Okay, the word “cheats” is a bit strong, and maybe giving them a production bonus is there to balance out some other deficit they have against an actual human being. BUT, if this is how the game works then the manual shouldn’t state that the AI uses the same rules as you on Noble level. This is just a lie and can mislead you into trying a banzai early attack strategy for example, because you think you have a chance of getting in before they build a warrior, whereas in actual fact you have no chance because of the secret head start they get in building them.
The combat percentages are not correct. The day I win 5 out of 10 combats at 50% each is the day I eat a cherry sno-cone in hell.
I’ve played Civ IV more than any game other than EU2, and I’ve never had a quibble about the percentages for winning. I win or lose in line with the percentages (often I’ll “cheat” and replay over and over to see what happens 
). And I’m much happier with the way combat is resolved than I was with the system used in SMAC, where as long as you were ahead by any appreciable margin, you were guaranteed a win.
And in Civ IV, the weights given to the units to make the odds more realistic seem to work well. When a tank unit loses to a drastically inferior unit, it’s more than likely because the hit points were lowered substantially before it attempted the attack; think not a Panzer division at its best, but rather a Panzer division with most of the tanks barely able to move, running out of fuel, with the soldiers manning the tanks mostly dead or injured. 
In every single version of Civ, a recurring complaint has always been that low-tech units like spearmen occasionally beat tanks. But there’s two things to keep in mind: First, remember that something that happens once per game in Civ is equivalent to something happening in the real world once in all of history, and something that happens once every ten games or so is something that would probably never occur in the real world. Second, remember that there have actually been a handful of very rare cases in history where low-tech soldiers have indeed defeated high-tech. Given that it’s happened multiple times in real history, but only occurs once per multiple games in the game, the real problem is that the game gives too low a probability for a spearman to defeat a tank.
I’ll have to differ here with a couple of the posters. I regularly have uninjured units (riflemen) get beaten by things such as crossbowmen or macemen where there is no substantial defensive bonus in favour of the weaker unit. And in sea battles the AI frigates always seem to have an advantage.
Further, how often later in the camp do you find an AI city being built on a crappy piece of terrain which later on turns out to containg that hard to find resource such as oil or aluminium? (Even when no one yet has the technology to find or see them?)
Why would you be surprised by the rifleman losing to the maceman? The rifleman is something like a strength 14 unit, and the maceman is a strength 6 or 8 unit (I can’t remember and am not gonna bother loading up the game to find out). That’s just a 2 to 1 advantage, meaning a 67% chance, if they have the same hit points, and no other multipliers.
You, too, can accomplish this by simply settling the blue circles off in the wastelands.
There are LOTS of cases of low tech armies beating high tech ones.
The game’s units have to be taken as abstraction, to some extent. If you think about the scope of the game, a “maceman” unit isn’t three guys with maces, it’s tens of thousands of soldiers - roughly going by the size of a really large army fielded by a powerful Civ, one unit is equivalent to at least a corps, two or three divisions. And it wouldn’t be 30,000 macemen; it’d be a corps dominated by macemen but also including supporting arms, just as a U.S. "“Infantry” division today includes not just infantry, but artillery, army aviation, anti-air units, and so on. Such an enormous force doesn’t normally get wiped out in a single battle; a co, or how a Roman legion was mostly comprised of legionary swordsmen but would have been supported by archers, ballista, and cavalry. A contest between two armies fielding multiple corps is a full scale war that would take years to complete. You would have to assume that the army fielding Macemen is going to attempt to find new tactics and ways of dealing with the army fielding Riflemen - better tactics, concealment, stealing rifles and using them (as the Zulu did against the British, for example.) It might usually fail, but sometimes they’ll succeed.
For all the complaints, I’ve played all four Civs and the army with the better tech had a pretty gigantic advantage. In Civ 4, being possessed of tanks or gunships if the enemy hasn’t got the right defensive unit usually results in the enemy suffering horrendous losses. It’s not a certainty, and it shouldn’t be - or else the only point to the game would be to be the first player to build a tank unit.
RickJay, I wouldn’t mind if it happened that my inferior units won every now and then but that so rarely happens 
I have Civ 4, Civ 4 Warlords and Civ 4 Beyond the Sword all installed.
