Currently in the April 2017 Community Challenge selection, we have two answers that have been drawing for a while now. Or at least near enough. And so it got me thinking, what should we do if two answers draw by the selection date?
Currently we do:
At the end-of-day on Friday, March 31st, the top-voted post will become the next challenge.
But this doesn't take into account draws.
How should we handle draws when selecting community challenges?
I think that if two answers score the same amount of votes by the end of the selection phase, both answers should be picked. This would require a small change to the current selection text:
At the end-of-day on {date}, the top-voted post(s) will become the next challenge(s).
It would also require an edit to the question to indicate that both answers have been selected. And so something simple like:
Both Chain (or belt) drive system and Simulate a Multi-Way Intersection won this challenge, you may pick to do either.
One possible way to avoid this situation is for the author of the question not to vote until seconds before accepting an answer.
You only need one person to break the tie after all, and all votes after acceptance of the answer don't matter.
I posted the chain drive challenge, which is currently almost tied for first place, so I'm biased but:
There are a lot of good challenge suggestions this time, and I'm tempted to try almost all of them. It's been the same for other rounds of Community Challenges.
This can lead to repeats from round to round - my own suggestion being one of them this time. Over time, we could end up with a backlog of well-liked challenges all vying to get picked each time, either reposted by the original author, or by someone else who really wanted it to win in the previous round. Or, more likely, some cool challenges just fall off the radar because they aren't reposted, despite having been quite popular and still holding a lot of potential. That seems like a loss.
So, like Peilonrayz, I'd suggest we consider either picking the top N as community challenges, tied or not, just to give the community more to choose from, and to clear out the backlog and keep more new challenges coming. (Ideally, the top N would have different levels of complexity or focus on different disciplines but that's probably too hard to ensure.)
Or we could have a list of "standing community challenges" that people can pick from at any time. Lots of questions on CR are already about programming challenges from other sites (e.g. HackerRank), so having a pool of challenges here might be interesting. Especially as these won't have to conform to the machine-checkable input/output that other challenge sites rely on, since implementations would be reviewed by the community.
While I do see the value in picking one as the challenge for the month, and focussing on that, the whole "there can be only one" style of it is also sort of frustrating. Sometimes I've seen a challenge I really wanted to try, but it didn't win, so rather than being encouraged to post something, it had the opposite effect (again, this time I have my own dog in this race, so I'm biased).
