Solution-Based Game Development

dh_epic

Cold War Veteran
Joined
Feb 10, 2002
Messages
4,627
Location
Seasonal Residences
I know this will be considered off topic... or even worse, an attack on our beloved Civ. But I was reading up on Galactic Civilizations by Stardock (someone told me it was pretty good). There's an interview with one of the key developers of GalCiv.

He seems to snipe Civilization. But looking passed some of the bitterness, I think there's an important point that Firaxis ought to keep in mind for Civ 4. The game doesn't become better by "wouldn't it be cool if...?" The game becomes better by identifying key problem areas, and discussing a solution to that problem that would make the game more fun.

Link here:

http://www.galciv.com/docs/qa.html

Lots of times games will come out and boast how they allow players all these paths to victory. And then you sit down and play it and it turns out that there’s really only one way to victory that’s any fun and the other ways are either virtually impossible or incredibly not fun. What we’ve done in GalCiv is work especially hard to make sure the different paths are enjoyable. In fact, the star bases, for instance, weren’t in the original design. We put those in along with the modules to upgrade them just this past Fall in order to ensure that winning through cultural and economics was as enjoyable as building fleets of ships and sending them into battle. Kind of a competitive “Simcity” type feel to it when you start building up your civilization and watching the little trade ships going back and forth by your star bases and such.

I can't help but feel like this is a dig at Civilization's "expand or die" gameplay. (Even though some people do the artifical challenge of a self imposed One-City game.) But by identifying it as a problem, they could begin to find a real solution, and expand gameplay in a variety of fun ways. I like the "competitive Simcity" analogy.

Here are their thoughts on AI:

The AI is multithreaded. What this means is that while you are taking your turn, the computer players are generating their strategies. That’s why there is no “please wait, computer players moving” screen when you hit the turn button. They’ve already calculated much of their moves. The only thing you have to wait for is the actual moving of units on the map.

The real benefit though is that it gives computer players much more time to “think” about their strategies. It means we can implement much more sophisticated strategies for them so that they play more intelligently.

None of this means that the game is “harder” to beat. But what it does mean is that we don’t have to dump tons of free money or whatever to the computer players in order for them to be competitive. They can play the same game you’re playing. We think many players gain a certain satisfaction knowing that when they’ve destroyed an Economic Starbase that it really did hurt that player. Most games I play I have to wonder whether blowing up some key building or unit really affects the AI or not. But in GalCiv, there’s no doubt because it’s playing the same game you are.

I know dozens of people who think this is a legit problem for Civ. Me personally, I don't mind if the AI cheats so long as they are realistically competitive. But there's no doubt that this kind of goal is a "holy grail" for AI design. By identifying this key problem, they look for a solution that makes the game top notch fun.

There are a lot of little tidbits in the interview. My point isn't whether Galactic Civilizations is a particularly great game. (Otherwise I would have called this thread "make Civ 4 more like galactic civs!!") I haven't played it, so I don't know. It could very well be crap. That's not the point.

But it would be nice to know that the Civ 4 developers "get it" -- that they're not just arbitrarily picking the coolest sounding feature. That's what impressed me about this interview, and why I use it as an example. I hope the Civ 4 developers are approaching the development of a new game from the same angle. They should identify the underlying gameplay problems in Civ 3 that make it less than perfect, and using those to discuss new goals for Civ 4.

Soren talked about a few things like corruption, or pollution. That gives me some optimism. But I'm curious if they're looking at the big picture about how to create a more perfect game than Civ 3?
 
Each incarnation of Civilization has been an improvement over the last. Granted, they haven't gotten it "perfect" yet, but I've ALWAYS had fun playing the game and winning by the different victories.

There are some things that I'd like to be changed (culture), taken out (corruption), or added back into the game (wonder movies!!!), but overall, each game has been a vast improvement over the last.

Just my thoughts.
 
I think that the devs have made a very strong case at improving or changing current concepts... That's why for the past year pretty much most of the things that stick to our minds is the goal of Firaxis to remove aspects which are not fun.

If this game was all about "more cool features" I think it would be the primary thing we would be talking about, but it isn't since the point is that they have been trying to improve gameplay.

This is not to say that a lot of cool aspects will not be in, because I am sure they will.
 
not fun aspects....pssh
i dont get what this means,,,
the whole time i have played civs, the only thing i found unfun was waiting 5 minutes on the playstation for the computer to makes its turn, but even that had its climax. since i had rectified the problem by building a super computer, there is nothing unfun about the game, even in doing what seemed like taking out the trash, i found fun....

my faith in the civ developers remains strong, i know i will be surprised and delighted when civ 4 is finally released.
 
