last day (1076 days later) » 

@PhiNotPi you wanna post a placeholder in the sandbox and delete the theme proposal?
Okay.
> Ordinary KOTHs require that a single user take up the job of hosting the competition: running the whole competition at once by themselves. Everything depended on them, and they can only run so many competitions or update ever so frequently. I think this can be improved... somehow.

Here are some ideas, many of which would require an encryption scheme or something to ensure correctness.

- Anybody can run the tournament and add their results to the current leaderboard somehow.
- When a bot is added/updated, only the new pair of contestants should be tried, and nearly anyone can update the
I have a few more ideas for this challenge. You guys can post your own ideas, too, if you have any.
1. A KOTH runnable via stack-snippet (an idea by Ypnypn)
2. A KOTH controller that uses a hashing mechanism to make the results verifiable.
3. A purely deterministic KOTH, where the results are the same round-to-round. (Like Nano-Core-War)
4. A KOTH that takes place over a long timespan.
(I have to leave now, be back in an hour.)
Maybe a way to run a tournament and have the (permanent) leaderboard auto-updated.
 
2 hours later…
21:23
If multiple people can run a tournament and the average is automatically updated, is there a risk of someone running lots and only uploading the ones that go in their favour?
If results can be guaranteed the same regardless of computer, so they only depend on the random seed, then the OP could provide a long list of random seeds (or a way of generating them in order). Then each new tournament has to use the next seed on the list so there can be no choosing.
We could have a stack snippet that only accepts new data if it uses the next unused seed. How could this be proved?
I guess that's a variation on using a hashing mechanism to make the results verifiable
21:45
I was actually planning a fully automatic Stack Snippet based KOTH for my 100th challenge. I got interested in making other challenges but I'm still planning to post it someday. It could serve as a the fortnightly challenge if anyone wants. I may Sandbox it later today/tomorrow.
@Calvin'sHobbies Would that be automatic for testing purposes, or would each person running the stack snippet contribute to the official average score too?
@Calvin'sHobbies Oh I thought you were gonna go for the answer-chaining DAG. ;)
@trichoplax Anyone could run the snippet to run the official KOTH battle or some subset of it (I guess the official battle would be run on my machine).
The simulation doesn't work at all but you can decide which entries will compete and whatnot.
@Calvin'sHobbies If it's still officially run from one machine at one time, then I guess it is separate from this fortnightly challenge so it should remain yours. Have I understood that right?
22:05
I guess. It would be asynchronous in the sense anyone could run the contest whenever they wanted to see the results, but no, not distributed.
@MartinBüttner I thought about it but I was more certain that a KOTH would be well received.
fair enough
any hint what the actual game is gonna be about?
ah k
I really want to do my first KotH some time soon, but I'm too busy with uni at the moment to work on a controller... maybe in April.
22:24
@Calvin'sHobbies I also have a long term intention to post a KotH which that question now might make a duplicate. It's basically the same but with pixels being moved instead of added. I think it's different enough but I need to write it up and see what the sandbox says...
22:45
I'm on holiday as of 8 hours ago, so if any C# needs writing or reviewing, feel free to prod me (if the quality of my code last time didn't put everyone off)
@VisualMelon It was appalling. I had to wash my eyes with sandpaper.
wow, that bad? I'd better close my GitHub account for the good of the people
(Seriously though, thanks for the offer. Might come in handy... although C# is probably not the best platform for a distributed KotH.)

  last day (1076 days later) »