Aside from the expansion specific campaigns that come with them are there any fundamental differences between loading the 3 versions? If I were to load the original Civ 4 and play it as America and then load Warlords and play as America would I need to make any strategy adjustments? Would any of the apparent inequities change?
I don’t care what anyone says, the combat percentages are a LIE and/or the computer cheats. I fight almost EVERY battle with a sizable technology advantage and beneficial combat percentages and I lose WAY more units than the AI does over the balance of a game, even on Noble.
It rarely happens the other way, too. I’ve had lots of inferior units win battles. They usually lose, but sometimes they win.
I know the game can be frustrating, but it’s observation bias. In a controlled test, the odds would always add up in the long run.
I still find it difficult to believe. In one on one combat, my experienced units are rarely a match for the AI’s.
However in other areas I don’t think the AI cheats. I have gone to islands that have started with no Civ but the AI has started cities on and there are still untouched “goody huts”. They may only contain piddling amounts of gold, but if there had been cheating it would have made sense for the AI to have entered them.
I find combat really, really frustrating, to the point that I usually cheat and turn it off. Even if it’s statistically fair, it feels as if the AI is far, far too good at combat. :smack:
It’s not that bad, you just have to accept that in war you’ll lose lots of units. It’s okay because the AI will lose a lot more.
This issue gets debated extensively on the CivFanatics forum. A 98% chance of victory equals a 2% chance of defeat, so you should expect to lose 1 of 50 battles with those odds. A typical game can include over 1,000 separate battles, so that figure adds up.
It basically boils down to a form of confirmation bias. We all tend to remember the battles we lose despite high odds for success, and we forget about all the 30-40% battles where we emerge victorious.
I don’t think the AI cheats. I also don’t think it is very good at warfare.
Generally, Civ 4 wars tend to go like this:
-
declare war
-
take a single enemy city
-
wait for the enemies’ big stack 'o troops to try taking it back
-
smash the stack o’ troops with the appropriate counter-troops waiting in reserve
-
heal up your troops
-
take enemy cities one at a time
The big mistake, I find, is attempting step 6 before you have smashed the enemies’ big concentrations. That allows the enemy to break up your invasion force.
The Civ AI tends to not concentrate its forces very well. It tries to defend everything rather too much. Once you have smashed up its field army, it tends to be easy meat.
Also, it tends to scatter seige troops all over the place to no purpose, rather than using catapults/cannon/bombers to reduce fortification values and weaken units before moving in for the kill. A big stack 'o bombers, and you basically can’t lose …
Aside from warfare, it seems that the AI doesn’t suffer from too much expansion (talking Beyond the Sword).
Be that as it may, my copy has now been locked in my broken computer for two weeks. I fear I will need to buy another copy- withdrawal symptoms.
I don’t have 30-40% battles anymore. I don’t even bother. I’ve resigned myself to the computer having its way with me during its turn, and I don’t even get mad at the stupid shit it does. But when the comp fucks up MY turn by cheating, I get steamed. Isn’t it enough that the comp gets to have free combat cheating during it’s turn? Can’t I at least have DEPENDABLE cheating? I don’t even care that the computer cheats, I only wish it showed the real odds adjusted for cheating.
That was exactly my complaint. For some reason (Shares in the company? Personal leather sex slave of Sid Meier?) certain people refuse to acknowledge that the AI in Civ IV cheats at combat. I know for an absolute certainty it does, because I have experienced situations where I supposedly have a 90%+ chance of victory, and then will lose, say, eight combats in a row. You do the math. You’ve got a better chance of winning the lottery or being struck by lightning where you sit – yet it happens regularly, often more than once in a game. I only need to “win the lottery” once to have a pretty good clue the AI is cheating, either by giving me false combat odds or by giving itself a secret combat bonus.
No, it doesn’t, but that’s ok, Smash, you continue to compensate for your poor ability at the game by blaming the program.
Or you could read carefully the analysis given by Normal Phase and see if it helps.
(Just checking: you do know that if you are running the same battle on re-load from save that the same result will occur EVERY time, unless you’ve dechecked the box at the very beginning of the game during set up that remembers the seed number, right?)
LOL! That’s probably it!