Sometimes is hard to know if a problem is original. I have in memory many problems that I’ve seen in previous contests, Also I check if it is googleable or OEISable. In despite of that, sometimes there are non-original problems that no one in the setting panel have seen before. We find many cases of repeated problems, once one setter found that a problem was in an Ukranian contest (with statements not available in english!), as a result the setter was banned.
The originality In long contests is noticed more than in short contests because there is plenty of time to search on the web.
Regarding OWCAFILE, I think it is a nice problem, I told Kasra that I was a bit affraid that maybe the trick is well known and that it would be a good idea to set the problem not so directly i.e add another layer to the problem with some process that maps (non trivially) to the original statement.
Since the setter of the hackerearth complained that it has an open editorial we had to make the problem non-scorable. Anyway, it is cool problem and I recommend you to solve it, also constraints of OWCAFILE are a bit higher.
In September Long we had the following after contest discussion
Arjun: Hey Alei why did you placed such hard problem in the first slot?
Alei: … Well, the intended cakewalk was not FIBEASY but CHEFINSQ. It is kind of trivial that the smallest sum is obtained by chosing the smallest elements.
Let’s hope that by now I learned better how to balance. I try to make the problemset to entertain you during the 10 days (is that even possible?) In order to do that we need super hard problems, the nearest was in May Long where we had Lewin as one of the setters. Of course legends like ACRush usually solve it in less than 8h. But recently the number of high quality problem proposals have decreased. Keep sending interesting ideas!
Btw did you liked my cakewalk (BRKBKS)? Actually I rejected two of my own ideas before I accepted BRKBKS, we have some boring cakewalks in our database. But I think even the easiest problem should have some beauty.
When solving problems I (and also many other setters) usually don’t look to example input/output, maybe that is why this time I neglected the example output of CHEFARMY. This issue doesn’t happen in CodeForces because polygon always runs the validator on examples.
Also there can be some typos e.g I wrote “The first line of input contains five space separated integers A, B, C, D” no one noticed that it is four and not five …Actually in the first version of the problem there was five integers.
When the problems go to statement verification they are editted, and sometimes there are confusions and we end up with a completly different problem! Also since the distance between lunch time and long contest is short, verification is finished very near the start of the contest and therefore translations are delayed.
There is a myth that we don’t answer clarifications, in the begining I didn’t answered questions that I considered not appropriate or too trivial, but now we try to answer all of them, even with a “No comments”. Some contestants are impatient and expect an immediate answer, however we answer at least once in a daily basis.