I think that some features come across as a solution to a problem... like culture was their solution to the problem of people sending settlers to 'steal land' that annoyed people.

But other features, like religion, I just don't see what problem it's solving. Is it an excuse to go to war? We don't need another excuse. It won't improve the modern age, since religion largely makes its impact in the middle ages before the populist forms of government appear (communism, democracy).

Civ 3 did make an improvement on Civ 2, but in my mind, not enough, or not in the right ways. Cool features are cool features, but there are other priorities.

The interview points out two of them:

- The AI needs to cheat to win
- The "multiple paths to victory" are really only quick detours off the "expand and dominate" game

Imagine if those two things weren't true. Kind of hard to imagine, I guess, because a problem is always a negative. And adding a new gameplay mechanism is always a positive. It's hard to imagine the positive gameplay mechanism that would cancel out the negative gameplay problem -- it could be a lot of different things, really. That's why you have designers.
 
Makes perfect sense to me. Developers need to be able to approach their games critically, and looking for all the flaws. Of course, sometimes this leads to a disappointing end result (I'm thinking of Fable here.), but in general I think it is a good plan. I most especially hope they find a way to make the AI powerful without it cheating - I love a challenge of outwitting a good tactician. I don't like the challenge of outwitting someone who starts with the capacity to found three cities, or is twice as rich as me, or whatever.

GalCiv good, incidentally. ;) I'm picking up a full copy very soon, but the demo was highly addictive. Think SMAC + Homeworld + Civ3 + something else addictive. Cocaine, maybe. xD Also keeping a strong eye on the sequel, which looks like being even more interesting.
 
dh_epic said:
There's an interview with one of the key developers of GalCiv.

Mr. Wardell is not just a key developer, he's the designer. He's also company president, I believe. He has a very open, hands-on approach. I've interacted with him on the Stardock forum many times, and I like his approach.

I'm a GalCiv fan. I led a pack of Realms Beyond players over to Stardock and we had some good fun with the game. I stopped playing about a year ago, and I don't think any of our players have played in the last nine months, but our empire is still perched among the top ten:

http://www.galciv.com/metaverse/empires.asp

If you want to learn more about GalCiv, you can visit Sirian's Restaurant of Eternity where I reported five of my games, including some first-of-their-kind ubervariants and a grueling come-from-behind-win every bit as dicey as RBE DSG2.


Brad Wardell said:
None of this means that the game is “harder” to beat. But what it does mean is that we don’t have to dump tons of free money or whatever to the computer players in order for them to be competitive.

Competitive with whom? Brad may be stretching his claim there. GalCiv has some very bright spots, and among the brightest is its diplomacy model, which involves the AI. However, the GalCiv AI is incapable of using starbases for cultural victory and won't ever beat a competent player to tech victory without the huge boosts given it on the highest difficulty setting. The only thing it does well is to play the military game. Yet as you have heard me say many times now, DH, the overarching issue is diplomacy. Who lines up against whom determines the course of games.

There are holes, BIG holes, in the GalCiv AI at the top end. Diplomacy, global strategy, adaptability. I don't want to spill the beans here, because GalCiv is a fun game, and some of you may decide to try it. In the same way I wouldn't offer spoilers for an RPG to "give away the ending", I'm not going to give away the flaws in Brad's AI. You have to play quite a bit before you can find them, and then there is still more good GalCiving before you run out of strategic challenges.


Soren's Civ3 AI is stronger and sturdier, but it is designed to be merciful. You can pay it off with teensy-weensy little bits of tribute and then you're (usually) safe for a good while, until the next time they come calling. There's the occasional sneak attack, but these actually become LESS likely at the upper end because the Civ3 AI tends to attack the biggest threats first, and the higher you set the difficulty, the deeper the hole from which you start, actually reducing your odds of being attacked early. It is also VERY easy to bribe, so if you do get into trouble, help is ready at hand.

Brad's GalCiv AI is ruthless and merciless. It is specifically designed to target the weak and eradicate hangers-on. It too has a merciful grace period to let you get started, but when it runs out (and it happens early) the gloves come off and the AI will rub you out. From a no-spoilers, no-reloads perspective, I lose in GalCiv a lot more. However, with a SINGLE reload at a place of my choice, I have turned all my GalCiv losses into wins except for "snuffed in the cradle" losses where I never had -any- chance at all to win in the first place.

If Soren's AI played cutthroat, it wouldn't need Diety-level bonuses to compete, and there would be no Sid level.

I don't know that Brad is digging at Civ and its AI. His multi-threading does save time between turns, but his AI beelines mindlessly at the enemy the same way every other game AI does, including Civ3's.

