy. The situation the US and Germany face may not but common, but it’s also likely to happen again – all that is required is for a final game to be played between teams which both have one win and one draw. Better game design always beats the alternative.
I don’t know much about this particular situation, but in general it’s still problematic if the ONLY incentive to win is for seeding where the next round’s teams are exactly known. If the situation is “If I win, I will face team X, and if I lose, I will face team Y” then you just know too much. Even if team Y is ranked lower, you might have a better chance of beating them than team X (maybe it was an upset, or maybe you are just really good vs team Y and never lost to them before).
This whole shady business stems from the concept of breaking a tournament into two phases. First, a qualifying phase of some sort (such as swiss or round robin) where you take the top RECORDS and then cut to a single (or double) elimination tournament consisting of only those with the best record. There’s practically always going to be a problem with that cut. If you super cared about that, you wouldn’t cut from one format to the other. BUT…there are reasons you’d want to use such a format. In a much smaller tournament than the World Cup, you’d want to use swiss or round robin simply let all the participants play as much as possible so THEY have fun, yet you’d want to cup to a single elim portion so the finals are actually hype and exciting.
In a huge money thing like the World Cup, surely there need not be a qualifying portion that creates corruption when we cut to top X, but maybe they want it so there’s more matches to televise. Perhaps that outweighs the corruption problems of this format to them.
If you wanted to absolutely maximize corruption, I think you’d do round robin. It’s practically guaranteed (or actually guaranteed??) that a round robin tournament will have a lame-duck case where someone who cannot possibly win the tournament will play a match where their win or loss determines if someone ELSE advances. Super bad. It’s actually best if round robin tournaments are run with all the matches as simultaneously as possible with as much confusion surrounding it all so that it’s difficult for participants to figure out where exactly the lame duck situation is before they have to play. But anyway “round robin, cut to top X” is a huge red flag.
For my own tabletop games, I had to design tournament rules. I chose swiss, cut to top X so that we get the hype of a real finals and the fun of the early rounds. Does it create corruption? Yes, but I’ve also tried to minimize way more than in most other formats. The more draws, the more opportunity for corruption. When played in a tournament format, my games don’t have match draws. There is always some sudden death thing or whatever to force a winner. This means these weirdo collusion situations happen less because it’s less frequent that the records line up to make them possible.
Then there’s something else that makes the collusion even way less than that. There is no “intentional draw” either. A result of “neither player won” is counted the same as if they both lost. So there is no point in them agreeing to skip the match with a draw and there is also no point in them playing a fake match where they “somehow happen” to draw. In both cases, they’d just both lose.
Maybe all this somehow helped, I don’t know.
Thanks for the lengthy comment, Sirlin. I absolutely agree that round robin (or any format that has separate phases) will always be problematic, so it’s really a question of making something less bad than it was before.
One idea which I think could help would be to have – I’m not even sure what to call it – but a non-full single-elimination bracket. The NFL and MLB already do this. For American football, the two division winners with the best records get to skip the first round of play. For baseball, the two wild card teams have to play a single game to determine which one gets the last slot in the bracket. The latter system, in particular, works really with differentiating the value of a division title vs. a wild card slot. Both are valuable, of course, but the division title is SIGNIFICANTLY more valuable, so no team is going to coast until they have their division locked up.
Perhaps, for the World Cup, the round of 16 could instead become a round of 24 where the eight teams that won their group get to skip the first round of eliminations while the second and third teams have to play each other. Having to play an extra game would be a lot more valuable than a possibly valuable higher seed.