Psychology, economics and mammoths
In Mammut you take the role of cavemen, sharing the spoils of the latest hunt. A bag of two-sided tiles is emptied in the middle of the table, giving a randomised assortment of tusks, fur, meat, fire, stone axes, and a bunch of other prey animals the hunters managed to snag. Why are they dividing up fire and axes as part of the spoils? That’s not really explained nor important here.
Each type of tile scores differently – for example, each tusk gets you two points, having the least amount of fire will get you negative points, and having a large variety of other animals will score a bunch of points at the end. What really makes this game different is the way the spoils are shared. The starting player can take as many tiles as he wants from the middle of the table – even, if he wants, all of them. The next player has a choice – take tiles from the pool, or take the first player’s tiles. If he takes the first player’s tiles, one of those has to go back in the shared pool. Play continues, with each player making the same choice – tiles from the middle, or someone else’s tiles minus one. The round ends, and scoring begins, when the last player not to have any tiles takes the remaining loot from the centre.
It’s part game, part social experiment. On your turn, if you take too many tiles, someone will simply take them from you. Take too few and you’ll score poorly. So there are tough decisions as you try to take more than your fair share, but not so much that anyone objects. That it’s a game with scoring means that you’re forced to be rational – and when it comes to sharing, people are not rational.
Of course people are not rational, you think. But neither are you. A thought experiment: you need to go to the shop to buy a loaf of bread. In the shop nearest you, the bread costs £1. Walk another three minutes and you can buy it for 50p. Most people will walk that extra three minutes and save 50p. Now imagine you want to buy the latest Fantasy Flight big box game. It comes with a whole load of miniatures and tokens and stuff, so let’s say it costs £100 in the nearest shop. But walk another three minutes and there’s a shop selling it for £99. Less people than before will walk that extra three minutes to save that £1.
This makes no sense. In the first example people feel that walking three minutes is worth 50p. In the second they feel it’s not worth £1. For an economist, this is stupid – if someone walked three minutes for 50p, surely they should do so to save £1. Our flawed human reasoning doesn’t see it like that, seeing the 50p and £1 savings relative to what we’re buying. It’s something to keep in mind the next time you see SAVE 50% stickers.
When we’re asked to share something, our reasoning is similarly flawed. The “Dictator Game” isn’t a game as such, but an experiment in psychology and economics. Two people who don’t know each other play the game – one is given a sum of money, and told to share it with the other person in any way they choose – they are the ‘dictator’. The other person has no input into this decision at all. If people act entirely in their own interests, as traditional economics says they should, then they should take all of the money, or at best leave a minimum amount for the person they don’t know. But they don’t. People tend to split the money more evenly than they ever need to, despite no compulsion to do so.
Add in a bit extra to this and you have the “Ultimatum Game”. The dictator doesn’t have the final say in this game – instead, the other player gets to accept or reject the offer. If the other player rejects the offered split, no one gets any money. So if one player gets £50, peels off a fiver and offers it, the other player can – and often will – say “no”. Again, to the economist, this makes no sense. The second player is getting £5, which is better than nothing – why on earth would you reject something in favour of nothing? But people will, when given too little, punish both of them for not playing fair. Similarly, the first player tends to see foresee this outcome, and in general will offer more than the minimum player two will accept. Rationally, he should offer the minimum, and rationally, his partner should accept it. But this is a far rarer outcome than an equitable split.
What this means is up for debate – but it does mean that people are not rational about sharing. Some see it as a evolved, altruistic behaviour ingrained in us. These games are designed to be played as one-off events, but in reality we have to share with the same people again and again. The ‘iterative’ game is different – people have memories, and if we screw them over this time then they might remember and screw us right back. It’s not just people who act this way – naturalists have seen patterns of altruism and punishment in animals.
Which brings us back to Mammut, which seems to be an exception. Because players score each round, it’s in their interests to act rationally, and grab all they can for themselves. Anyone taking too little through some misguided altruism will simply lose the game. If economists want to hold on to the simplicity of ‘rational actor theory’, then they could do worse than play board games.