I and the entire RBGC community urged the Stardock team to upgrade the AI in their expansion pack, but they focused on other stuff. I bought the xpack anyway, but without a new AI there isn't much for me to do with it. It's the same as when I had mastered the Civ1 AI: at some point, you've seen it all and there's nothing new left to experience. That took me a year with Civ1, fifteen months with Civ3, and six months with GalCiv. In fairness to GalCiv, though, its games are MUCH shorter (I don't even play with "wait at end of turn" it's such a sleek game), so the actual number of games I played is about the same as for Civ3.

The key difference between Soren's AI and Brad's AI, in my mind, is that the flaws in GalCiv are more fatal: I can still enjoy a game of Civ3 with the outcome uncertain well in to the game. With GalCiv, if it doesn't beat me with a grunt rush in the first few turns after the grace period wears off, it never will. Thus I reach a winning position much sooner and that's boring.


Oh, and if you don't think you need to expand in GalCiv, wait until you see the Drengin fleet closing in on you and your three planets :eek: and tell me how you're going to fight them off. :lol: These are both empire games, and they have more in common than they have that is different. :crazyeye:


- Sirian
 
This guy might be on to something if Galactic Civilizations was any good... :dubious:
 
Does that mean GalCiv got panned in the reviews?

fwiw, I recently bought Victoria because everyone says the government model (lots of sliders, based off the EUII model) rocks. I played it, and I reckon the government model is weak. Changes in government aren't driven by your decisions so much as in-game events, which are mostly a fixture based on your nation and the scenario. Compared to the pre-scripted stuff, your decisions seemed to me to be mostly cosmetic in influencing anything. Doesn't help that the manual did not explain the in-game effect of anything at all.
 
Ah well, turns out Brad's words were hollow. That's too bad. But I do have a tremendous amount of respect for him even identifying the problems. I think that was my real point.

I like the fact that Soren identified "unfun" concepts like pollution and corruption that could be replaced by something more "fun" -- like health. But I've yet to see any kind of big picture thinking on Civ 3.

Religion and civics kind of came out of nowhere, you know? Not that I'm opposed to them, but I'd like to know that they're looking at the big picture. Like "We find that players too often isolate themselves, and wars are generally one on one, if not for a temporary alliance. We hope to encourage more globalization in Civ 4." They could even say "We hope that religion will encourage more globalization in Civ 4" -- if religion is their solution.

I guess I'm holding my breath until the first Civ 4 developer interview.

Sirian, do you think the AI will be inherently flawed? I almost think so. Also, do you think it is possible to create a game where expansion DOESN'T give you easy access to every other victory types-- spaceship, culture, and domination?
 
GalCiv economy has two components: trade and expansion. Trade is half the economy and literally independent of expansion, so when running a One Planet Empire you are running with a half-speed economy, rather than economy that is a tiny fraction of what a large empire would have. This is an effect that must be played to be appreciated in context, but you can get some kind of feel for it by reading my two OPE reports at the Restaurant.

GalCiv also has a working Alliance Victory condition that is loads of fun. It lets you out of won games sooner. Unfortunately, once you know what you are doing, the games can also be won much sooner. So in the end, this turns out to be kind of a wash, but it is loads of good gameplay for any who are still short of complete mastery over the game.

I don't know that Brad's words are hollow. Some points are overstated, but I think he wrote that at a point before players like me got in there and exposed the framework. It took a lot of gaming to find deeply rooted design flaws. Civ has its flaws too, of course.

Brad was also speaking more to Master of Orion III than to Civ. MOO3 was GalCiv's more direct competition, and by comparison to that clunker, GC is a total masterpiece. :cool: That is one thing Soren and Brad have in common, the tenacity to avoid complexity contests in their designing. Sometimes I think a good possibility or two gets left on the cutting room floor, but there is no denying that overall, Civ3 and GalCiv are at the top of the food chain in the genre today, one for past/present and the other for futuristic.

I'm a bit concerned about GalCiv2, though. I went to the site the other day and the hints and teasers seem to indicate that they are going to follow the lead of Master of Orion II and introduce an Uber Race in to the mix. That would not be the way that I would go. I hated MOO2 with a virulence that must be seen in action to be fully understood. :lol:


Sirian, do you think the AI will be inherently flawed?

I think the AI faces very stiff challenges. The only way to write a great AI is actually to write a heckava lot of it, because no matter how clever a single algorithm may be, once players learn the shape of its performance, they get ahead of it. Unpredictability is the key difference between playing human opponents and a game AI. The first game to overcome this problem is going to reset the bar for everybody. Writing five times the AI for one game is not likely to be the top priority for most companies, though.


