Блог пользователя AURRUM

Variety of topics in latest rounds
Автор AURRUM, 2 года назад, По-английски

Recently I've noticed that in almost every latest CF round variety of topics is very little. I dont know about the harder problems that are highly above my level, but at least for problems C,D (sometimes even E) of div2, almost all of them have the following structure. "Find one observation and code it in few minutes". No DP, no graphs, no data structures (segment tree and so). Why is this happening? Or maybe it's just for me?

  • Проголосовать: нравится
  • +55
  • Проголосовать: не нравится

»
2 года назад, # |
  Проголосовать: нравится +70 Проголосовать: не нравится

i am looking forward to your 1400 rated segment tree problems

→ Ответить
  • »
    »
    2 года назад, # ^ |
      Проголосовать: нравится +12 Проголосовать: не нравится

    Sorry, I was speaking mainly about C and D but I expressed it wrong, now it's edited

    → Ответить
»
2 года назад, # |
  Проголосовать: нравится +38 Проголосовать: не нравится

You can try AtCoder Beginner Contest.

→ Ответить
»
2 года назад, # |
  Проголосовать: нравится +33 Проголосовать: не нравится

I thought that had been the case for the last 3 years. Graph problems that aren't dfs-tree are especially rare too.

→ Ответить
»
2 года назад, # |
← Rev. 2   Проголосовать: нравится +30 Проголосовать: не нравится

Problem topics of the first few problems that you haven't solved in recent contests, i.e. what you should solve to improve:

  • #779: constructive, bitmask, games
  • Edu 125: sorting + binary search, dp + combinatorics
  • #777: binary lifting
  • Edu 124: graphs, matching + combinatorics
  • #776: greedy + data structures, dp, dp + graphs

I can see a broad variety of topics here.

→ Ответить
  • »
    »
    2 года назад, # ^ |
      Проголосовать: нравится +15 Проголосовать: не нравится

    To be fair, your appeal feels somewhat justified to me, as many of the problems that I solve require some observations before relatively short implementation. However, I don't see how this is bad.

    Every problem that isn't well-known will require some observations, including a segtree one. Also, if you have a good template for segtree (say AtCoder library), then you will notice that actual implementation is also short.

    → Ответить
»
2 года назад, # |
  Проголосовать: нравится +8 Проголосовать: не нравится

They are running out of questions.

→ Ответить
»
2 года назад, # |
  Проголосовать: нравится +4 Проголосовать: не нравится

Yes, absolutely correct. To find good DP/Graph/SegTree problems, one has to look at problems rated in the range of 1900/2000+. Nowadays, A, B and C problems require just one observation and there you go. Seems more of like a waste of time to me.

→ Ответить