Schrödinger's cat and the non attended battle against AI fleets.

Ka Dosh
Ka Dosh
Minor Nuisance
Joined Sep 2015 Posts: 130

My biggest concern about those battles we play against the different events available in each event is this:

I've not entered in battle when I attack x AI fleet using autoauto function and not using the auto function. Also I've entered in battle with auto and without auto. The results have been the same for each kind of experiment.

In the viewed battles without human intervention no more than summit expectation, the received damage is low and lower if you use auto function. Nothing special of review.

The things get spiced when I don't watch the battle. Is supposed that the results must be the same as above. The is not intervention at all. But the damage suffered is tremendous. Even it appears high level fleets destroyed with AI fleets level 60, 65. What is happening here? Kixeye's AI attacks specially any of our ship commanded by cats? Or there is other reason for this battles behavior?

This becomes important because the server has disconnected me several times since I've logged in my first time in the half of the battle when I'm winning and 2-3 attends later when finally the server deigns to connect me again, my fleet is destroyed or almost destroyed. Only one time I've returned in less than 5 seconds and my fleet capable of destroy 20 fleets of that kind started to BOOM separately each ship. The fun thing is over leaved them very compacted and 5 second later they where spread throughout the whole battlefield.

I would like some CM would like to ask the battle/AI programmers in search of an answer for this situation. I hope they don't come with cats' answers... ;)

  • CM Dade Murphy
    CM Dade Murphy
    KIXEYE Community Manager
    Joined May 2015 Posts: 513
    I double-checked with our Engineers to make sure - if auto is used for both an attack being viewed viewed and one that's not, the AI behavior should be the same.
  • Ka Dosh
    Ka Dosh
    Minor Nuisance
    Joined Sep 2015 Posts: 130

    Well, dear CM Daddy Murphy, Kixeye AI must hate my little niece account and mine too just because both charges the name of my baby! Because the repair times we have is tremendous when we don't attend the battles. And with attend, I mean log into the battle. The ship the always suffers greatly is the number 3, no matter what formation we use. Normally, the battleships receive less damage than the cruisers. Is funny that a fleet with levels 96-104 suffer damage level between 10-20% with the funny Riot lvl 60-65 fleets. And I remark that the damage is taken when we don't log into the battle. If we log in and stay immobilized the damage percent is about 1-2% in the worst of the cases.

    Again: Kixeye AI hates blonde haired, blue eyes funny baby's faces! ... And cats!

  • LilBasterd
    LilBasterd
    Force to be Reckoned With
    Joined Mar 2018 Posts: 1,871
    its not though. AI vs AI should give same results every  time.  When same fleet by player  is used,same commander, to fight same enemy fleet, but it isn't

    Something  is occuring that  changes our AI or enemy AI  to give varying  results..
    Sealcorps is my biggest fan, ty Sealcorps  for always reading what i write. :)
  • Ka Dosh
    Ka Dosh
    Minor Nuisance
    Joined Sep 2015 Posts: 130

    I am going to do some statistic tests for determinations about if the game use stochastic behavior and/or quantum mechanic principles. If my conclusions are different from what Daddy Murphy says, then I will demand Kixeye for not less than 300 millions of coins! ADVERTENCE: I will not accept Blood Amber as payment!

  • ToxicBait
    ToxicBait
    Greenhorn
    Joined May 2017 Posts: 7
    edited 14 Sep 2019, 5:50AM
    Ka Dosh said:

    I am going to do some statistic tests for determinations about if the game use stochastic behavior and/or quantum mechanic principles. If my conclusions are different from what Daddy Murphy says, then I will demand Kixeye for not less than 300 millions of coins! ADVERTENCE: I will not accept Blood Amber as payment!

    you know they are not going to compensate you like 300 million coins or even one coin right? The AI behaviors are a bit RNG based, I have tested a couple time one paladin vs the event 75 hobliers on auto and the repair varied. Honestly you can just do manual battles or get a better connection. The fact that you asked them for 300 million coins if your conclusions are different is rediculous
  • LilBasterd
    LilBasterd
    Force to be Reckoned With
    Joined Mar 2018 Posts: 1,871

    Makes no sense. It is AI vs AI. Each one has a preset reaction to what the other does. Thus battle results should be very similar... some varience is given but large variances multiple times is well uncommon...

    Sealcorps is my biggest fan, ty Sealcorps  for always reading what i write. :)
  • Ka Dosh
    Ka Dosh
    Minor Nuisance
    Joined Sep 2015 Posts: 130

    ToxicBait: if you follow my sort of commentaries, you will note that I'm always in a critic mood with acid/silly jokes that always have hidden thoughts and references... In fact, all my commentaries in this post are plentiful of them!

    I desire for you the best! And 300 millions cats! All alive! What comes first!

  • GoodmanGroup
    GoodmanGroup
    Potential Threat
    Joined Aug 2013 Posts: 66
    Don't be surprised that Kix cheat. When You enter a casino you know tthey cheat.
    I's all about making money
  • cderalow
    cderalow
    Minor Nuisance
    Joined Oct 2015 Posts: 268
    It would not surprise me in the least if Kixeye revealed that their AI system for both players and pve fleets was random number generated.

    It has always been a bit of my suspicion all the way back to auto rancor fleets for events.  every battle v same target would generate a slightly different result, despite the fleet compositions, arrangements and armaments being identical.  minor variance in most cases, but maybe 5% of the time there'd be a major variance.


  • airdrop airdrop
    airdrop airdrop
    Potential Threat
    Joined Aug 2019 Posts: 26
    edited 18 Sep 2019, 1:25PM
    I had the same feeling during the event this weekend. I was farming 91 cybernauts with the 5 free bastion cruisers and 1 zeus mk5. When I was viewing the battle (full auto, no action from my side) it ended up with 4% damage. But during a battle where I wasn't watching (but auto still active of course) I've lost over 20% ! That shouldn't happen ...
  • LilBasterd
    LilBasterd
    Force to be Reckoned With
    Joined Mar 2018 Posts: 1,871
    I double-checked with our Engineers to make sure - if auto is used for both an attack being viewed viewed and one that's not, the AI behavior should be the same.
    can you explain the plain as day results we are getting then.? that  completely contradict what they are saying?
    Sealcorps is my biggest fan, ty Sealcorps  for always reading what i write. :)
  • Ka Dosh
    Ka Dosh
    Minor Nuisance
    Joined Sep 2015 Posts: 130

    Then I wasn't alone! "Meow! Meow" to all my brothers in arms!

  • GoodmanGroup
    GoodmanGroup
    Potential Threat
    Joined Aug 2013 Posts: 66
    edited 17 Oct 2019, 9:08AM
    "I double-checked with our Engineers to make sure - if auto is used for both an attack being viewed viewed and one that's not, the AI behavior should be the same. "


    Should? Not is , should!

    This is Gibberish for: We dont know or we wont tell.
  • LilBasterd
    LilBasterd
    Force to be Reckoned With
    Joined Mar 2018 Posts: 1,871
    edited 21 Oct 2019, 1:26AM
    yes    notice the  lack of CM responses. clear have to  video record battles ourselves and then see what other crap excuse they  manifest
    Sealcorps is my biggest fan, ty Sealcorps  for always reading what i write. :)
  • Ka Dosh
    Ka Dosh
    Minor Nuisance
    Joined Sep 2015 Posts: 130

    Recently, CM Daddy Murphy posted that the game is done completely in their servers.. So what happens in Las Vegas... I think I'm like Angol Mois from Sargent Keroro! A la orden!

Sign In or Register to comment.