Блог пользователя Thaumic_Executor

Автор Thaumic_Executor, история, 3 месяца назад, По-английски

When I got 1550 perf solving ABC in the last round I became angry, but, somehow curious.

I made a survey on the Standings of my 5 last rated rounds.

I define the "pro alt rate" = P(you haven't reached cyan | you got GM perf in this round). Pretty Good.

R967 (Recent): 16/77 participants. 21% pro alt rate.

Edu R169: 9/71 participants. 13% pro alt rate.

Edu R168: 13/72 participants. 18% pro alt rate.

R960: 9/68 participants. 14% pro alt rate.

R958: 13/61 participants. 21% pro alt rate.

Man! What can I say! It's terrible! 1 in 5!

Mike says no alts, but somehow that's useless.

(Tips: The Chineses and Americans made the most alts, not the Indians. I noticed that, do not argue.)

  • Проголосовать: нравится
  • +57
  • Проголосовать: не нравится

»
3 месяца назад, # |
  Проголосовать: нравится -14 Проголосовать: не нравится

Auto comment: topic has been updated by Thaumic_Executor (previous revision, new revision, compare).

»
3 месяца назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

Auto comment: topic has been updated by Thaumic_Executor (previous revision, new revision, compare).

»
3 месяца назад, # |
Rev. 2   Проголосовать: нравится +7 Проголосовать: не нравится

(Tips: The Chineses and Americans made the most alts, not the Indians.)

So the reasons seems to be that indians don't have as much GMs as the Us and China

»
3 месяца назад, # |
  Проголосовать: нравится +104 Проголосовать: не нравится

this is sad and a misuse of alts. alts should be for harassing people without any consequences

»
3 месяца назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

If this stats is accurate, it's quite alot

»
3 месяца назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

as a noobie i need to ask 1. what is perf? 2. P(you haven't reached cyan | you got GM perf in this round) what is this? 3. what is alt?

  • »
    »
    3 месяца назад, # ^ |
    Rev. 2   Проголосовать: нравится +1 Проголосовать: не нравится
    1. Performance: to know where your current level is in that contest
    2. Probability of people who haven't reached cyan had a GM performance in listed contests (I guess so).
    3. Abbreviation of alternatives, that means it's not someone's main account.

    • »
      »
      »
      3 месяца назад, # ^ |
      Rev. 3   Проголосовать: нравится +6 Проголосовать: не нравится
      1. _
      2. No, it's a conditional probability: among people with a GM perf, how many haven't reached cyan yet?
»
3 месяца назад, # |
  Проголосовать: нравится +21 Проголосовать: не нравится

Some people may have the ability to get high perf before they start to compete in codeforces.

»
3 месяца назад, # |
  Проголосовать: нравится +11 Проголосовать: не нравится

You're not finding any indian "pro" alts because they can't reach GM performance kiddo and they are generally bad, they only know how to cheat. I noticed that, don't argue with me

  • »
    »
    3 месяца назад, # ^ |
    Rev. 2   Проголосовать: нравится +11 Проголосовать: не нравится

    Additionally, have you considered the fact that these alts originated from people coming from other platforms such as luogu and atcoder? If they're smurfs, so what? They converge to their true rating anyways and you won't beat them. What's more concerning, is the amount of Indian cheaters that's ruining contests.

»
3 месяца назад, # |
  Проголосовать: нравится 0 Проголосовать: не нравится

Losing or gaining around 20 places doesn't change your perf all that much.

For example, if we take my perf 2967 (the highest perf of an official participant below me in standings) and go 20 (rated) places below we will get 2612, which is a big difference.

However, if we take your perf of 1527 and go 20 places higher we will get 1529, which is not that big of a difference. Even if you go 200 places higher, your perf will only increase to 1544.

Hence you can assume that pro alts have almost no influence on the perf outside of top places.