Sensible game play

If the model of "Spread Misere" is combined with that of Tit-for-tat, it may be possible to provide a viable alternative to trust for establishing collaborative associations on the Internet. Sensible play would entail collaborators being skilled enough to set up sequences of mutually profitable associations that build up slowly towards a position of justifiable confidence.

If it is understood by collaborators that the rules don't include any notion of loyalty, friendship, commitment, altruism allegiance or obligation, then decisions about agreeing to collaborate would have to be based upon a more tangible value system. This would mean making sure that defections from collaborative associations are always seen as having negative values. In other words, anyone in a collaborative association will have the responsibility of making sure that their associate collaborators have a credible expectation of further benefits at a future time - after the time of any current or proposed collaboration - at least sufficient to ensure that it would be unwise for them to defect.

This can be arranged in practice by starting off collaborative associations with activities that involve small risks and small rewards, then, gradually increasing the value of the collaborations as confidence is built up that the other person knows how to play the game and is unlikely to defect. In such a strategy, much effort would have to go into establishing the confidence in the beginning - for little or no real benefits. This foreplay would involve a substantial amount of time and effort, which would be wasted in the event of a defection. This loss of build up effort is often sufficient in itself to make defections less likely - especially by players who understand how to play the game sensibly.

The small risks at the start of a series of collaborative associations will also allow collaborative associations to be built up from nothing, even between complete strangers, because it insures against too much being lost by collaborating with somebody who doesn't know how to play the game - and defects at an early stage.

This is not a radical strategy: it is the way collaborative associations are normally built up in the non Internet world of bricks and mortar. The difference is that in the non Internet world, relationships are forced to proceed at a slow pace; whereas in the world of the Internet there is often the mistaken impression that relationships can be established as easily and instantaneously as it is to communicate with people. Such a misconception is a serious strategic mistake, because there is every reason to believe that reliable relationships for dependable collaboration via the Internet will take much the same length of time to become established as in the non Internet world.

Sensible game play should make suitable allowance for collaborative associations to build up slowly, with risks at each progression being suitably counterbalanced by adequate penalties for any defection by one or other of the collaborators.

The idea of taking little steps towards building up a mutually beneficial series of cooperative associations is accommodated neatly within the framework of a Genetic Algorithm strategy. The separation of any collaborative project into many generations will offer the opportunity for all collaborators to cooperate or defect. Just as in the game of Tit-for-tat, collaboration could be seen as a series of plays that are used to establish a relationship that can blossom into a long standing series of mutually profitable subsequent associations.

Just as a Genetic Algorithm strategy can cope with many simultaneous ideas or approaches at various stages of development, so it can cope with collaborative associations. A few, already established, profitable collaborative partnerships, can provide a reliable income while, at the same time, new relationships are being developed for future projects. In this way, sensible play would see a collaborator's list of contacts evolving: becoming increasingly more extensive and reliable as time goes by. There is no short cut to this process.