Game Design

The Pirates Who Don’t Do Anything Irrational

The salt air stung my eyes as I awoke, tied to the mast of an anachronistic sailing vessel. Five pirates, with about 7 eyes and 8 legs between them, stared back at me. “Please don’t kill me,” I begged them. “Haven’t I praised Libertalia enough?”

“Yarr,” the oldest one said in a perfect West Country accent. “We need your skills, matey. You see, we had a dilemma with a chest full of gold, and thought it would be interesting to turn it into a board game, but our initial playtesting didn’t turn out well and we threw it into Davy Jones’ locker out of frustration. Fix it, or walk the plank!”

“I’ll help you, then,” I said. “What was the dilemma?”

“Well, there be a chest with 100 pieces of eight, and all of us be rational pirates who operate by the pirate’s code…” My blood froze. It was going to take all of my wits to escape this situation.


Tabletop game design stands with one leg in the bucket of psychology and one leg in the bucket of math. Many big-name game designers, like Reiner Knizia, Richard Garfield, and, weirdly, John Nash, all have substantial experience in mathematical fields.

It then stands to reason that a lot of games, especially older games and European-style ones, are based off math problems and logic puzzles. Most notable is the traveling salesman problem, which sparked the venerable “Pick Up and Deliver” genre and encompasses games like Spiel des Jahres winner Elfenland.

However, not every logic puzzle is as easy to adapt to the tabletop environment as the traveling salesman problem. This article talks about the famous pirate game, why board game versions of it have run into issues, and how to fix them.

The “Pirate Game” Game

To break down this logic puzzle as a game, it roughly goes like this:

  • One player is the “chooser” each round – this role usually rotates.
  • A number of random rewards are drawn from a deck or similar; the chooser distributes these rewards among the other players.
  • The other players then vote on whether they like how the chooser distributed the rewards; a majority of “yes” votes confirms the treasure distribution, while a majority of “no” votes does various things including eliminating the chooser, electing a new chooser, and so on.

That’s the base, but of course any game is going to embellish it with various other mechanics to add complexity.

The highest-profile published game that uses this mechanic is Goodcritters, but I’ve seen numerous prototypes at playtest events that used this mechanic, and I’ve probably taken a crack at it once or twice myself. It’s a very appealing game on its face because it combines things that make for great gameplay – greed, group psychology, and shocking moments where the vote goes completely opposite how everyone expected. But in reality, the Pirate Game has a lot of issues that make it not that fun to play.

The Problem With The “Pirate Game” Game

  • Over 50 percent of gameplay is just one person choosing stuff with no input from other players. You know when you’re playing Apples to Apples and the judge takes forever? It’s incredibly frustrating, especially because you know that speaking up with your own opinion will actively prevent your answer from getting chosen. Conversely:
  • Being the chooser feels really bad. When you’re the chooser, you often feel like you don’t have any good options. Either you try to benefit yourself, and the other players instantly vote to disapprove your division, or you don’t benefit yourself whatsoever, don’t advance your victory plans, and even then there’s a chance the other players will vote no just because.
  • The distribution ends up being roughly the same every time. In order to ensure that nobody votes against the distribution, the chooser is often forced to divide everything up as evenly as possible. While some games can make this difficult or impossible, it doesn’t produce the variety of game states that make a game interesting to play repeatedly.
  • People aren’t rational. The 5 Pirates Problem assumes that all of the pirates involved in the transaction are perfectly rational and make decisions that result in the optimum benefit for them. However, no human is truly rational. Even when presented with the theoretical best option for them, people will often vote against the chooser for emotional reasons or even arbitrarily. Besides, even if people were rational, it wouldn’t be an interesting game as there would only be one obvious answer to each distribution round.
  • A lot of point-scoring mechanics get constrained. There aren’t a lot of interesting point-scoring mechanics that you can use when someone arbitrarily decides who gets what kind of loot. Set collection doesn’t really work because nobody would let people have a meaningful number of set items, negative effects will unanimously get voted against, etc.

The Solution to the Problem With The “Pirate Game” Game

This section has some proposed solutions to the problems outlined above. I don’t think that these are necessarily guaranteed solutions, but they’re ways of exploring the premise and producing a game that’s fun and still recognizably based on the logic puzzle.

  • Too much downtime: The two issues associated with this concept are “one player acts at a time” and “that player has too many choices, so it takes a while.” The first issue could be resolved with simultaneous choosing, or with voting happening one-by-one instead of all at once after everything has been distributed. For the second issue, the game New York Slice (a member of the adjacent I-split-you-choose family of games) neatly solves this by making the “treasure” a pizza that can’t be moved or rearranged – only split differently. This significantly collapses the potential decisions into a few that, while still tough, are much quicker to puzzle through.
  • Being the chooser feels bad: There has to be some way to reward the chooser that can be affected by a smart distribution, but that doesn’t feel “greedy” on the part of the chooser. Alternatively, some sort of hidden information element might make it easier for the chooser to feel like they can get away with something.
  • The distribution is roughly the same: The only way to get your players to take risky or uncomfortable actions is to make them. In the case of the pirate game, equal distribution is the “safe” tactic, so you have to disable the chooser from doing so, at least easily. “Parceling” treasures or otherwise making distribution less granular is a good start here.
  • People Aren’t Rational: No matter how much you try to force people to be rational, or to reward rationality, people just will not be rational. Instead, reward your players for identifying exactly how their competitors will be irrational, or at least make the consequences for irrationality funny and memorable.
  • Constrained Scoring Mechanics: If a scoring mechanic rewards players for meeting some condition, and the chooser has total control over how rewards are distributed, that condition will basically never be met. It’s better to either leave an element of hidden information or randomness, so these conditions can be better achieved unknowingly, or force them to happen through some other system. (For example, if the chooser has to impress one specific player, they may choose to do that by eg. grouping set-scoring rewards together.)

These are just surface-level thoughts on all of these issues, but you can clearly see that how, though flawed, the pirate game can be changed into something better oriented towards the flow of tabletop gaming.

Avast!

Math and strategy are irrevocably linked, so it stands to reason that classic math problems might be inspirational to game designers. Not every classic math problem can be transferred 1:1 into a strategy framework – however, that doesn’t mean that there’s no value there.

I’m orienting this article as an example of how to take a non-mechanical influence and turn it into a good game. Vlaada Chvátil once noted that game designers aren’t movie directors, and neither are they logic puzzle writers, storybook authors, or anything else. There’s a gap between inspiration and execution, but it’s a gap that can almost always be bridged by good design.

Edit: Jay Treat replied to this article with some interesting points.