Regret buying the game?
I do, for a few reasons...
from 
GalCiv2 Forums
GalCiv2 Forums
I play more games than I buy, because, quite frankly, most are only good for a few hours. After playing around with GalCiv2, and noting the developer commitment to both continuing improvement of the game and meaningful participation in the user community, I decided that GC2 was one of those that I wanted to support and bought it.
Well... maybe I should have played the game a little longer. At this point, the honeymoon is officially over and what I'm left with hardly seems worth $45.
The AI, to put it mildly, was more impressive in print than in person. Certain fundamental structural issues Stardock got exactly right, and seriously, props to them for that. They made the right choice with regards to parity ("no cheating") between human and AI players (well, almost--the diplomacy skill is a break from total parity). They understood how the AI should change its decision making process as the difficulty level changes, and from what I have read implemented this successfully (although I can't say, not having played on the lower difficulty settings). And, in general, the basic mechanics of the game are pretty solid and create an environment in which an immerisve, exciting, and sophisticated game play experience is possible. But the more I play, the more problems crop up with the AI; problems which, in aggregate, make the it so uncompetitive on "Intelligent" that, despite the great potential, the gameplay is shallow, predictable, and without challenge (for me).
For example, I recently completed a game (1.1B, huge map, 9 opponents, tough) and noticed these problems, among others.
1) Dedicated raiders devaste the enemy, where by raider I mean a cargo hull with 1 beam attack and maximum engines. Undefended troop transports, constructors, freighters, and unreinforced starbases are easy prey. The AI never builds or anticipates raiders, and as a result has no defense against them.
2) Enemy planets, as revealed by conquest or espionage, display poor utilization of special resources and suboptimal building strategies. I saw the screenshots that previewed the improved 1.1B buliding AI, and the screenshots looked ok. I didn't see that in my game. What I saw was mismatched buildings on bonus tiles, a profusion of unproductive buildings like orbital fleet managers, hopelessly long build times (upgrading to stock markets before upgrading factories), a propensity to build farms only where they're not needed, and a shocking underemphasis on production in general. And yes, this was in 1.1B.
3) Poor shipbuilding combined with poor tactical deployment by the AI resulted in a kill-death ratio for me of over 15 to 1. It's hard to summarize the mistakes; it was many small things. I had a huge advantage in surveillance and speed, and the AI never managed to muster a serious threat.
4) The AI for diplomacy, in particular for tech trading, is highly dysfunctional. The AI is not equipped to appropriately value technologies, taking into consideration the possible actions of the other players. This opens the door to aggressive arbitrage strategies for the human player. No change in "weighting" values can correct this problem. To put it simply, the AI needs to understand that it is better to be a participant than to be an observer. The key words here are "game theory" and "economics."
The only major problem not related to the AI, in my opinion, is these forums. This issue has been discussed here before. Seriously, though, sales were good. Put some of that into better site design, new servers, whatever. I don't know, I expect better.
I have some hope that future patches will resurrect this game, and deliver on some of the latent potential. As it is, with the AI problems, it's just, well, disappointing.
Edit: See comment #44 /Edit
Well... maybe I should have played the game a little longer. At this point, the honeymoon is officially over and what I'm left with hardly seems worth $45.
The AI, to put it mildly, was more impressive in print than in person. Certain fundamental structural issues Stardock got exactly right, and seriously, props to them for that. They made the right choice with regards to parity ("no cheating") between human and AI players (well, almost--the diplomacy skill is a break from total parity). They understood how the AI should change its decision making process as the difficulty level changes, and from what I have read implemented this successfully (although I can't say, not having played on the lower difficulty settings). And, in general, the basic mechanics of the game are pretty solid and create an environment in which an immerisve, exciting, and sophisticated game play experience is possible. But the more I play, the more problems crop up with the AI; problems which, in aggregate, make the it so uncompetitive on "Intelligent" that, despite the great potential, the gameplay is shallow, predictable, and without challenge (for me).
For example, I recently completed a game (1.1B, huge map, 9 opponents, tough) and noticed these problems, among others.
1) Dedicated raiders devaste the enemy, where by raider I mean a cargo hull with 1 beam attack and maximum engines. Undefended troop transports, constructors, freighters, and unreinforced starbases are easy prey. The AI never builds or anticipates raiders, and as a result has no defense against them.
2) Enemy planets, as revealed by conquest or espionage, display poor utilization of special resources and suboptimal building strategies. I saw the screenshots that previewed the improved 1.1B buliding AI, and the screenshots looked ok. I didn't see that in my game. What I saw was mismatched buildings on bonus tiles, a profusion of unproductive buildings like orbital fleet managers, hopelessly long build times (upgrading to stock markets before upgrading factories), a propensity to build farms only where they're not needed, and a shocking underemphasis on production in general. And yes, this was in 1.1B.
3) Poor shipbuilding combined with poor tactical deployment by the AI resulted in a kill-death ratio for me of over 15 to 1. It's hard to summarize the mistakes; it was many small things. I had a huge advantage in surveillance and speed, and the AI never managed to muster a serious threat.
4) The AI for diplomacy, in particular for tech trading, is highly dysfunctional. The AI is not equipped to appropriately value technologies, taking into consideration the possible actions of the other players. This opens the door to aggressive arbitrage strategies for the human player. No change in "weighting" values can correct this problem. To put it simply, the AI needs to understand that it is better to be a participant than to be an observer. The key words here are "game theory" and "economics."
The only major problem not related to the AI, in my opinion, is these forums. This issue has been discussed here before. Seriously, though, sales were good. Put some of that into better site design, new servers, whatever. I don't know, I expect better.
I have some hope that future patches will resurrect this game, and deliver on some of the latent potential. As it is, with the AI problems, it's just, well, disappointing.
Edit: See comment #44 /Edit
Stardock Forums
