Combat system issues

[PinballWizard]

I have now played CIVIII 3 times, I've won a cultural, diplomatic, and space race victory. But I won't go to war. I find I can in no way predict the outcome of any fight or battle. Example: I land 7 troops... 2 rifleman (6/6/1) and 5 cavalry (6/6/3), knowing they are going to be attacked by 8 aztec knights(4/3/2). The fight is grassland to grassland, and all troops (mine and theirs) are veteran. According to most the formulra for computing attack success rate is given the attack of the attacker over the defense of the sum of the attack of the attacker and defense of the defender. So, for a knight attacking a rifleman, (4/(4+6))=40% chance of a successful attack. The chance of overall victory will be slightly more skewed toward the defender in this case, based on probability of successive attacks being the same. However, all of my units died. Every knight that attacked was victorious except one, and the 8th knight polished him off. The odds of that happening would scare the pants off casino owners everywhere. However, this is not an isolated occurrence for me. It has happened many times. I will generally attack only when I know I will win (based on the numbers). But more often that not I lose... especially at sea (yes I am aware of the rate of fire influence). I'm tired of losing destroyers and battleships to ironclads. I'll be damned if I'm going to war when the computer has decided that on the day of my campaign, all of my units have contracted debilatating disease. Is there some horribly skewed unlucky factor I'm unaware of?

-PinballWizard

[SirSchwa]

Not necessarily an answer, but...

Speaking of odds, try this. Before you attack, save the game, if you don't like the way it turned out, just go back to the saved game and try the same exact move again, it will most likely turn out differently. If it doesn't, run away and wait until you have some nukes, the AI must be pissed!

[pauli]

a: the combat system is b0rked.

b: the ai cheats aggregiously.

i avoid combat till i have modern armor most of the time :/

[PinballWizard]

thanks for the replies

Sorry about the tone of the last post... my login to Apolyton timed out and I had to retype the whole letter!
This much I do know... the computer chooses the outcomes to all battles at the beginning of each turn. In order to change the outcome of a battle, you must either a) go back a turn or b) alter the order of your attack (i.e. instead of attacking first with a knight, use a cannon, then the knight). If you save and go back to the same turn before a fight, and do everything the same, the outcome will be the same. Firaxis did this to counter the "oh I lost a big fight, I'll go back to my save this turn and fight till I get the outcome I want"
I have no problem with this... I just want to understand why supposedly superior units die so easily to inferior ones. I would be fine with a cavalry losing to a knight every once in a while, even 40%, but I should NEVER lose every unit I have in one turn, when all of my units are superior, unless the opposition has a far greater number of units.

Thanks for the comments, folks

-PinballWizard

[simwiz2]

because combat is extremely random, it doesnt matter if its warriors vs mech inf the warrior would still have about a 40% chance

and because (i think) there is a hidden AI cheat (fanboys dont flame this plz) where the AI will decide that it CANNOT lose these units, this attack, this city, etc. And its units will become almost invincible. I've seen it happen too many times to be a coincidence. Like one knight on defense (plains, unwalled town) took out 6 of my attacking tanks