Instead of giving the amount to the captain, they could throw it away to the sea. Nobody would cheat then.
In real life situations, you may want to replace the sea by the government or some kind of common fund in the domain.
> Nobody would cheat then.
Depends. Perhaps somebody would try to bully (i.e. play a game of chicken http://en.wikipedia.org/wiki/Chicken_(game)): "Either you lower your share in my favour, or I will claim too much and you get nothing."
It wouldn't work if it was explicitly known. The others could just as easily decide to cut him out of the distribution entirely.
Yes. Or if the victim could commit to not yielding to threats like this beforehand. Being able to commit oneself is of huge importance in games like chicken (http://en.wikipedia.org/wiki/Chicken_(game)).
Re:Self-Enforcing Protocols
by silentOpen on Tuesday August 11, 2009 @02:48PM
(#755993)
Interesting idea but no parties in the protocol would be motivated to do that. This might be useful in a situation where a group entity exists that is approximately the same size as the division group -- e.g. an endowment, shared property, common business, or co-op.
If I were a pirate who had little gold to lose, and a grudge against some of my fellow pirates, I could always play the spoiler and cause everyone's gold to go to the captain or into the sea...
Yeah, I like the comments in this sub-thread. These protocols can suffer from Maginot Line type of security. Also, they can provide an illusion of propriety that cheaters can point to when accused and use that to distract, say, a jury from however it is they really cheated. It's important to understand these protocols' failings.
Re:Self-Enforcing Protocols
by pavel_lishin on Tuesday August 11, 2009 @11:33AM
(#755427)
Isn't the point of a self-reinforcing transaction that it doesn't need a third-party mediator?
Re:Self-Enforcing Protocols
by jedi_stannis on Tuesday August 11, 2009 @11:37AM
(#755441)
I think his point was that current third party mediators will not be happy to be cut out of the equation. Even though the system will be more efficient without them, they will try to do whatever they can to still get their share of the transaction.
Thanks - that is exactly what I meant.
I think the property tax example is just an inverse pirate scenario. (the landowner is the pirate needing the cooperation of the captain (the head of the land acquisition authority)). Granted a much weaker position to negotiate from.
One thing that I wonder is how is a self reinforcing transaction affected when the third parties that used to mediate it are shut out. For example, certificate authority companies that are no longer needed.
Furthermore, the "all hands" in the room as opposed to secret ballot voting - while on the surface it is self reinforcing from a social perspective it wouldn't really work (i.e. the coercion he mentioned and also the lemming effect where someone votes with their crowd delaying raising their hand just enough to see what the consensus is).
Finally, the collusion and breakage of a self-reinforcing system like the pirate example shows that a self-reinforcing protocol is situationally dependent. What are the chances that at least one pirate will collude with the captain? (I bet the chance is quite high considering the nature of pirates) As opposed to the property tax example which seems to be a very good use of such a protocol.