[clay_to_n]

I think the best solution for hackathons is to stop focusing on the prizes. When you have a "grand prize" worth many of thousands of dollars, people are going to compete for it. I think a better idea is to go with just a top 6 / 8 /10, and a lot of sponsor prizes. At Treehacks (the one in this article), there were what felt like 25+ sponsor prizes - and that is cool, because a lot of those are going to kids who are too inexperienced to land in the top 10 at something with 600 students.

But remove the Grand Prize, because it makes the whole thing feel like a competition to people, and cheating / pre-building projects is unavoidable. Even the grand prize winners at Treehacks (built the robot arm, mentioned in the article) partially faked it - they were asked by the judges how much of the hardware they built before the hackathon, and they said they built all of the hardware beforehand. How much of the software had they already tinkered with / sketched out / built? Who knows.

So remove the idea that pre-building is cheating, and instead of 3 big prizes give out 50 small ones.