The enemy AI is certainly the area of this game most in need of improvement. In war it does okay, but only if it has enough ships, troops, money, ect. When it comes to the process of going from one system (or even several, if a young empire) to a large empire, the AI just cannot keep up most times.
As a player, I vastly out-expand the AI for two main reasons:
1.) I colonize where it makes sense, regarding imperial integrity. This game doesn't necesarrily invoke the laws of thermodynamic exchange, but there are very real and tangible benefits to creating your first large group of colonies close-by, and preferably in a neat little group. If you're careful about your first few colonial acquisitions, the rest are much easier, because...
2.) Fleet Management. AI is better than it used to be, but still fails to make some pretty basic risk-reward calls when it comes to building new fleets, basing them around effective locations, and safeguarding their posessions. I tend to give each world at least a few frigates--usually much more, if I'm not poor. And I'm usually not poor, because...
3.) Taxes. AI doesn't do taxes right. At all. I remember the first game where I used manual taxes, and it hit me: The AI is really bad at this. By keeping your people happy and taxes low you can have worlds of 2000-3000M after only a decade or two at most. By that time, you can begin to tax them heavily in order to pay for all the new worlds that you AREN'T taxing. In order to avoid over-taxation, you decrease the load on your older worlds and spread out the tax-load with time. With revisions only every couple of game years, I can manage my taxes infinitely better than the AI, apparently, and I don't think I'm doing anything all that special.
So, the problem lies in the AI's decision-making process, IMO, and not in some core mechanics of the game. If the AI only thought a little smarter, it would do what I do. By now, I have a system--my Empires all look the same after a few decades, and are much too powerful to make continuing anything other than a test of "How unfair can I make this against myself?" If the AI would consistently do the things that I do (which aren't, I don't think, any different than what most of you do--and probably a lot less impressive to be honest) then perhaps it would stand more of a chance. As it is, the game is single-player only.. which means providing a challenging AI should be a very serious and main concern.
Did you catch that? I meant three reasons. My dog, bad. Also, as a caveat, I understand that nothing in video games are ever as easy or simple to fix as they might seem. I don't mean this to sound antagonizing, but rather as an attempt at helpful insight and constructive criticism.
< Message edited by Gelatinous Cube -- 12/31/2011 3:44:08 PM >