AI cheating

[alsoDavo]

I'm really sick of the AI cheating. Has anyone noticed a pattern? I think you automatically lose any 3rd fight in a single round but not sure of this. It seems ridiculuous to me when an elite medieval infantry 4,2,1 is dropped when attacking a 1,2,1 defense unit on grassland. If anyone knows the pattern I'll do some cheating of my own.

 

[Strider]

The AI does not cheat in unit vs unit combat. It's a proven fact.

 

[Chris1111]

The only cheating that is going on is the wonderful RNG.

If it's really frustrating just play with preserve random seed off and reload the BS losses. You'll save yourself the heatburn.

Unfortunately with it off you won't have any proof when that godly spearmen beats your tank and you come here to post about it.

 

[alsoDavo]

Care to share that proof? If I were quick enough with screen shots, I could show you some prime examples to back up my argument.

 

[KaNick]

Originally posted by alsoDavo
Care to share that proof? If I were quick enough with screen shots, I could show you some prime examples to back up my argument.

. The programs made a RNG. That's all the proof you should need. But if you want proog play a few more games and it should be clear to you.

 

[binyo66]

Originally posted by alsoDavo
Care to share that proof? If I were quick enough with screen shots, I could show you some prime examples to back up my argument.

I dont think AI cheats in this case. But I would be delight to get your proof.

 

[widdowmaker]

People have made scenarios where the player has like 10000 warriors. And the enemy has 10000 warriors. And then the 2 sets of 10k go at it. At the end the calculate when a person was hit when they got a level up (from veteran to elite for example) and did the math at the end.

The RNG in this game is goooooood. It really is almost completly random!

 

[alsoDavo]

Hey I give up....sheesh whenever I'm 100 points ahead of the other civs from now on I'll just Y Y Y. Well maybe a few fortifies. Better still, I think I'll see what's on HBO.

 

[Gengis Khan]

Originally posted by alsoDavo
Care to share that proof? If I were quick enough with screen shots, I could show you some prime examples to back up my argument.

I hope you have better examples then a Med Inf losing to a forted spearman.

The AI DOESN'T cheat during combat, there has been countless trials & tests to prove it.

 

[SuperBeaverInc.]

Show me a stack of 20 Modern Armour losing to a single unfortified Warrior, then I will believe that the AI is cheating.

 

[alsoDavo]

After cooling off for a bit I realized what was happening. I had assumed a unit from another era, i.e. Med Inf would be vastly superior to a stone age unit. It looks like from the posts here and what is happening in game is that he only slightly superior. I can deal with that. Although it does seem a bit strange that stone weapons would be almost equal to iron, in point of fact a crushed skull from a stone weapon is probably just as letal as a split belly from steel and doubtless less grusome.

 

[Tatran]

In my last game I had that familiar feeling again.
Facts : 4 Egyptian towns size 1(3x) and 4(1x)
I want to sell electricity to them and guess what.
They offered 43 gold per turn!!!
They had no excess resources.So they must have
sold a tech to others.I did gave Fascism earlier
to help them in the war against America.Lincoln tried to
to take over his continent and I decided to decimate
him.I don't think those 4 towns could produce 43 gold
per turn.

 

[Greyhawk1]

Not this again....

You will find the Random Number Generator to be a stern mistress. She will punish you at the most awkward of times and reward you when you least expect it. That bad result may have you tearing your hair out but when your one Rifleman defeats an entire Knight Army single handed in some forgotten war, the Mistress will be smiling on you...

 

[alsoDavo]

Well Greyhawk I could say something about that - like isn't it coincidental the worst shape you are in the better your luck is? From now on when I get really frustrated with my units being defeated i'll just play on a tougher level.

 

[Grille]

Originally posted by alsoDavo
Well Greyhawk I could say something about that - like isn't it coincidental the worst shape you are in the better your luck is? From now on when I get really frustrated with my units being defeated i'll just play on a tougher level.

