What is "Cheating"??

[Lt. 'Killer' M.]

TheNiceOne: the teleporting rules are different for the AI! They can teleport to the nearest square no matter what, but human troops cannot teleport to a square the AI has a unit going to (with the AI unit-go-to-command-equivalent).

 

[Zachriel]

Originally posted by cracker
In an empire of 500 civilians, knowing that citizen number 327 needs to stop producing food for a turn so that her city does not grow by another person because there will not possibly be enough food to feed that extra person, thats what computer AI are supposed to do without even hesitating.

You could try playing this strategy game like a human, rather than a computer. Delegate those simple and tedius tasks to the governors and concentrate on strategy. I believe that if you treat the map as a bunch of square blocks, and the game as a way of exploiting those digital constructs, the game loses all meaning and fun.

 

[alver]

Zachriel: I think Cracker meant that was a task the AI could be expected to do quite well, as opposed to a human who wouldn't bother for the reason you just stated. So I think you're in agreeement

 

[cracker]

Alver, you are spot on.

and Zach, you and I are actually in the same play mode.

The computer should always be better than me at things like micromanaging civilians or adaptations of the pathfinder algorithm that say "move-irrigate-move-mine" in one circumstance and then "move-road-irrigate-move-road" in another. The ability to evaluate the future impact of a 6 to the 21st power possible permutations of move sequences and then choose the right sequence is what computers do well.

The argument about humans having superior memory versus the AI is also one argument that I usually think is backwards. My computer has so much memory that it can easily remember everything that I have ever known and perhaps even forgotten already in my life. Virtual memory means the AI can create internal records that provide it with instant access to all the unit positions and possible unit positions for every unit it has ever detected. Humans can do this too, but only at the price of time, resources, extra program support, and a potential shift in focus.

Computers figure out the nuts-bolts and ins-outs of tactical game play in a flawless sequence of micromanaged steps if they are programmed well and they do this in nanoseconds while their human players must take lots of time to accomplish the same objective.

Some of this is just algorithm depth in terms of what tools the AI has available to use in its decision processes.

AI programming is the toughest part of advanced gameplay design, so I am not trivializing this task in any way. Sometimes the way AI programs function is driven by the way the "father" views the objectives of the game play process. When we see a galley full of something that is being escorted across the map at 3 steps per turn by a battleship, it reflects on the depth of understanding of the game decisions that was present in the creators and on the time resources that were available to implement and test decision algorithms.

 

[sumthinelse]

I think this is correct: If the AI wants to kill a worker, since you will probably capture it back immediately, it can do it, but the human can only capture workers. (That's why your workers disappear rom the ma[ when the AI attacks them sometimes). I think this means that the AI can get a Great Leader by killing workers while the human cannot (except in the following situation)....

I know this sounds crazy but etj4Eagle had the bug after a nuke exchange where his combat units attacked his workers, and he got a great leader from killing his own worker!!!

The AI can probably get promotions and great leaders from killing workers at any time, but the human can only do it during the "nuke bug" phenomenon and only (AFAIK) from killing his own workers.

How could I test this theory? It would be interesting if Firaxis would let us play a human civ but we would be allowed to do the same things the AI does....

 

[Zachriel]

I'm always agreeable, especially when you agree with me.
. . . .
Anyway, just tell me what end to stuff the shell into and let me at 'em. (You'd better stand back a bit.)

 

[Lt. 'Killer' M.]

Originally posted by sumthinelse
I think this is correct: If the AI wants to kill a worker, since you will probably capture it back immediately, it can do it,

You sure???? I think they don't kill them, they capture and disband them

 

[etj4Eagle]

Originally posted by Lt. 'Killer' M.


You sure???? I think they don't kill them, they capture and disband them

That is my belief too. Also when an AI "kills" a worker of mine in capturing it, it does not go through the combat sequence. My mech infantry fire of a few rounds and then the worker dies under this situation. I forget if the workers have 1 or more hit points. But it definetly is combat.

 

[cracker]

I still prefer the term "Flying Monkeys" to the overused AI cheating rant.

