So, first of all, I would like to apologize for the issue with the problem B in today's ER. We tried to fix it as quick as possible, but unfortunately, still lots of submissions were made and this caused a big queue during the round. I am sorry for this.
How did this all happen? Well, this is really a sad story about one commit.
About three hours before the round, while writing explanations for the problem B statement, I noticed that the answers were different from the correct values by 1. I opened the model solution in another tab in Polygon, wrote +1 to the value I printed, clicked "Save" and then committed the changes (including the notes for the samples). But, unfortunately, that change in the model solution wasn't included in the commit.
The cause was that the model solution didn't save in time for the commit (and actually never saved for some reason, maybe some bug in the browser or Internet connection issues). I didn't notice that, and while rereading the problem again, I never checked that part because "well, I fixed that thing, it is definitely correct now!"
The point of the story? Always recheck what you're actually committing. Always double-check even the parts of the problem you have already fixed, even if you're sure that after all the changes you've done, they're 100% correct.
I am again deeply sorry for the problems it has caused, and I hope that it won't happen again.