Sunday, 7 December 2014

Game theory in nature?

I recently saw a BBC documentary and it covered Clark's nutcracker. This bird stores tens of thousands of nuts in stashes of 10 across area spanning upto 100 square miles for retrieval during winter and manages to remember around 70% of them.

This also happens to be a symbiotic relationship between this bird and tree (whitebark pine) from which it gathers these nuts. The 30% forgotten seeds help the tree sow its seeds in the coming spring and reach out farther. The bird's beak and the seeds have matching size and shape.

While marveling at the incredible memory of this bird, I have a feeling that it's not only that. It is perhaps possible that it remembers more than 70%, but chooses to leave the seeds in the ground for the relationship to remain symbiotic. I couldn't find out if the seeds 'forgotten' are at the locations forgotten (it seems so). If however these are forgotten even within a stash, it would reinforce this hypothesis.

Now let's see about ultimatum game. An example runs like this - Alice proposes dividing $100 between herself and Bob, say keeping 90 for herself and 10 for Bob. Then Bob can agree or disagree. If he agrees, both of them keep the agreed split. If he disagrees then none of them get anything.

Standard explanation of rational behaviour says Alice should propose 99 / 1 split because 1 is better than nothing for Bob so he should accept that rather than reject and Alice can keep the rest. However empirical runs of the game seem to split the pie 50/50 to 70/30, but also dependent on priming, cultural norms, and possibly the size of the pie.

Ultimatum game, along with dictator game are often given as examples to illustrate success/failure in a public development project where different stakeholders need to get consensus. This is perhaps more explainable by principle of fairness or reciprocity, especially evident if the game is run multiple times and by switching players. An interesting diversion on this topic is the empirical run of dictator game giving evidence of 'power corrupts', on Dan Ariely's blog.

70% is a convenient number in this case because Clark's nutcracker seems to be Alice in our example and Bob the tree, and they seem to be respecting the reciprocal benefit to continue the repeated game.

It is entirely possible that this symbiosis and the ultimatum game are unrelated, and the matching number is just a coincidence. However I will not be surprised if this bird and tree implicitly knew about this game before us humans. I have argued before that we need to respect other life forms better. This does reinforce that view.

PS: I have deliberately not put this on the other blog which I intended to use for economics-related posts because it seems more inter-disciplinary. Also labels make multiple blogs somewhat irrelevant.