Better AI for an expansion????

this sounds a little silly to me - what does everyone else think?

Hey there,

First let me say that Stardock and all involved have to be the best support team in the history of any game I've played. Course I was one of the ones burnt from the MOO3 debacle. But this is a great team and seem to have the best interest of the consumer in mind.

Also let me say that I'm no gamer fanatic - a few games come along and I'll play them. Gal Civ's have been some of those games and this game GalCiv2 (1.11) is also one that is wrecking my personal life.

Now after trolling through the boards and such - it seems to me that there is a "general" consensus that a great many of the posters would purchase an expansion pack if the AI was improved. I find this logic silly to me.

If the issues with the AI are indeed problems - then in my opinion I think they should be fixed with a patch. Expansion packs are to be used to extend the life of the game - new races, story lines, additional features, etc, but not to fix what is lacking in the original game. That's my opinion.

Personally, I'm not winning every game, so the AI is more than a match for me at this moment. However I have experienced the so called "getting over the hump" in some easy games. This seems to come a little premature in GalCiv2 as opposed to some of the other games I've enjoyed - so yeah I can see where the AI may not be performing up to par.

I guess I wanted to start a little debate on whether people feel that the AI should be fixed/patched/modded/improved via a patch or an expansion pack. As I said personally I think those kinds of things should be fixed with a patch.

What are some other peoples thoughts?

Keep up the good work stardock

5,519 views 6 replies
Reply #1 Top
Hi!

I'm not sure people are saying the AI is not good enough - it is very good in comparison to similar games, so its not that it has to be patched (ok there are a few people who think this but I think not the majority).

its more like the AI is good, but (naturally) can still be improved, and many people bought it for the AI and would value an AI expansion which (possibly as an option at the expense of speed) further improves the AI as opposed to an expansion which adds only features.

so, just as you could not argue that an expansion adding features should be a patch because the game should originally have had these features and needs fixing, I dont think you can argue that a potential AI-improving expansion pack adding more intensive algorithms (which would entail a LOT of work - AI is NOT easy) should be a patch because the original should have had these algorithms and needs fixing. if you see what I mean.

so, in my case, even though I think the AI is good (for an AI), and certainly in no way has to be 'fixed', I would certainly pay for an expansion which improved AI even more, just as happily as I would pay for one adding features.

of course the problem with adding AI in an expansion is that AI is 'under the hood' and so is not so immediately obvious as new features, which could make it harder to sell. you would have to trust that the AI has indeed been expanded significantly as advertised without any non-subjective way of actually seeing the improvement. of course, in the case of stardock this wouldnt be a problem for me!

thats my view on things anyway

Mark
Reply #2 Top
However I have experienced the so called "getting over the hump" in some easy games.


This is by design and a good thing. At levels below Tough, the AI is designed to make random mistakes. The less intelligent the AI, the more mistakes (in theory), and the easier for novice players. Any complaints about the AI on levels below Tough are really not that useful because it is hard to distinguish between an AI bug, and some random mistake that comes at a critical decision point that turns out to be disasterous for the AI. The reports of odd happenings on levels Tough and above, however, should be repaired in patches to the current game.

Expansions should give us new CONTENT, like new ship jewelry, civs, AI packages to apply to different civs, adding real features to the game. There are some diplomacy issues that I and others have raised that might require enough work (gfx and AI additions not just changes) to accomplish that they may fit in the Expansion category.

Fixing bugs and AI logic issues in the existing game are NOT in the purview of an Expansion, but of a patch. But I think Stardock already know this and are acting accordingly.

Cheers,
Reaver
Reply #3 Top
I think Brad posted that he thinks he fixed the AI's wont to build diplomacy and entertainment centers with reckless abandon on the 1.2 patch, so Stardock does indeed patch the AI.

I could forsee the AI in the expansion being better by being able to take advantage of the advanced espionage and diplomacy that was alluded to be under development.
Reply #4 Top
"Better AI" is usually done either through more AI development (duh) or restricting user options.

The reason why people rave about how good the AI in GalCiv II is is because the game is much more open ended than other games and yet has computer players that adapt very well to very different styles of play.

The AI does not satisfy every single user. There are people who have developed strategies that beat the AI. But that isn't a "bug" that should be patched. This isn't an MMO where people are paying $15 per month for continued tweaking. A more sophisticated AI is a feature that should be paid for.

If I were to rate the AIs of strategy games I've played on a scale from 1 (terrible) to 10 (perfect) I would rate them as follows:

Civilization 1: 3
Civilization 2: 4
Master of Orion: 3
Master of Magic: 1 (later improved to 2)
HOMM 3: 5
HOMM IV: 3
HOMM V: 5
Master of Orion 3: 3
Master of Orion 2: 4
Civilization 3: 5
Civilization IV: 6
Space Empires IV: 3
Galactic Civilizations 1: 7
Galactic Civilizations II 1.0: 7
Galactic Civilizations II 1.2: 8

There is no strategy game available of this kind that has a computer player that plays as well as GalCiv II's. Most computer games have their computer players cheat like mad. Civilization IV has a decent AI but the game is so restricted, the diplomacy is so tied down that you are hamstrung by limitations to the game to prevent players from doing anything unexpected.

If Stardock limited players to no more than say 4 of the same building on a planet and only 1 starbase of each kind in a sector, these calls of "easy aI" would come to a quick end. It is the game's open endedness that lets players find ways to defeat the computer player. But would that be fun?

I would rather not be restricted in how I play but instead have Stardock put more time into helping the AI deal with such open endendess even more in an expansion pack.
Reply #5 Top
With all my respect for those who think GC2 is more "open ended" than Civ4, i have to say I fail to see where. I think Civ4 gives the player more possibilities.

True, you can build only one factory per city in Civ4, but the role factories play in that game equals the role starbases and economic capital play in cg2, they increase base production with a percentage. In civ4 you use tiles to build mines, workshops, mills... You can do a lot of city specialization in civ4, and then you have also a lot of civic combinations which give a lot of room to different strategies, like communism+workshop spam for example. Then you have the Great People Concept... There is also more variety in victory conditions. I don't know about other people but I was able to play Civ4 in many different ways, using different strategic aproaches.

The only part where the player is limited to fewer options in civ4 is technology, but even there, the decisions you make on the path you follow have a greater effect on the game than in cg2.

True, the AI cheats a LOT in civ4, and can be really annoying some times, but a restricted game compared with cg2? i don't see how.
Reply #6 Top
With all my respect for those who think GC2 is more "open ended" than Civ4, i have to say I fail to see where. I think Civ4 gives the player more possibilities.


Let me count the ways...

1. You can only build one improvement per city.

2. There is nothing like the starbase unit in civ. There is no way to expand your sphere of influence in the game other than city imps which goes back to #1.

3. Diplomacy is hard coded to prevent you from doing very much. Want to trade some troops to your friend in exchange for a tech? Tough. You can't.

4. You like great people? How is that open ended? They are little more than a bonus that peeters out.

5. The tech tree in Civ 4 is more like a Tech trunk. You will always research the whole tech tree. There are few choices.

6. In Civ 4 you can theoretically win in lots of ways but try to win through culture on anything but an easy level. Nearly all games end in the space race unless you turn it off.

7. You really can't specialize cities in civ 4. Galciv can have an industry world or a tech world.

8. Players have little control over population of a city in civ 4 compared to a planet in galciv where they have a great deal of control.

9. Improving tiles on teh map is not open ended at all. The terrain is predetermined on start up. There is a very limited number of things you can do with them. Your best cities in Civ 4 are totally determined by terrain. In GalCiv, a class 10 planet can trump much higher class worlds through starbases.

10. Civ 4 has nothing like galciv's freighters limiting to how you can influence the computer player's liking or disliking of you.

11.Gifting of units in civ 4 doesn't change the ai's opinion of you and is a crummy ui. Galciv has a lend lease type tying in it.

all this aside, civ 4 really does confine you to a very narrow and controlled path of play. i luv the game, don't get me wrong. But it's a very controlled experience. Let players build buildings that pump up surrounding cities production and culture and take the diplomacy restrictions for trading away and then we can talk.