From: Leeuwarden, Netherlands
Of course the AI cheats! Does it break coding rules the AI and we people have to follow probably not. I have found the SC tactical AI to be pretty good, the operational AI so so, and the Strategic AI unsubtle and navy pretty bad. We humans have to understand without help (cheats) commercial gaming AI would never beat a decent human player. Playing the AI should be considered training for playing people. Although such training can lead to bad habits (I know I have some.)
As a retired USA Artillery Officer I had an opportunity to work briefly on military simulations. The amount of time and effort put into a military computer war game was huge and it still needed people to provide strategic, operational, and tactical sanity to the game.
So if we want to play AI verse human we have to understand the AIs limitations. It is fine to note issues and propose solutions but for a commercial AI to be really really military grade the game would probably cost $1,000 or more and few would buy it at that price (so no game.) Look at the effort (labor of love) that went into “World in Flames” 3 rule books, years of board game play, costs $100, and still needs humans on both sides.
So constructive criticism great, bitching is pointless. You want realism join the Army.
I agree. I haven't found much games that don't cheat some way or other. Learned to live with it, happy if the AI manages to pull off a brilliant move. Then again reallife Brass tends to do ... awkward ... things as well. Some players like to mimic this behaviour with the throw of a dice. They know they're going to battle with this or that AI, but let the declaration of war depend on the throw of a dice.
All in all if and when the AI bores you to death it's time for a human opponent. :-)