TGT: Four Maxims of Tournament Design

In Tourneygeek’s Guide to Tournaments I set out a proposed table of contents for a monograph I may (or may not) write some day. I’d like to present the important things I think tourneygeek teaches in a more straightforward manner.

In this post, I’ll share my first attempt to write a section of this book. It’s about the maxims of tournament design. Regular readers may recall that so far I’ve discussed only three maxims – the fourth is new. Scroll down to see what the fourth one is.

Maxims of tournament design

Much of the advice in this book applies to very specific situations. But there are a few principles of tournament design that are so generally applicable that they can be thought of as maxims. In this section, I’ll introduce four of these.

Bear in mind, however, that there will be good reasons, sometimes, for violating even these four maxims. Whether a particular reason is good enough will be discussed in various places, and in other chapters there will frequently be recommendations that tend to derogate one or more of these maxims. But they will provide guidance in many situations.

Maxim 1: Maintain Balance

Unbalanced designs are likely to be inequitable. Avoid them unless there is some good reason that a design needs to be unbalanced.

Balance can be a problematical concept, so it’s not clear in all cases when this rule should be applied. Sometimes, a move that creates an imbalance in one respect also tends to restore balance in another. In most circumstances, however, a good tournament designer has an intuitive sense of balance. She feels a little uncomfortable whenever imbalance appears. One of the objectives of this book is to help you cultivate such a sense.

A number of salutary rules are, in essence, specific applications of this maxim. Here are a couple of examples:

     All of the drops from one round in a bracket should go to a single round in the other bracket.
Unearned byes should be spread as evenly as possible in the first round of a bracket.

Maxim 2: Do Not Reward Tanking

In a well-designed tournament, there should almost never be a situation in which a team or player has an incentive to lose a match rather than win it.

This maxim becomes harder to observe as a tournament’s incentives become more complex. It’s fairly simple to observe where there’s a single goal in play. A plan for distributing more prize money to the loser of a match that to the winner, for example, will stand out as absurd.

But where there’s more than one set of incentives in play, it may be difficult or impossible to keep them all pulling in the same direction. For example, in chess tournaments there is often a prize money payout for the top player in ratings class “C”. Some players who are really better than class C status would suggest have an incentive to lose enough games to keep them in class C so at to be eligible for the separate prize money.

The most pernicious violations of maxim 2 happen because there are incentives that are outside the formal control of the organizers. Gambling tends to create these. When large sums of money depend on the outcome of play, there can easily be situations in which a gambler can profit by bribing a player to throw a match.

Maxim 3: Keep incentives balanced

Even if no player has an outright incentive to lose a match, a tourney can be distorted when the incentives for victory in a match are one-sided. If one player is indifferent to winning or losing when the other is highly motivated to win, bad things are likely to happen.

This is a particular problem in the late rounds of round-robin tournaments. The round robin is, in some ways, the fairest tournament design, as it gives each player the very same set of opponents. But in the late rounds, it often happens that one player’s future is already set while the other’s is still in doubt. This happens frequently in sumo wrestling, for example, where it’s been shown statistically that the player who badly needs a win will me much more likely to get one when playing a player for whom one more win is not valuable.

The worst violations of maxim 3 can lead to collusion, but they can also cause trouble even if there’s no bad behavior on anyone’s part. The indifferent player or team may not actually throw a match, but it’s sensible for them to take an opportunity to rest their best players or play conservatively to avoid injury.

Maxim 4: Decide in Advance (and Appoint a Tournament Committee)

When a tournament organizer is presented with an unanticipated problem, there’s often more than one plausible way to proceed. But no matter how careful and disinterested the decision, it will always bear a taint of suspicion if the question is not resolved before the director knows which specific players or teams are involved.

In a larger context, this is known as the rule of law. An autocratic ruler in a dictatorship will usually be able to offer seemingly neutral reasons that justify decisions that just happen to aid his friends or hurt his enemies.

The metaphor of dictatorship is distressingly common in complaints about tournament directors. If you wait to decide what the rules are until after you know who they’re going to apply to, the decision will always smell fishy.

To the extent that you can, you should settle in advance all the rules that you might need to apply. For Example, don’t wait until there’s an unlikely tie before announcing the rule that will be used to break ties. But in tournaments, as in life, sometimes there are decisions to be made that even a very careful tournament director does not see coming.

In the context of governing countries, this is the role of an independent judiciary. For tournaments, this is often given over to a tournament committee—a small group of players who are well regarded by the playing community. In a well-run tournament, there will usually be no work at all for the committee. But you want to have one, just in case.