Блог пользователя gs15120

Are old problems easier?
Автор gs15120, история, 3 года назад, По-английски

I'm upsolving contests at 2017, but I feel a bit easier than these days even the difficulty is same.

Are the old problems easier?

  • Проголосовать: нравится
  • +132
  • Проголосовать: не нравится

»
3 года назад, # |
  Проголосовать: нравится +57 Проголосовать: не нравится

Yes.

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится -34 Проголосовать: не нравится

are u trolling?

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится +54 Проголосовать: не нравится

True, Old problems with the same rating seem much easier than new ones.

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится +78 Проголосовать: не нравится

Of course

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится +87 Проголосовать: не нравится

It's probably because the ideas we take for granted now weren't as well known back then.

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится +18 Проголосовать: не нравится

Yes.

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится +30 Проголосовать: не нравится

Yes

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится +85 Проголосовать: не нравится

In most cases the statement is true. It's mainly because novel idea becomes a routine as time goes. However, some notorious tasks remain quite hard.

→ Ответить
»
3 года назад, # |
  Проголосовать: нравится -19 Проголосовать: не нравится

I think it is true, especially very old problems.

For example, 6D - Lizards and Basements 2 can be solved by an easy dp, but it is $$$2600$$$. I think it is only $$$1800$$$ now.

I think it is because in that time, some algorithms/tricks are not popular.

→ Ответить
  • »
    »
    3 года назад, # ^ |
      Проголосовать: нравится 0 Проголосовать: не нравится

    The example problem you give has a clear reason why it's higher rated: at the top, it says

    This is simplified version of the problem used on the original contest. The original problem seems to have too difiicult solution. The constraints for input data have been reduced.

    That's why nobody solved this problem in contest. They were literally given a different problem than the one you see now. So the rating calculation is likely solely based on upsolves which is a far more biased sample.

    → Ответить
»
3 года назад, # |
← Rev. 2   Проголосовать: нравится +17 Проголосовать: не нравится

Just my observations:

I have thought about this in the past and I don't think it's a particular trend. Sometimes a problem of rating X may seem easier than other problems of the same rating for a particular contestant. I am regularly surprised by the difficulty of some problems in recent rounds which have been awarded ratings like 2400/2500 but take the same time for me to solve as some old 2200 problem. I have also experienced the opposite: some high rated old problem being easy to solve. So I don't think its a pattern with old problems: just personal subjective difference of difficulty.

In fact, there have been multiple instances of an old CF problem idea being repeated (sometimes verbatim) and the new CF problem getting a higher rating score than the old one. This has happened several times in Edu rounds and also in some div1/2 rounds.

(but extremely old problems (Round <100) may be very easy for their level, because the participation was not that high during the early CF days so rating system wouldn't have worked).

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

    Also "back in the day," there used to be a lot more FSTs, which could explain high rating tag.

    → Ответить
»
3 года назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

Just imagine 10 years from now a 2700 rated problem with FFT tag become easy

→ Ответить