Also, do you think it is possible to create a game where expansion DOESN'T give you easy access to every other victory types-- spaceship, culture, and domination?

Sure, but not in this genre. Empire games are ABOUT expansion: imperialism, colonialism, brush wars and hot conflicts between major powers. I'm not sure I understand the urge to remove expansion as a determining factor. The way forward lies in requiring the player to make difficult tradeoffs between further expansion and more strength in the core, where the ideal balance shifts from game to game, requiring players to apply thinking and strategy intead of cookie cutter game plans. Better AI is a key element here, as it is from the AI that pressure must come, making it hard for player to expand rapidly and still manage to defend all of his holdings.

Of course, one can expand peacefully and play nice, and there should be ways to win that way, but if you conquer all opposition, you're going to be a in position where you can pick the ending of your choice and no one can stop you or beat you. I think this is only a problem when it comes to SCORING SYSTEMS and measuring performance. It would be tough to create a scoring system where, say, an Economic Victory will score more for Peaceful Pete in his 1/6th of the map than it will for Warmongering Wade and his 3/4ths of the map, having crushed everybody in sight.

In my mind, though, the important thing is that one can win the game without HAVING to occupy 3/4ths of the map. Setting score aside (or not even bothering with score, as in the original Master of Orion), there should be fun and interesting (and challenging) ways to win without having to roll out the military hardware and steamroll your neighbors.

That's why I found the GalCiv Alliance Victory fun. After making friends and together killing off the opposition, you could win together with your allies, instead of then having to turn on your friends and eliminate them too! :eek: I hope Civ4 has top notch diplomacy. I also hope it comes up with stronger victory conditions than Civ3's diplo and cultural options, but we'll just have to wait and see.


- Sirian
 
Maybe I'll take a peek at Gal Civ 2 when it finally comes out, instead of trying Gal Civ one. It seems to me like it has some merit :)

I happen to think that an AI is inevitably exploitable. After studying four years of AI, there are certain kinds of problems where there IS an ideal solution and it cannot be achieved algorithmically. Only the most tenacious of humans can achieve an ideal solution, if it exists. Not to say that they shouldn't make the AI better, but I think a little bit of cheating is acceptable. But I don't think that cheating should be a static advantage, but something the AI does on occasion only if the game needs an injection of challenge. (E.g.: the player is starting to get way ahead, so the AI should catch up).

I do find it odd that people can be holding out for that holy grail of AI when they can't yet cite a game that has achieved a challenging enough (legitimate) AI. Do we really need to have built Skynet, Hal9000, and the Matrix to have a challenging AI?

I also find it interesting that you consider Civilization to be an "Empire" game, where the genre is about expanding and conquering. I don't consider it to be in that genre, although it's definitely on that fuzzy line. I'm not sure what I'd classify it as, but it's definitely not as simple as Empire. It's got many "Sim" like qualities that I think it should embrace. I think that's why I kind of liked Brad's comment about "competitive Sim City".

There are plenty of reasons why Peaceful Pete with 1/6 of the map could beat Warmongering Wade with 1/2 of the map (although maybe not 3/4, that's a little powerful).

One is because Warmongering Wade is spending heaps of money to keep any semblance of unity in his empire. He may arguably have a "larger economy", but he's also losing a lot more money than Peaceful Pete is in his minimalist Republic. It doesn't matter to Warmongering Wade, though, because he's very close to dominating the globe, if he can hold his empire together.

Two is because Peaceful Pete's Republic isn't exactly an island. He trades with the remaining 2/6 of the world, which have shunned Wade's Warmongering ways and put sanctions on him. 1/6 of the world trades with Pete voluntarily. But another 1/6 of the world trades with Pete because he secretly sponsored a regime change that favors him. He pulled a dark-diplomacy deal where he sold weapons to Rebel Ray, helped Rebel Ray take 1/6 of the world, and now Rebel Ray siphons off huge amounts of funds to Pete. And to make matters even better, Pete doesn't have to deal with the problems that Wade does. Rebel Ray manages his own people, so Pete doesn't have to pay to solve Ray's problems. Pete gets the profit without the pain, while Wade's wealth won't win him the game anyway.

The game is as much a conflict between Wade and Pete as it is between game styles and personal challenges. Pete focuses on economy, and his biggest personal obstacle is his shrinking number of trade partners. Wade focuses on conquest, and his biggest personal obstacle are the costs coming from holding a multi-ethnic empire. And time is working against them both.

