Ok, word: The AI gets cost cheats at nearly every difficulty level:

[Markgraf]

I´d called it cheating if the AI wouldn´t have to pay anything at all to upgrade its units, or if it were able to see resources before they appear (like in Civ III), but thankfully this doesn´t seem to be the case.

I have observed the AI building its first cities far from it's capitol in unlikely locations that ended up being right next to coal and oil. The latter in the midst of a desert. I believe the AI still "sees" all the resources. Hard to justify building cities in the locations I've seen other than the only available resources in the area.

[Apocalypse Cow]

It might not techincally "see" the resources, but it might know the percentages. Oil has a higher percentage of showing up in the desert than anywhere else, correct? If thats the case, maybe the AI knew that there was a (hypothetical) 25% chance of oil showing up in the desert, and settled a city near a sizable desert.

I know I've done this before, granted not early in the game, but before oil shows up, definitely. Its a gamble, but sometimes it pays off. Was the desert the only one in the area? It may go back to the AI being able to see the whole map.

[RalphTrickey]

Originally posted by -Ab-


The AI cheats because its badly made



couldnt the AI be modeled on the strategys these strong players use
i would imagine that an AI thats simple and just has bonuses on high levels requires very little playtesting compared to one that needed the game to be made and then played extensively and only then an AI created that would need considerably less "cheats" to be competative and engaging
and thats the point isnt it? how many games come out now and the first thing that gets said is "omg was this thing playtested?"
from a business point of view there is nothing to compel a developer go the long way round when they can just bang out a crap AI and then build the game up around it so it give some illusion of being able to play

eg. resources,teaming up,sharing tech,those mysterious barbs that pop up and win an unlikely victory and in the case of civ3 and to some extent 4 the whole map is constructed to hand the AI an advantage

etc etc etc

Originally posted by -Ab-


The AI cheats because its badly made

Throws down gauntlet.

Progam a better one then. Or post the rules for a better AI. Make sure to use exact programmable terms, for example, you can't say attack city x, or pillage y, you have to say attack city x with z units with a relative strength which is q times the defenders, etc.

Don't forget that the diplomacy AI also has to be 'fun', unlike the military AI which just has to win. That's a different challenge for you.

I personally think that the civ AI does an excellent job. It actually groups units together for attacks in waves instead of dribbling them over, it knows how to use boats. Both os these things are very difficult to program.

I've heard that the AI will be programmable, so why not try to write your own. Even better, the FreeCiv project would LOVE to have your help.

[nbarclay]

Developing a great AI is a whole lot easier said than done. In chess, each turn, each player makes one move. In Civ, each turn, the player can make up to tens or even on larger maps hundreds of moves (including not just units but build orders, changes in civics, and so forth). In chess, you can see exactly what the situation is at any given time. In Civ, unless the AI cheats, there are a lot of things about what the other players are doing that the AI doesn't know. In chess, there are a maximum of two players. In Civ IV, there are up to 18. In chess, there are 64 squares. In Civ, there can be thousands or even tens of thousands. In chess, the exact outcome of every move is known in advance. In Civ, some types of moves have a random element involved in their outcomes. And so on.

When you compare the games' complexity, Civ is enough more complex than chess to make Deep Blue (the computer that could beat the world's top chess players) look slower than an Apple II. Using anything resembling a brute-force approach, a Civ-type AI cannot possibly consider more than a tiny fraction of all the different combinations of things that could happen over the next several turns.

That means AIs in Civ-type games are very heavily dependent on programmers to tell them what strategies to use in different types of situations. But even if the people who provided the strategy for the programmers were the best Civ players who would ever exist, AIs would still face several problems. First, something would inevitably get lost in the translation between what goes on in the best players' thinking and what gets programmed into the AIs. Second, the AIs would still be stuck working from general rules rather than being able to invent new situation-specific strategies on the fly. Third, in the absence of a genuinely massive investment in giving the AIs a variety of alternative strategies, human players could take advantage of learning to recognize patterns in the behavior programmed into the AIs. And in any case, after the game's been out a while, there will be human players who are better at the game than anyone was at the time the AI was originally developed and who share their knowledge with others.

So with anything resembling current artificial intelligence technology, it's just plain unrealistic to expect AIs to be able to compete with the best human players with both sides playing by the same rules. We can expect AIs to come closer over time as computers get more powerful and AI programming techniques improve, but if top players want AIs that can give them a serious fight, we have no choice but to give the AIs advantages on the higher difficulty levels.

[Proteus_MST]

Originally posted by nbarclay


When I was little, my father was a lot better at chess than I was. In order to have games that were reasonably even, he would start without a queen or a rook, setting up a game that I could win if I didn't make too many mistakes but could lose if I did. I don't view that sort of thing as cheating because it was agreed to by both sides and was necessary to make for a competitive game.

