One issue with Zendo that doesn’t necessarily make it a great game for actually improving rationality is that the rules are made by humans, so you can use psychology to get information. For example, I was playing with my family a few nights ago and my mother guessed a certain rule, my little brother immediately said “That’s not Josh’s rule. it isn’t elegant.” And sneakier players can do even worse stuff using sidechannel attacks. In particular, how long it takes for the Master to judge whether different configurations fit the rule can give a lot of information about the rule form.
Upvoted for truth. Hopefully this game will avoid those pitfalls because:
1) We’ve never met
2) The time it takes me to judge a rule is swamped by the time it takes me to notice somebody posted
Also, using psychology to get information and improving your performance that way still requires rationality.
Aren’t those all just additional aspects of the game, additional constraints in the optimization process?
If people will make inferences from how long the Master takes to judge whether different configurations fit the rule, then the Master should either systematically manipulate his times to give the wrong impression or introduce some random waits. And if you are known for only using elegant rules, then you’re using a sub-optimal strategy for choosing the patterns, because you’re giving away more information than necessary.
Humans are the only things capable of reliably generating things that seem non-obvious to humans. The only reason the universe seems so good at it is because we pay less attention to the obvious things. I don’t think we can improve on this issue easily enough for it to be useful for the purposes of the game.
Doug Orleans told me once of a version like this he made to be played with an IRC or MUD bot (I forget which). A rule was a regular expression. (This came up when I mentioned doing it with Lisp s-expressions for the koans instead.)
Rules clarifications:
Mondo guesses may be changed at any time before the judgement is posted.
If someone posts and it hasn’t been judged yet, other people can feel free to post while you wait for me to show up.
Meta stuff goes here.
One issue with Zendo that doesn’t necessarily make it a great game for actually improving rationality is that the rules are made by humans, so you can use psychology to get information. For example, I was playing with my family a few nights ago and my mother guessed a certain rule, my little brother immediately said “That’s not Josh’s rule. it isn’t elegant.” And sneakier players can do even worse stuff using sidechannel attacks. In particular, how long it takes for the Master to judge whether different configurations fit the rule can give a lot of information about the rule form.
Upvoted for truth. Hopefully this game will avoid those pitfalls because: 1) We’ve never met 2) The time it takes me to judge a rule is swamped by the time it takes me to notice somebody posted Also, using psychology to get information and improving your performance that way still requires rationality.
Aren’t those all just additional aspects of the game, additional constraints in the optimization process?
If people will make inferences from how long the Master takes to judge whether different configurations fit the rule, then the Master should either systematically manipulate his times to give the wrong impression or introduce some random waits. And if you are known for only using elegant rules, then you’re using a sub-optimal strategy for choosing the patterns, because you’re giving away more information than necessary.
Humans are the only things capable of reliably generating things that seem non-obvious to humans. The only reason the universe seems so good at it is because we pay less attention to the obvious things. I don’t think we can improve on this issue easily enough for it to be useful for the purposes of the game.
Since nobody else has done it, here’s a post which I will keep edited with the current Master requests and responses:
As of 03:29 Tuesday January 4 UTC:
Doug Orleans told me once of a version like this he made to be played with an IRC or MUD bot (I forget which). A rule was a regular expression. (This came up when I mentioned doing it with Lisp s-expressions for the koans instead.)
Rules clarifications: Mondo guesses may be changed at any time before the judgement is posted. If someone posts and it hasn’t been judged yet, other people can feel free to post while you wait for me to show up.