It's possible, if you stop thinking about the game as a narrow "empire" game.
 
OK, let me explain "empire" games. I absolutely LOATHE the term "4X game". I find the four X's to be inadequate to describing the gameplay. I prefer to call them empire games, because an empire is about expanding beyond your homeland to control a larger share of the habitable area, and not necessarily doing so by warfare. Colonialism -and- conquest together make an empire game.

There were empire games before Civ (I can think of a few, including Reach For the Stars) but Civ reinvented the genre by offering a peaceful alternative to conquering the entire world/galaxy/universe: the spaceship launch.

With the exception of OCC and other variants, Civ involves expansion. How much expansion is needed depends on the game and the player.

Sim City is also an expansion game, but I'd put it in the Tycoon genre (another one reinvented by Sid Meier). "Business empire" games.

Building empires of one kind or another is just a very appealing and fun thing to do. :)


- Sirian
 
Sir Schwick is onto it. You can build an empire without necessarily occupying more land. I think it's more than colonialism and conquest... it's building relationships (with the carrot or the stick) and influence. Those require land, but they don't require but a quarter of the the land in the world. Not to mention "business empire" -- the accumulation of wealth.

Peaceful Pete DOES stand a chance against Warmongering Wade. And it's not the self imposed challenge of a One City Culture Victory. It's that Peaceful Pete and Warmongering Wade do the significant steps of expanding so as to provide enough living space and resources for their people. At the crossroads, both choose two different challenges. Wade goes for the challenge of maintaining order in a growing empire with people who do not get along with each other. Pete goes for the challenge of wealth accumulation and building trading partners, while avoiding the costs that would come from being the President of half the world.

And that's without even talking about the challenges that Pete and Wade pose to each other.
 
sir_schwick said:
I think the real complaint I have is how one-dimensional expansion is. Currently you either own land or you do not. THere is no way to gain use from territory without having your flag flying over it.

I agree. I think the concept of a peaceful commercial empire is one that is not really well represented in any empire game, although I feel you can get a rough approximation of it in Civ3 by playing the weakling uber trader than can buy its way out of bullying but have a relatively small slice of the world pie.

That said, Sirian has a point. Empire games are about empire building. Commercial power is just one aspect of empire building and a purely commercial empire is not going to save your ass. You need military power and the economic base, both in population and territory. Peaceful empires are possible and can win in Civ3, its just harder and you're unlikely to get a high score from the game. That's probably one thing I would like to see fixed in Civ4. The mechanics are there to win peacefully in Civ3, but the scoring system doesn't adequately capture the commercial side of a player's activities.

On the issue of gaining use of territory, there is the ROP in Civ3 that lets you use territory that is not yours to travel. I'm not sure what more you want. If you mean access to tiles, resources and get production from tiles that are not your own, that doesn't quite work in reality and in an empire game. Empires have historically profited from lands it did not directly control by demanding taxes and tributes, and this model is already approximated in Civ3 with the 'client system' whereby you can lock in smaller Civs into your sphere of influence by making them your dependents, technologically and resource wise and you effectively gain control of their capitals and their tax revenues as they pay you all their free gold in exchange for your technology and resources. I am currently writing a treatise on this, which I loosely call the Machiavellian Doctrine -- the idea that players use of soft power to control and profit from your planet (in conjunction) with going to war and taking territory.

Everything considered, I support the general idea that the commercial aspect of empire building needs to be boosted abit, and Civ is the franchise to do it.
 
Maybe you have a smarter AI, but I find it hard to get any tribute without periodically invading something. Even then the tribute is usually in erratic forms. Basically getting tribute from the AI is so frustrating war is much easier, and more profitable then anything htey give me. Also, there is no way to reap the benefits of mercantilism. Europe made a fortune practically stealing resources from colonies and processing them and reselling the new goods back to the colonies.
 
Not tributes in the terms of demanding them pay you gold. But these are trades that sap gold out of their coffers into yours and you reinforce that dominance by being their chief supplier for a variety of goods.

I'll try to post an early draft of my treatise as soon as I can finish writing the important sections.
 
Actually, somewhat O/T, but I found very interesting what DH_Epic said about the challenge Wade has in holding together a multi-ethnic empire. The reason it was interesting is because I have just finished reading 1984 again, and the reason given for the perpetual war is that none of the three powers can EVER conquer the whole world because, to do so, they would have to integrate a massive number of people from a new ethnic group into their empire-and doing so might allow the people to realise that these 'foreigners' are not actually so different from themselves. It might be interesting if this fact could be somehow reflected in Civ4 :)!

Yours,
Aussie_Lurker.
 
Back
Top Bottom