Same is true for Go.
Go players are ranked according to a System of Kyus and Dans and players with a higher Kyu (or lower Dan), i.e. those who are waeker/mor inexpirienced in the game, get from the beginning on 1-9 of stones in their color which are placed at certain locations on the Go-Board, to give the stronger player a handicap and increase the chances of the weaker player to win the game.

(btw. in contrary to chess, where it was able to develop AIs which have a chance to beat players with master level, the best AI for Go is only able to beat players up to maybe Kyu 12 [with Kyu being pupils levels and Dans being Master/Professional levels] although [or maybe rather, because] Go has much less rules than chess)

[Sirian]

Originally posted by nbarclay
My only real objection to the advantages AIs get is that I think they should be documented more clearly.

The XML is easy to read. The files are intuitively labeled in most cases, including this one. You open the file and you look inside it. How much more clear are things supposed to be?

- Sirian

[Kassiopeia]

Even if XML is easy to read, not all are computer savvy enough to go trudging around in the file directory of the game - some may just lack the courage to do that, as odd as it sounds, for fear of breaking something. They resort to the manual for information.

[dearmad]

Originally posted by Kuciwalker


I take it you don't read /.

Huh? I take you just can't think and observe then... because the inflation is clearly listed in the economy advisor.


Moving on:
For those of you getting all pissy that some people are disatisfied with the AI and who just make excuses for developers- go play Imperialism 2. The strategy is complex. The "AI" gets it. It plays the game. It's tough to beat even when it is not cheating.

You can program AI to play the game. It's not impossible.

[Skanky Burns]

/me laughs at the cute n00bs who think that writing an AI is easy.
Amusingly, some couldn't even be bothered making a complete criticism of the current AI yet were complaining of others taking shortcuts.

[Senethro]

Theres no way we can have a AI that is good at Civ any time soon, according to my SMAC techtree. We haven't even discovered Cyberethics or Advanced Military Algorithms!

[nbarclay]

Originally posted by Sirian

The XML is easy to read. The files are intuitively labeled in most cases, including this one. You open the file and you look inside it. How much more clear are things supposed to be?

- Sirian

From your perspective, I imagine hunting down information in XML files seems pretty easy, but consider the situation from the perspective of a gamer who isn't all that computer literate and who hasn't discovered resources such as Apolyton to point him or her to the XML files. Further, even once a person locates the correct XML file, the format is extremely clumsy for comparing difficulty levels with each other, and quite possibly a bit daunting for the uninitiated to try to read.

What is really needed is (1) a list of the advantages AIs get on all levels, (2) a table showing how the bonuses that vary by difficulty level differ where it is practical to present them in tabular form, and (3) a description of how things like hut outcomes and starting techs that aren't so easy to fit in a table compare. It would have been nice if Firaxis had done that so all players can have the information readily accessible whether they visit places like Apolyton or not. But since you haven't, I guess someone from the player community will need to do it. (I'm not volunteering at present, but I might do it at some point if no one else beats me to it.)

[Markgraf]

Was the desert the only one in the area? It may go back to the AI being able to see the whole map.

It was very conspicuous. It was a large desert. And the Roman AI went out of its way to place a city far from its own cities and only two squares away from a Spanish city, squeezed between the Spanish border and my [French] border. Really up in the grill. Also, there was a single non-desert/non-controlled square that the AI could develop. Oil appeared right next to the Roman city later on, after it was "converted" to Spain via culture (my own culture was so close to this that it "captured" the oil by going right up against the former Roman city). Such a bad placement -- with plenty of better sites available from a kingdom standpoint -- made it crystal clear that this was exactly the "resource siting" that occurred in Civ3.

AI City placement by seeing the whole map is one of my most hated cheats. I think it's been toned down a tiny bit, but it's still there. For sure.

[Electrius]

Originally posted by -Ab-


The AI cheats because its badly made

Spoken like somebody who's never tried to program an AI in their life. It's hard. That it works at the level it does and is responsive and adaptive speaks monuments to AI design of Civ4.

[baboon]

I'm fine with the AI cheating against Barbs.

Just as long as it does well against the players and AI players, who cares about how they do against barbs.

Barbs are for humans to get pissed and to have a lot of extra fights and some free cities

[Tiberius]

Originally posted by baboon
Barbs are for humans to get pissed and to have a lot of extra fights and some free cities

Speaking of humans and barbs: is 5 the maximum experience points a (human) unit can achieve against wild animals? I had a warrior who won at least 4-5 battles against wolves and lions after he reached 5 experience points and still remained on the same level. Later on he killed some barbs and his exp. level went up again.

Is this a known thing? Sorry, but I haven't heard/read about it.