Designing rationalist games?
14 views
Skip to first unread message
Emile Kroeger
Feb 23, 2012, 5:59:36 PM2/23/12
to lesswrong-...@googlegroups.com
An idea that regularly pops up on LessWrong is making a rationality video game ... it's been discussed in the threads here: http://wiki.lesswrong.com/wiki/The_Less_Wrong_Video_Game
I've been rereading those threads and there are some interesting ideas, though there's also stuff that sounds like a lot of effort to develop and wouldn't deliver much rationality bang for the buck.
How about a new round of kicking ideas around?
For example, how could a game teach someone to not deceive himself? That would require:
1) A situation in which one is likely to deceive himself because of a specific bias (the conjunction fallacy, confirmation bias, hindsight bias, etc.)
2) A game system that rewards overcoming the self-deception (either because it is detected by the system - which is hard! - or because true beliefs are instrumentally useful in the system)
This seems significantly harder than the cleverness/deduction games (like mastermind, minesweeper, Professor Leyton, etc.). One approach is to focus on calibration, and measure something that the player has reasons to get wrong to give a better image of himself (for example, chances of completing an angry-birds-like level in less then three tries, or confidence in one's answer to a trivia question). Another approach is to take a fully scripted path, with a story where one could expect the player to be biased, and then have a revelation - which is nice, but is a bit of a one-trick poney.
It seems that each bias would require a special game mechanic (some of which would be quite hard to design - hindsight bias???), and players may be more likely to just learn how to game that particular mechanic through trial-and-error without overcoming the underlying bias.
Maybe just going through http://en.wikipedia.org/wiki/List_of_biases and coming up with game ideas would work. And there are also rationality skills that aren't about a specific bias ...
Thoughts? What kind of rationality games have you made, or have you thought of?
Emile
I've been rereading those threads and there are some interesting ideas, though there's also stuff that sounds like a lot of effort to develop and wouldn't deliver much rationality bang for the buck.
How about a new round of kicking ideas around?
For example, how could a game teach someone to not deceive himself? That would require:
1) A situation in which one is likely to deceive himself because of a specific bias (the conjunction fallacy, confirmation bias, hindsight bias, etc.)
2) A game system that rewards overcoming the self-deception (either because it is detected by the system - which is hard! - or because true beliefs are instrumentally useful in the system)
This seems significantly harder than the cleverness/deduction games (like mastermind, minesweeper, Professor Leyton, etc.). One approach is to focus on calibration, and measure something that the player has reasons to get wrong to give a better image of himself (for example, chances of completing an angry-birds-like level in less then three tries, or confidence in one's answer to a trivia question). Another approach is to take a fully scripted path, with a story where one could expect the player to be biased, and then have a revelation - which is nice, but is a bit of a one-trick poney.
It seems that each bias would require a special game mechanic (some of which would be quite hard to design - hindsight bias???), and players may be more likely to just learn how to game that particular mechanic through trial-and-error without overcoming the underlying bias.
Maybe just going through http://en.wikipedia.org/wiki/List_of_biases and coming up with game ideas would work. And there are also rationality skills that aren't about a specific bias ...
Thoughts? What kind of rationality games have you made, or have you thought of?
Emile
Alexei Andreev
Mar 3, 2012, 5:56:45 PM3/3/12
to lesswrong-...@googlegroups.com
I've been working on a game, Alchemy College, based around Eliezer's The Simple Truth. I believe most of you participated in the design stage of that game. Right now it's about 60% complete, with most art and music assets already finished. However, I have to abandon it (I'll go into "why?" in another email). If anyone wants to pick it up and finish it (in whatever form you want), let me know. I'll assign all the copyrights to you, and email you all the assets: Unity3d project, C# source files, art, and music. And if you are actually serious about finishing the game, I can put you in touch with the artist and the composer that created the assets, so you could finish the game.
Alexei
Alexei
thomblake
Mar 4, 2012, 7:29:27 PM3/4/12
to lesswrong-gamemakers
That sounds awesome, though I'm in no position to take on a project
like that right now. I'll have to see what the status is, when I
finish my dissertation, if nobody else has picked it up.
On Mar 3, 5:56 pm, Alexei Andreev <alexei.andr...@bentspoongames.com>
wrote:
like that right now. I'll have to see what the status is, when I
finish my dissertation, if nobody else has picked it up.
On Mar 3, 5:56 pm, Alexei Andreev <alexei.andr...@bentspoongames.com>
wrote:
> I've been working on a game, Alchemy
> College<https://docs.google.com/document/pub?id=1DPr3ZwIl2pty0h97pIFsaQo328Ip...>,
Reply all
Reply to author
Forward
0 new messages