Well, I think it to be an instrumental idea as well…!
My only concern is : It being a somewhat hint to the same…,might be protested against by some individuals, giving rise to a somewhat unwanted and unhealthy environment for the coding platform…
To avoid the same, i suggest to make the rating mechanism a bit different for the long challenges, and the introduction of prerequisites (as mentioned by @the_pythor), for some(or all) questions…
What I mean by a different rating ,mechanism is to either set an upper bound to the contests…or decreasing the rating affected by changing it to some arguable percentage of what it presently does,or by some other calculation yielding a less affecting result…
What my motive behind suggesting this is quite obviously to unharm the coding spirit of the participants…Since, making it unrated would decrease the competiton,coz many won’t even care to participate, even if its for their good… and making it usually rated might face a backlash from some of our fellow friends…
So, why not keep everyone happy…
Just a suggestion though…
I agree on your that it would be a good source of learning but it won’t be a competition then, working for days on a question thinking you are getting closer to the solution but actually you have been wasting your time since attempt 1 is kind of soothing.
I think educational competitions like in atcoder and codeforces should be held with prerequisites available when the competitions begin( and they should be general i.e. not in any order - just for all n question prerequisites listed together ) , but unlike atcoder it should contain more than one topic.
As for long challenge the rating system should be changed there are videos of solution of JULY20B out and in just 2 days the submissions of a question rose by 300. I can’t think of a perfect way of avoiding cheating , so what I suggest is that long challenges should not increase you rating this much . They should affect you rating but not as compared to cookoff and lunchtime.
The problem with your idea is that - people will not spend as much time as they do know learning things just to solve 1 question instead they will either spend time (let’s say studying coin problems of DP just to get that question correct ) or not try at all, thinking the they have not studied that topic yet.
After the challenge is over, in the editorial we can learn them as well. I think your suggestions is best for editorials, not for ongoing contest. If these prerequisites are mentioned before the problem in long challenge I think its not good, but those can be mentioned in the editorials after the contest is over right.
I think you’re right. I spend almost 6 to 7 days on single question, as I’m almost a beginner it was very difficult to know which concept is to be used. I tried different approaches, but some test cases were too tight that they were not solvable by my simple approach.
I think not mentioning topics helps a lots in increasing thinking power. As if pre-requisites are mentioned and we learn those topics, then after seeing the problem statement, it may not take a lot of time to know which topic does problem belong and then it would be of no use in solving the question to increase thinking power. If you want to learn new topic or practice problems on a topic, solve them in archive by searching through topic or upsolve after contest. In this way, people who really want to participate in it as contest will not feel discouraged. And for learners also it will help in thinking in a different way and also after seeing editorial, they will learn those topics for future.
