Announcement

Collapse
No announcement yet.

Combat.... say WHAT!?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Combat.... say WHAT!?

    Just reviving an old whine.

    Situation:
    My axeman vs AI archer
    98,5% odds in my favor. Die a steely death weakling!
    Archer wins. wtf¤%#¤¤ etc etc etc, but again, the odds were not ALL in my favor.

    Later same situation, same game. My axeman vs enemy archer: 98.4% odds in my favor. Archer wins!

    Now, if this had happened as it should 2.5 times every 100, then I wouldnt really mind. In this case 2 out of 10. Are my math skills that bad? There seems to be something askew with the rolls when I play against the AI. I lose battles that I should win. A lot. Anyone else stuck with the same feeling? That the roll will favor the AI far more often than the other way around? Is this adjusted with difficulty level or something?
    Diplogamer formerly known as LzPrst

  • #2


    You got unlucky. Suck it up. It's been shown the AI doesn't receive a bonus in combat.
    You just wasted six ... no, seven ... seconds of your life reading this sentence.

    Comment


    • #3
      You remember the bad stuff. That's one thing. You probably won battles you should have lost, too, but they fly from memory because they don't result in your losing a city, etc. If you're not convinced, you could test this, say, 100 times with the "reseed" option properly set.
      I've been on these boards for a long time and I still don't know what to think when it comes to you -- FrantzX, December 21, 2001

      "Yin": Your friendly, neighborhood negative cosmic force.

      Comment


      • #4
        I've been blessed with relatively few 95+% losses, but in this last game I've had several -- vs. barbarians yet (noble/huge/marathon/Terra, Better AI, barbs can build cannons & several culture-producing buildings)!

        My first Better AI game. I think I'll end up rescinding those barb cannons.

        Comment


        • #5
          It happens. I lost three units in a row with a greater than 90% chance of winning - playing as the Celts on Earth. Since I had only two cities in the medieval age, my not-so-large army was decimated.

          If you want to see some of the victories you shouldn't have won, look at the Combat log. You'll see it evens out over time. I've won when the civ attacking has had 99.7% chances of winning... and one when it was >99.9!

          Comment


          • #6
            I played the Omens scenario a few days ago. The odds were often off. There were generally clear odds in favour of myself, or of the enemy. Yet, many fights looked like a walk in the park for the winning army (not necessarily the enemy's: I benefited from it too)
            "I have been reading up on the universe and have come to the conclusion that the universe is a good thing." -- Dissident
            "I never had the need to have a boner." -- Dissident
            "I have never cut off my penis when I was upset over a girl." -- Dis

            Comment


            • #7
              2 out of 10 means nothing.

              In theoy, you could lose 10 out of 10 battles at those odds. It would just be a sheer run of bad luck.

              Or you could win 100 out of 100 battles at those percentages.

              It's all a matter of weighted luck.

              Comment


              • #8
                I once lost 2 in a row that showed 100.0%

                Needless to say, I was a bit upset, but what can you do?

                Comment


                • #9
                  I was playing this morning. Lost a 15 xp War Elephant against a badly wounded Longbow - there was a 2% chance of that.

                  OTOH, I had a catapult beat a fortified longbow in a city in the same game (the odds were something like 3% of that happening).

                  So... sh!t happens.
                  "I'm a guy - I take everything seriously except other people's emotions"

                  "Never play cards with any man named 'Doc'. Never eat at any place called 'Mom's'. And never, ever...sleep with anyone whose troubles are worse than your own." - Nelson Algren
                  "A single death is a tragedy, a million deaths is a statistic." - Joseph Stalin (attr.)

                  Comment


                  • #10
                    I throw a lot of units against terrible odds, and I remember when I achieve a miraculous victory.

                    For example, when I'm on an island alone, and all the AIs gang up on me (let's be honest, the AIs rarely war against each other - they represent a united force against the player) and stacks of doom regularly land on my shores, I throw my cannons against them not expecting to win - just expecting to cause collateral damage. Maybe 1 out of 200 such encounters I will pull off a victory with 10-30% chance. Never, ever, ever, ever have I won with a 1-2% chance. ****ING NEVER!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

                    Yet when I'm on a continent with other nations, and I'm forced to become a war monger because the game is unwinnable as a peacefual builder (except on easy difficulties) I am in the position of throwing very powerful units against AI weaklings. In such situations, the AI pulls of 1-2% odds about once every ten battles.

                    So let's imagine the odds aren't rigged - what are the real odds factoring in my chronic 'bad luck':

                    Me: 98% chance of victory = 40-60% chance of victory
                    Me: 50% chance of victory = 25% chance of victory
                    Me: 10% chance of victory = 000000.1% chance of victory
                    Me: 1-2% chance of victory =

                    AI: 98% chance of victory = 100% chance of victory
                    AI: 50% chance of victory = 100% chance of victory
                    AI: 10% chance of victory = 98% chance of victory
                    AI: 1-2% chance of victory = 10%

                    Sure, the odds aren't rigged, but it would be nice if the odds could factor in my chronic 'bad luck'.
                    Voluntary Human Extinction Movement http://www.vhemt.org/

                    Comment


                    • #11
                      Ten battles, each 98.5 or 98.4% in your favor. I'll just average that at 98.45% if that's okay.

                      Chances of you winning all ten battles: .9845^10= roughly 85.5376%.
                      Chances of you winning nine of the ten battles and losing one: 10*.9845^9*.0155= roughly 13.4671%
                      Combined together, that's a 99.0047% chance of winning at least nine of the ten battles.

                      That means there's also a 0.9953% chance of losing at least two battles. It's definitely low, but not entirely far-fetched. Weird spells of bad luck like that happen every once in a while.

                      Now if you lost five or more of those ten battles, the chances drop below one in 4733153, by which point I'd think something weird might be going on.

                      Like losing two 100.0% battles in a row...
                      Known in most other places as Anon Zytose.
                      +3 Research, +2 Efficiency, -1 Growth, -2 Industry, -2 Support.
                      http://anonzytose.deviantart.com/

                      Comment


                      • #12
                        It's all a matter of weighted luck
                        more like weighted dice...
                        Diplogamer formerly known as LzPrst

                        Comment


                        • #13
                          my experiences seem to mirror bkeelas....
                          Diplogamer formerly known as LzPrst

                          Comment


                          • #14
                            The combat odds thinge (if I remeber right) does not count first strikes.
                            “...This means GCA won 7 battles against our units, had Horsemen retreat from 2 battles against NMs, and lost 0 battles.” --Jon Shafer 1st ISDG

                            Comment


                            • #15
                              Yes, it does count first strikes.
                              LandMasses Version 3 Now Available since 18/05/2008.

                              Comment

                              Working...
                              X