Make sure you actually know about the %-chances of a win or loss result. The A/D stats (etc) of the fighting units define these chances. The system is very simple, nothing like "an ancient unit can never beat a modern type unit because of realism" or something such. A statistical method is used and you can use a CombatCalc (look into the C&C subforum) to get an idea about win/loss chances of e.g. Medieval Inf vs Spear (btw, the normal RNG rant starts with Tank loses vs Spear). Now if you got used to the system/chances and still complain, I'll tell you that an e.g. 0.01% chance of losing a battle does not mean this rare result would never happen...

 

[alsoDavo]

I've been playing Civ since Civ I. In it you could depend on a higher level unit defeating a lower level one. Not so with Civ II. In it if you thought the result of a fight was BS you could reload and often get a dif result. From what I have tested so far, in Civ III you get exactly the same result regardless of how mnay times you reload. I don't doubt that the system is simple and the calculations straightforward. I do think those calc results are run past another algorithm(sp?) before the end result. If it were absolutely random then it looks like you would get a dif result if you reloaded the game especially from a couple turns back. At any rate I'm really going to make an effort to drop this. I like the game and Civ I was boring when the outcome was always predictable. Actually all I was trying to do was get a handle on the combat and have some relative certainty about the outcome. In fact I think I have that handle now - if you can't afford to lose a unit, don't attack.

 

[wilbill]

alsoDavo... First, go to the game setup and click "Preserve Random Seed" off. With it on, you WILL get the same results when you reload. With it off, you generate a new random number.
Second, combat results are, as you observed, a result of Attack vs. Defense combat strengths, not era. A spearman (1.2.1) isn't stone age, but bronze age and a MI attacks it at a 2:1 strength (a little less since a defender always gets a 10% bonus, less still if the spearman was fortified).

Here's a combat calculator. Play around with it a little and you'll get a feel for the kinds of results you can expect to see...


Combat Calc

 

[alsoDavo]

Very nice.
thanks Wilbill

 

[King Kruhl]

Cheating AI…try this on for size.
I like to do the initial rush of archers and clear as many Civs off of the world early in the game. A very large percentage of the time when I find a new Civ and move my archers next to their fledgling capital city there will be only a warrior shown on screen defending the city. Yet when the AI gets its next turn, the Warrior magically transforms into a spearman.

OR …(this happened to me last night.)

I was playing as the Vikings. I did my initial sweep with the archers and some help form Ancient Cavalry (those poor Zulus and Spanish), capturing a nice sized Civilization of about 20 cities total. Then I settled into the mad dash for Invention in order to start building Berserkers. I know I was running at the front of the pack in Science because at the discovery of Invention I received a Scientific Leader. So, There I sat, Zeus was churning out Cavalry for me and my cities were on the verge of giving berth to a horde of crazed berserkers. The Mayans were the only Civ close to me in size and in being so they were the target of my next phase of world domination. Just as my Berserkers were born, I established an embassy in the Mayans capital to see how difficult of a fight it was going to be. I was SHOCKED to see the Mayans had amassed not just a few Musketeers but rather a Multitude of them, roughly 10 per city. Oh ya, I also knew the Mayans had the Knights Templar but they had just completed it about 15 – 20 turns earlier, yet the number of Crusaders they had built was way beyond 20 turns worth. Within 10 turns of this discovery The Mayans not only initiated the war with me but they had effectively reduced my Civilization by 50% . I was amazed at the stack after stack of troops that marched across my once rich and growing Kingdom.

OK, can one of you “Ripleys” explain those to me??


King Kruhl

 

[wilbill]

Originally posted by King Kruhl
...there will be only a warrior shown on screen defending the city. Yet when the AI gets its next turn, the Warrior magically transforms into a spearman.

If the population of the city drops by one, it's the "scared to death" cheat - the AI pop rushes a spearman which shows up a turn before it really should because of something to do with the order of AI/Human turn segments.