Does AI always know the location of Iron?

[r_rolo1]

Late to the party, but as no one awsered to this in detail ( a shame, since the post is actually well contructed and has some good points ):

-- Most people wouldn't mind higher levels giving the AI bonuses, but this kind of bonus whoring must not look that obvious to the player. Example: The city expands to 10 cities by 1500 BC or rushes with bronze units @ 2500 BC; in these cases it feels like the AI is playing an entirely different game. A AI with bonuses should ideally always feel a little ahead of the player and not deliver killshots at impossible timings.

Civ IV does that pretty well ... until Deity and this even on BtS ,where they downgraded the bonuses hard. IMHO Deity bonuses, especially the starting units, break the assumptions behind the devs idea so badly that it fits in the "AI is playing other game" .Note, I'm not saying that Deity is not winnable, I'm just saying that when you get to Deity any remainging ilusion that you can win by not leeching the AI in some way ( in other words , tranfering part of the AI bonuses to you ).

There are two good examples of that:

1 - The fact that AI has 2 settlers in Deity makes that in normal games barb animals do not appear ( in Deity quick, the situation is that the barb animals will hae exactly 1 turn to spawn before starting to be replaced by regular barbs ) and it also makes the barb date to enter borders go down the drain. I'm pretty sure this is not intended behavior

2- The fact that the AI starts with a of archers makes that it can start wars far earlier. Note, not because of power ratings or such, but simply because the AI will only DOW of their own if they can get a stack of atleast 4 units , being one of them a unit capable of city attack. This means that in deity, the AI is normally 1/2 warriors away of being ready to start a war ( and no, there is no grace period on this , unlike so much people still say ). If a AI discovers you soon enough and it is close enough, you are in high danger.

-- The AI must practice the basics. Basic defense, expansion, economy, and army. Obviously it won't be the most efficiently done, but I better not see trash like one AI expanding to 4 cities and never teching.

The civ IV normally does that decently ( unless flavouring enters in the show ... more on that below ) ... except that small and irking coding error that creates the cottage-> something-> cottage cicles that everyone that plays this game already seen ( basically the original code does not evaluate the value of cottage ->something transformation as the inverse value of something-> cottage ... it is a small error, but with big consequences ). You can see it clearly when you play against a AI that is far less constricted by roleplay ( say, Cathy ). I do not think that, besides the above point, that civ IV is badly served ... could had been better done , OFC

-- Flavor is fine, but being too extreme creates joke AIs that ruin the game by softplaying. Reluctant to tech trade civs can remain that way, but in Civ IV it's comical how the AIs cripple themselves by trading so little. It's stupid.

The best way to promote flavor is to use a strong base build, and use a dice roll to pick several branches that have a chance of being viable and will see occasional use. But the foundation must be there to begin with. Civ IV AIs simply does not have this.

Yup, the devs unfortunately gave roleplay too much of leeway in the AI actions. Better said, the roleplay enters where it shouldn't and it is too inelastic to be a good thing. A monty that would think twice before bashing again against a foe that beated him conclusively 10 turns ago in spite of having a smaller army ( this is because of the power rating being a too much simplistic measure for what it is tortured to do in the current AI code ... just for starters it does not count with promos and completely disregards the fact that a 20 srt unit will eat for breakfast 2 16 str units most of the times , regardless of the power ratings ) or a Toku that actually sometimes teched with someone besides some completely unprobable conjuction of diplo modifiers and map would be nice

-- More variety. The AI is disturbingly uniform when it comes to tech choices with a few variations. This means that certain tech paths just completely blow because everyone runs towards it? Why?

Yeah, basically. In fact the only parts of the game where there is a genuine group of good tech path alternatives are the beginning, the rifles age and the Modern eras. The classic and the medieval eras are pretty dry on that ... and the best path is actually the one that the AI usually follows , the one that humans normally shun due to tech trade oportunities Going through the bottom side of the tech tree during the medieval ages will give you a good bunch of economic boosts, a good group of units both to defend and to attack ( hence why humans hate medieval warfare vs the AI ), some nice civics and a couple of decent buildings. Compared to that, going anywhere else is weak unless you can buy the other techs

-- It must try to win. I understand some want to be more peaceful, but there must be higher chance to build military and go to war for some of the peaceniks. As it stands, the peaceniks are utterly worthless and often become nonfactors. This doesn't mean it should refuse to vote for you in the UN regardless of stance, but come on! Someone that hates you and sees you're about to launch a spaceship should indeed have a reason to attack you.

Well, barring that I really don't agree with you on the "beat the guy that is close to win " since it can lead to unfun and/or stupid AI behaviour if not very well thought, it would not hurt to make the AI leaders in general to perceie that in this game land is power and that agression can really pay up .... better said, besides not having scripted strats for most of the possible VC ( most likely due to the September release of BtS , instead of the more logical and common Christmas relase, that cut a lot of the BtS AI development ( there is a lot of AI behaiour that is the code equivalent of a abandoned construction yard and the VC strats are one of them ) ), the AI has a very poor perception of power balance and cost/reward ratios for most of it's actions ( want a example ? Do you know that basic checks for a AI to peaceassal to someone are that AI power and that AI diplo towards the possible master ? There is not even a threat asessement or , for heavens sake, even a check, by the part of the AI master to be if this new vassal will not bring him to wars he would not declare by him self or to a undesired diplo situation.... ). That alone is enough to derail the AI unless, by acident, they enter in a vassal chain reaction or something similar ...

Having AIs that implode, or just provide extremely easy picking, or even worse peacevassal whores is really not challenging or fun. It's just annoying.

Now that is probably the big issue on civ IV AI performance. Like I said above, IMHO the leaders that tend to perform better ( this without considering traits and UU/UB ) are the ones that have high flavour to expand or the ones that are less restricted by roleplay issues. OTOH the ones that shoot on their feet quite often ( Pericles, Wang Kon ( I'm playing a AW SG with raging barbs where Wang in all of it's wisdom, decided to found budhism + Judaism before going for city #2 or aparently before getting BW .... result : the barbs now have a double holy city and Wang is no more ... ) ... ) are leaders that focus in either religion or research too much for their own good. Other thing that happens often is a AI that, due to it's high "shiny" ( wonder addict meter ) factor tries to make a wonder before having enough realistical conditions for it ( that is before having workers, worker techs or a second city that can continue the empire expansion in decent time ) and ends losing the wonder, not making setters in due time or not improving enough land to fund expansion in a realistical time frame ... or ( what happens to Wang from time to time ) it gets so fixated in getting some building that will not resume expansion or even worker building before ending that building ( wang definitely loes his libraries and unis too mcu for his own good ) ... or hire too much specs and ends not working enough land to be competitive ( Phi leaders can fall easily in that trap, especially if they found a religion ) ... or good ol'Toku.

In resume, the AI in this game roleplays too much for it's own good.