Results 1 to 4 of 4

Thread: Precaution chances, random seeds and spring blossoms.

  1. #1
    Sometimes Hero Demi Moderator silktrader's Avatar

    Join Date
    Jan 2007
    Location
    A woman's womb
    Posts
    5,918

    Precaution chances, random seeds and spring blossoms.

    An assumption will follow based on a few observations:
    a) Thuls heavily rely on their "rawpaws", "blackhides" (their elite counterpart) are not a viable option for various reasons I won't bring up here. As a consequence, the first squad that is targeted by precautions ends up being "rawpaws".

    b) The most useful precaution to single out a particular squad is "Worm Tongue". "Plague" chooses squads randomly (although it's incredibly useful when the player hasn't researched elites). "Worm Tongue" (which is misspelt in the game) is extremely effective, when successful.

    Now, during several matches I could notice an odd repetition:

    c) I would get "Worm Tongue", one after another.
    There were times when I got three "Worm Tongues" in a row, playing against the same opponent in three different matches. The chance of having three consecutive "Worm Tongues" is 0.11^3 or about 0.1%.

    I've had six "Worm Tongues" in the last eleven matches. The chance of having six "Worm Tongues" in eleven matches is 0.045703% (through a binomial distribution of a Bernoulli trial) — which is higher than Trefort's chance of finding a good looking girlfriend but generally, incredibly, low.
    ***


    I now wonder: could it be that the random seed from which stem precautions success chances is somehow dependent on previous precautions successes. I believe that is a question to which only people with access to source code can provide an answer.

    If precautions' random seed was actually saved and reused for successive precautions, or even modified according to already failed or succeeded precautions it would be possible to manipulate things so to have distinctively higher chances of succeeding a late "Worm Tongue" or any other precaution with low chances.

  2. #2
    Sometimes Hero Demi Moderator silktrader's Avatar

    Join Date
    Jan 2007
    Location
    A woman's womb
    Posts
    5,918
    Note to self: the odd string of "Worm Tongues" continues.

  3. #3
    Sergeant

    Join Date
    Nov 2010
    Location
    ,Germany
    Posts
    55
    1.) Precautions aren't reworked yet.

    2.) Worm tongue checks on every single unit which exists from that type and checks if it gets applied. Affected units get spawned first. That's why it seems to have such unlikely sucesschances.

  4. #4
    Sometimes Hero Demi Moderator silktrader's Avatar

    Join Date
    Jan 2007
    Location
    A woman's womb
    Posts
    5,918
    Quote Originally Posted by Nedom Wolven View Post
    1.) Precautions aren't reworked yet.

    2.) Worm tongue checks on every single unit which exists from that type and checks if it gets applied. Affected units get spawned first. That's why it seems to have such unlikely sucesschances.
    That's a helpful answer, thanks. So, to recapitulate:

    "Worm Tongue" actually has a chance to succeed (11%) on each recruit of the targeted kind. Squads which are struck with "Worm Tongue" are then deployed before others. This behaviour seems in contrast with the wiki and game's description but I would nonetheless be very happy to understand — please correct me.

    In fact, "Worm Tongue" would be reported as successful when at least one recruit is affected. For 40 recruits, it's almost a certainty it will happen (about 99% — which still refutes my sample's results). A viable option would then be to sensibly decrease the number of available recruits, to reduce the number of possibly affected squads.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts