Scanian8

Join Date 11/2009

Random issues with this game that keep bothering me

I guess I should start out by saying that this is in no way a post to begin yet another rant of "this is broken and must be changed now or I will be giving up on this game. These are merely issues that I just can't get out of the back of my mind and serve to irritate me the more I think about them.

1. Almost all the races have connections/backgrounds that involve each other. Iconians create Yor, Altarians have seen humans and are seeking why the are so similar, Drath and Altarians from same world...etc. Heck, even the biggest event in the universe, hyperdrive, managed to be passed from the humans to all other races. So here it comes...why do I have to research a universal translator? Did everyone just start forgetting their ties to the other races as they began to implement the hyperdrive tech? Did we scrap all our databases on the other cultures for no real reason? I get that the exact same Altarian and Drath leaders that once spoke to one another are most likely long gone. I can even take the idea further and think of all the different dialects that may develop in the time apart from the other races that may make communications difficult...but really? So much so that no one is keeping up with the other racial languages? This is only further complicated by the fact that every race instantly recognizes me and knows my history. STudy the history but forget the language? WHY? "We will dedicate many teams to learning all about the galactic empires that we may be facing, but as for keeping up with all these crazy languages...let's just build a machine to spit out translations when we have that technology, but for now...let our people live in darkness."

I'm not sure my feeling are coming out clearly. When I first had to research techs just to talk to others I thought, "That is awesome!" It adds a lot to the story and feeling of space exploration (obviously I'm new to the space-faring 4X genre, stuck to Civ in my days). But the more games I go through, and the more I have to hear "Good to see you again human" or some variation that implies we have a past, the more it keeps killing this novel idea for me. ANy chance someone can throw me some good fiction for why we forgot all these languages?

2. I am absolutely sure at this point that the AI in ToA knows exactly where habitable planets are located without scouting for them. I have run through a few test of this idea but the one that sealed it for me was in a regular metaverse game of mine. To explain quickly...my empire started in a corner with one star behind it. This star had 3 habitable planets on it, and I found this out ASAP, my flagship spotted all 3 planets on the first turn. Knowing that I could see every angle coming at this system, I sent my colony ships out towards other systems. As time passes by, not a single flagship nor scout passes into this system from another empire. Next thing I know, 3 races, the Yor, Humans, and Altarians are all sending 3 colony ships through my territory on a direct route for those planets. I could forgive one...maybe the AI thinks "empty system with 5 planets, ill send a colony ship to explore and maybe get lucky." This was obviously a direct effort by 3 different races to colonize an area they had never seen...trust me, I've never seen the AI send out 3 colony ships to a system in the *hopes* that the system will have 3 habitable planets, let alone 3 of them do it. This only bothers me because I could have sworn that I had read somewhere that the AI specifically does not do this.

3.The AI...don't get me wrong, this is not a rant on how horrible this game's AI is after all the praise from the reviewers... There are plenty of posts on this if you feel so inclined to join that discussion. I am actually quite satisfied with the AI. It isn't perfect. Of course it can, and will, be exploited by those who learn the game. But overall, it is entertaining and actually manages to switch things up a bit. I can't tell you how happy I was when my second game told me of the shift in Drengin ethics to good...they even played in a different fashion than normal. Hell, I could list a lot of features that keeps the AI interesting. They may lean towards certain scripted behaviors, but when they break from those it can always throw you a bit and add a lot of flavor to the game.

My complaint is more from the idea of why I feel that so many people have been disappointed by the AI. When the AI is struggling, it slips into a total spiral out of their own control. No good way to explain this, but when an AI has lost his footing in this game, it becomes clear that he is an easy target. Then we are soon left with the often mentioned planets pumping out ships only to die. The AI just cannot think of a long term plan to come back. It lives in the moment and sees "danger! danger!" and never realizes that if it makes a few concessions (as humans do) it may have a hope of resistance.

Well, thanks for sitting through my ranting. Love this game so I am really just looking for people to bounce these frustrations off of. Feel free to also tell me anything I am wrong about...Seen the AI actually scout for planets? How about an AI who actually came up with a good plan to rise from last to first? Happy to hear about it.

Locked Post

Advanced Search

Watch this post

Do not email me updates for this post Email me updates for this post

Successfully updated karma reason!

qrtxian

Join Date 03/2010

+30

My thoughts on the subject:

1. For most of the races, this isn't actually that big an issue. Drath and Altarians seperated a long time ago... I think millions of years. I've read the Altarians barely remembered the Drath except as myths. Humans and Altarians have never actually met before. Drengin and Torians (btw, a similiar concern of mine: if the Drengin need to research Planetary Invasion, how did they conquer the Torians before?) quite possibly never learned each other's language - Torians might have but I doubt the Drengin would bother. I believe the humans distributed hyperdrive through unmanned probes. The only ones that really give me conceptual trouble are the Iconians and Yor, Drengin and Arceans, and Drengin and Korath. The Yor probably wouldn't even have their own language, being created by the Iconians; the Drengin and Arceans have had a cold war going on for millenia, and, worst of all, the Drengin and Korath were until very recently the same species and should definitely be able to talk to each other.

That said, I really don't think this is as much of an immersion-breaker as a lot of other things (such as the amazing ability of my population to double in a few years). There's all sorts of backstory related oddities in any given game, such as the Drath telling my Terrans "You look like the Altarians who stole our world" when there are no Altarians in the game and never have been, or the mere presence of Torians without Drengin. I suppose, when you come right down to it, it's actually a weakness of having a detailed backstory - having the races begin with shared histories can make some aspects of the game confusing. Like I said, there's plenty of other techs most of them should have that they don't.

2. I'm pretty certain the AI does not know where habitable planets are, but I am 99% certain they do know the location of galactic resources. I've actually developed a resource-finding strategy where I follow a minor race's constructor to a resource, then use my much faster ship to claim it first. I know for a fact that the minors cannot have found these resources with the pitifully small number of scouts they've built.

In your example of AI omniscience, had most of the other planets nearby been claimed? The AI does send colony ships out fishing several times in my games when I claim their neighboring planets. Also, they build such a ridiculous number of scouts - I don't build many at all in the early game to keep my economy under control - that I doubt they "cheat" in knowing where the planets are. I also might ask what your difficulty is; I have heard that the AI does know where planets are on the harder levels. I don't play much beyond Tough yet, so my AIs have no bonuses yet.

3. Yes, the AI does suffer from a lot of problems when put on the defensive. Between their weakness at long-term planning and the idiocy of their use of military score alone to determine strength ("You've conquered my homeworld, crippled my economy, defeated my fleets in every battle, and bribed half the galaxy to finish me off, but my military score is five points higher than yours! Beg for mercy!") they aren't able to react well to losses. There have been quite a few wars where I would have gladly made peace with the AI, but because they won't admit I'm winning, I wind up crushing them completely.

I have seen the AI make a few comebacks, but it's rare and typically involves some sort of event. I recently posted a discussion of the 4 AI personality types, and in my experience, AI personality 11 (I think), which is used by the Arceans, Altarians, and Korx, overwhelmingly does the best at long-term planning. It's hard to quantify it, but these races seem to strike a good balance between aggression and build-ups and are by far the best at fighting wars. They've come the closest to acting against that "danger! danger!" tendency, although they're hardly perfect.

Successfully updated karma reason!

There have been several threads on what the AI knows, etc. Posters have run tests and experiments based along lines similar to your experience, Scanian8.

IIRC, one poster placed colony ships near planets in a corner such as you described and varied via saved games on exactly what turn he colonized the high PQ planet(s) there. No scouts ever showed, but colonizers would appear if the planets were not colonized, but would not come if they were. Enough variations were run to convince me that the AI always knows about all uncolonized planets while the AI is in the expansion/colonization phase.

That is, I am not sure but, IIRC, one can abandon a colony and make the planet uncolonized again. I think someone ran a test in late game and the abandoned high PQ planet may have remained uncolonized. I may have the wrong game, though.

One poster said that the AI had been revised in an earlier patch to colonize initially closer to its home planet to reduce its observed tendency to abuse its knowledge.

I, too, have quickly built or bought fast constructors and chased after minor race constructors in hopes of beating them to resources no one could see. Unfortunately, often I find that the resource had already been claimed even earlier and the one I was following or leapfrogging was simply a constructor heading to upgrade the base.

Successfully updated karma reason!

The backstories of GalCiv are great, they imbue this game with spirit, and some consider the roleplayign aspects of this game the #1 reason to play it. I guess esp. modders & AAR writers can sing a song about this...

Sometimes things doesn't add up well, but couldn't that be said from any movie/series on TV todays, if one takes himself the time to look into it?

Nevertheless, I have nothing against constructive criticism, and hopefully folks with such sharp look will join the beta in GalCiv3.

BTW I think the Yor (being droids) speak like this: 1101011100011... (in 256 bit)

***

I fully agree the AI knows everything of whatever on the map. I say this not to bother people, but because I believe it is actual truth, and I have witnessed that many many times, even though I haven't looked for it. My "guts" tell me...

It would have been surprised me if it was different. I mean, "the AI" is nothing but some logical sub-routines in the hard-coded side of the game, and to differ/seclude these two things will be hard to do. Now I'm not saying that this couldn't be done, no, no, but it simply would require much more programming to do so - esp. when you have a full setup of StockAI/Minor races, and the programming would have to hold different infos for each of them on each turn. It is much easier to do it like this, say, here is the map containing all info's, now RaceA play out your pattern with it, and next race etc etc.

In fact, there are many indices available directly in the game that will "hint" at this. For example, the statistical curves in the "Timeline" section, or what you can see on the Minimap. You will get to know how big the AI's Military is, how and where his production is, his pop, almost everything that is important in this game - without ever encountering him. And the AI plays by this rule, too - if you have a high MilRating he will actually treat you nice (and not declare war) although the possibility could be there that he has so far only encountered one single planet from you (somewhere in the outer rim of your empire) and by this can't know that you have massive fleets sitting at your core worlds.

Or you want to know what he's doing on a planet, but you don't have a spy ready or an espionage level. Ok --> search on planet-list *bingo*. See? The game doesn't differ its info, and the devs didn't cover up the holes too well. The game is full of these examples, which I don't mind as I tend to see/use them for my advantage.

But what is bitter if devs/CEO make comments to updates like "The AI has to scout now before he can colonize"...well, well.... yes, the AI will built alot of scouts (esp. AI 11) but he's not using them to scout out anything. He will burn his money rushbuying them, then sending 5 to one direction, and the other 5 to the other direction, and they follow each one another with the distance of one turn, in a straight line. I've played 100+ games in which I could see that behaviour, because I always buy every scout/miner the AI will ever built.

That is NOT scouting. Scouting is clearing the FOW, going zig-zag, sending only ONE ship in a direction, and sending the next parallel to the previous course. Actually, an AI could compute such routes much more better and efficient that a human ever could, but do we witness that?

BTW on AI-behaviour, personally I like "Thalan" the most, as he will not "burn" his money on things he doesn't really need like the others, and thus will still have most (of *my*) money on first contact.

2nd BTW "Aggression" is actually separated.

3rd BTW Jim, if you abandon a colony it becomes an uncolonizable PQ0. This was different in GalCiv1 where you could leave a planet and colonize it again, and each time the planet will gain +2 PQ.

4th BTW Scanian, if it becomes too easy you might want to play for score then, there, the AI's are your friends, and you'll be in love with their weaknesses. But you'll have to face a new REAL enemy: Time.

Have fun!

Successfully updated karma reason!

I'm actually glad I read everyone's wall of text. I'm not thoughtful enough to add anything. But I do understand every word written.

Successfully updated karma reason!

I am glad everyone got the point of this post...Thanks for all the great responses! There are a myriad of topics I may want to come back to, but just don't have the time as of right now. I do want to argue against one of my complaints though...

The very next game I played after posting this was an attempt at a ZYW (succesful) and in that game I played against the Yor on a tiny map...Crippling difficulty if I remember right. In this game the Yor were mass producing scout ships and actually sending them in all kinds of directions. Not only this, but they had colony ships at home that were not bee-lining for planets. I even spotted a planet pretty far off...his scouts never made it there, and he never sent a colony ship in that area. Guess it's back to wondering what the AI really knows...that or the devs are reading my complaints and messing with my AI

Successfully updated karma reason!