Proposition Game Crit. point of the game http://tunedit.org it would have been nice to have seen this site earlier in the semester to give a better idea of the motivations behind the prop. game. but more collaborative personal exp. guilty: My group posted incorrectly formatted propositions/examples more than half the time. (usually they were called out and corrected, but since there were no consequences there was no incentive to be more careful.) Many of the challenges/agreements we and others got were just ignored. Especially after an assignment has finished maybe explicitly include which part of your grade comes from piazza so people have to be more responsive. This means you actually have to tally everyone's postings each week- I'm not sure how helpful piazza's administrative tools are. problems difficult to play with more than 3-4 pairs. everyone posting their own claims doesn't work well usually the 'game' would cluster around 1 claim, and the other 15 or so would be neglected, or get some perfunctory agreement. went well: improving algorithms incrementally, getting ideas about what to post from other people's solutions. don't change piazza is great definite changes no grading solutions by # seconds, too system dependent fully automated avatars are a bad idea - even pseudo-implementing them wasn't much use. possible changes make it more formal? have a formal scoreboard where pairs compete no editing claims after the fact strongly enforce scoring points lost for using the protocol incorrectly should probably compete on # machine instructions/specific examples to be fair pros: students are motivated by competition gives the protocol legitimacy speed of making claims effects score: people will participate more to optimize their score cons: will mostly motivate top 2-3 pairs competing for first zero sum game leads to a depressing scoreboard speed of making claims effects score: may lead to undesired effects/gaming the system machine instruction counting is a pain/unfeasibly slow make it less formal? people discuss/improve algorithms openly on piazza maybe once people have agreed they must explain how they got their running times/talk about their algorithm implementation compete on asymptotic running time/general code no real scoring system pros: better learning experience, especially for students who are having trouble/aren't engaged develop better final algorithms collaboratively less aggravation with the formal protocol/posting with an incorrect format cons: sloppiness students can piggyback on others work instead of coming up with algorithms themselves