OT - Strategy games with GOOD AI

[LDiCesare]

I far preferred galciv over galciv II. The ai's about as clever in both games but galciv is somewhat simpler and the ai seemed to handle the game better.
c-Evo is very easy to beat imo.

[Nightblade]

Master of Orion (the 1st one) has a very decent AI, i enjoy a lot playing from time to time and it always gives me a good challenge and does not seem to cheat in any ways.
playable under Dosbox for modern computers without any problems.

I remember reading that the AI in the RTS "Battle Realms" was too strong in the unpatched game and they toned it down a lot in the patches. I have no idea how it was, having only played the standalone expansion "Winter of the Wolf" that include an already patched Battle Realms, i would really have liked to try.
Winter of the wolf had a challenging skirmish AI though.

An immense disapointment regarding the skirmish AI for me was with Red Alert 2, it was so bad that even at the highest difficulty, it was actually very difficult to not cripple or destroy it.

[LDiCesare]

Originally Posted by Nightblade

Master of Orion (the 1st one) has a very decent AI,

Strategic ai that is. The tactical battle ai was utterly moronic.

[Sombre]

I just want to add I am in no way suggesting anyone plays galciv II - I had a similar experience to calahan - it seems like a good game but I just don't enjoy it. At all. It feels like running a spreadsheet.

Nor am I saying the AI is great. But it's better than any other PC strategy game AI I've seen.

[vladikus]

It's real time strategy so I may be digressing, but Empire Earth had some awfully challenging AI. Although, it was probably too hard.

[analytic_kernel]

I have to agree with those who are recommending computer chess, even if it was tongue-in-cheek. As Agema has mentioned, games with simpler sets of rules can generally get deeper-thinking AIs programmed for them.

I've actually had some experience at working on improving AIs for open source TBS games. In one case it was for an existing TBS game engine that could run different kinds of games, depending on the game definition files you fed to it. I was younger then: more ambitious and less reflective. As I got further and further into that project, the full enormity of the task became apparent. It is truly a non-trivial undertaking to make an AI that plays well in a variety of circumstances when a lot of variables are in play.

I also started to develop a Python AI for Wesnoth about a year-and-a-half ago, but became frustrated by the limitations of the sandbox environment they had set up (for security reasons). Someone else actually had put together a Python AI that was a decent tactician, but it was computationally expensive and there was no "strategic level" thinking. I took some of his ideas and worked on one that performed some strategic analysis and kept coherent goals across multiple turns. It was doable, but I don't think it was a huge improvement, because there are so many situations where one kind of strategic analysis doesn't work well. (Closeness of starting positions, victory conditions, etc... can all force significantly different approaches to the game.)

So, I think you can sink a lot of thought and a lot of code into making a good AI, but after a certain point there are diminishing returns. "Use cases" that someone cares about do slip through the cracks. My guess is that professional game development shops realize this, determine a threshold that they deem good enough, and then don't expend much effort beyond that point.

As mentioned in an earlier post, humans are generally more challenging intelligences to play against. If I could go back in time and change some of my participation in the TBS engine project, I would not have touched the AI and would have instead focused my efforts on getting the networking code back in good shape. Working multiplayer functionality trumps having a decent AI, in my opinion. (Of course, if you have both, then more power to you.)

[Trumanator]

Well, if you're looking beyond TBS then you could try WC3. It managed to stomp me and 2 of my buddies at just the average setting. We do suck pretty bad though, so who knows...

[DakaSha]

one of the first things i noticed about the AI in galciv (and im noticing again now) is how it thinks it has to colonize planets deep in my sphere of influence...

thanks for the free planet

again... the AI might be the best in the world of semi-complex TBS AI's but why does it have to do something that moronic?


and yeah games with more simple rules are def better AI-wise.. im sure thats why galciv is actually so simple under all that make up....

im pretty decent with GML in game maker so ive been letting game concepts swirl around my head... Im trying to get a strategy game system down that includes all of the "cool" things like espionage, technology and my personal fav supply lines but is done in a simple enough fashion to allow the creation of a decent AI. An AI that can plan on simple mathematical facts (like chess) while allowing some deeper strategy (like uhm...well other TBS games)

of course ill prob just start coding and stop after a month but i can at least dream :P

[chrispedersen]

I would say battlebots has the best AI i've seen.. at least the most challenging to beat.

[Foodstamp]

Even AI chess can be pretty poor. For instance, Chess Titans will swap equal value pieces with you no matter if you are a piece up on the AI. This scenario plays out every single time unless AI has a move that forces you to lose a piece without the AI losing one.

So basically you can beat Chess Titans every single time if you get one piece up, just by swapping pieces the rest of the game and once your down to nothing you are the only player with a power piece left, putting you in a great position to win the game.

I am sure other Chess AIs are probably better (I mean they have chess programs that beat champions), but I figured the AI was pretty standard for the types of chess games most of us are playing.