Page 2 - Contest Reorganization

Yes, this is the kind of judging script I was talking about. Bascially my advice is to have all of the contest infrastructure setup before the contest begins, so all that you need to do at the end is plugin the entries and testcases. This is more upfront work for the admins/judges, and it requires some dicipline on their part, but it doesn't really make for more work in total and should make for a smoother event.



I'm not sure this is a good idea. Why make things unnecessarily subjective? What advantage does it give you?

I guess it depends on the nature of the contest, but by limiting each entrant to a small number of test cases I think you'll have more interesting (and thus, I would say, higher quality) testcases. If you are going to filter the testcases that are submitted, I think there should be some objective metric: like you take the N cases with the greatest range between hi-score and lo-score, or you take the N with greatest statistical variance in score. You might also want to constrain things so that you don't accept more than M testcases from any one person. I agree that the judges should be able to add their own testcases to the mix, and that contestants should not be obligated to provide testcases.

Anyway, just my two cents.

randomly picking testcases is good I think. Because lets say I submit 5 testcases, that my script performs perfect on, then I have an advantage, and can just hope they choose 3/5 test cases, which would get me in the top scorers. You don't want something objective, because if there is a pattern to the chosen test cases, if you design your script around a known test case, and you know the objective method for choosing it, then you can exploit that. I'd just pull test cases out of a pool of them randomly.

I'd also suggest having several judges. Some people who have alot of free time. We understand that everyone who runs this site has a life, and a job, and all of those things, so why add that stress to someone who has a mess of things to do. I'd set up a judges pool, then ask who has time to judge them in the near future, and call it good. That would allow someone who has the time to judge it to do so, instead of having to overload someone who is already immensely busy