For the most part, the AI can't cheat in the strictest since of the word because it just plays by a different set of rules that we haven't necessarily been able to figure out yet.

After I have seen Flying Monkeys, Flying Lemurs, Flying Orangutans, Flying Gorillas and a Flying Howler or two, all in the same game, then I start to think I am playing some other game besides Civilization, but then again I do not play with the perspective of being the Monkey king or Mrs Gulch. Perhaps for them, Flying Monkeys are a reasonable part of any civilization.

 

[zeeter]

My god. I've just sat through reading three pages of this drivel and am spent. I cannot believe that so many people are so uptight about "cheating". What are you so upset about? The AI zigged when you thought it would zag? Occasionally that spearman makes a run at immortality when its the best that the AI can offer to defend against your modern armor? The galley that somehow makes it through the high seas to your shores and drops a settler on your one and only open coastland square?

Why can't we get past this? It seems to me that players get their strategies together, then get angry because the AI doesn't perform according to the "rules". You know, I haven't read exactly through the manual, but as far as I can tell the rules there are for the player, not for the AI. The "cheats" make the game more interesting. Guess you should have broken that spearman down with a few artillery blasts before sending half your army at it. How stupid are you who sends ten tanks at the biblically superior spearman anyway? Send one and if it dies, you know you're in over your head.

Maybe the AI programmers realized that on a map where the two main land masses are over five squares away you won't contact anyone until the late middle ages. How boring is that? Lets lower the sink percentage for the AI in these situations to make for better gameplay. We can't do it for the player, because then they'll know how far away the nearest landmass is without actually looking. We can do it for the AI, because after making this adjustment we'll just remove that information from the AI Civ's "memory".

And while we're at it, when the player is in the late modern era and goes to war against an ancient era civ it will be too easy. Lets make some of the ancient civ's units a little tougher to make a game out of it.

While we're talking about "cheats" here, lets not forget the biggest "cheat" of all. The player gets to do whatever he wants in the game, while the AI only gets to do what its programmed to do. Sure, maybe a patch will even it up a bit here and there, but how long does it take to overcome this - "Whoa! Never saw that before. Guess I'll have to do this from now on." Does the AI get to say this?

 

[sumthinelse]

Originally posted by zeeter
My god. I've just sat through reading three pages of this drivel and am spent....How stupid are you who sends ten tanks ....

I think if you read carefully you will see quite a few well-written thoughts in this thread, some defending features of CIV3. Do you think that if a topic doesn't interest you that nobody else should think or write about it? A little respect for others would not hurt here.

 

[cracker]

Zach,

Its easy to agree with you since you have adopted my way of thinking.

 

[sumthinelse]

Originally posted by etj4Eagle


That is my belief too. Also when an AI "kills" a worker of mine in capturing it, it does not go through the combat sequence. My mech infantry fire of a few rounds and then the worker dies under this situation. I forget if the workers have 1 or more hit points. But it definetly is combat.

Ok, I guess I was wrong about that. If we had a multiplayer game and a human opponent captured and disbanded the worker, it would look exactly like what I thought was "killing" the worker via combat, so no difference between human/AI.

Thanks guys for noticing my error. I'll be updating the leader guide with the new info you found, though.

 

[JFL_Dragon]

AI Cheating

I dont mind that the AI has an advantage, as it makes it more of a challenge. But Im with the majority here, in that unknown advantages are really not necessary.

I dont understand why they make it so complicated ? Just by adjusting AI trade rates, tech disc rate and build rate is really all they need to do to make the game more difficult. Not all this other crap.

Im sure if diety was 50% research and build rate. And the AI trade rate was the same this would more compensate than all the free units and myterious settlers appearing.

Cheers

JFL_Dragon

 

[Grey Fox]

Originally posted by sumthinelse
I think this is correct: If the AI wants to kill a worker, since you will probably capture it back immediately, it can do it, but the human can only capture workers. (That's why your workers disappear rom the ma[ when the AI attacks them sometimes). I think this means that the AI can get a Great Leader by killing workers while the human cannot (except in the following situation)....

I think it's rather that they capture the worker and then disband it, just like we can do.