Disenchanted

[Harovan]

Velociryx: Great posts, great suggestions! I like these ideas, especially the "Sell units to allies" option and the "peace keeping forces", but also all the other, without exception. Most of them should be not so hard to implement, and would be the difference that makes an already very good game great.

Firaxis: I'd like to see you to implement at least half of Vel's suggestions. Don't look too much at the money. Heck, I would pay twice as much for a game twice as good. Civilization has always been worth the money I spent for it.

[Asmodean]

Vel: Great ideas. My own Civ3mod.bic already has a lot of these implemented.

I remember we had a debate some time ago, over whether it was okay to change the game rules. Some called it cheating - some said that you wouldn't be playing the game that Firaxis wanted you to play. I say: If the game makes sense to you/is more fun to play for you, if you change the rules, then by all means do so. Fun is the keyword here, and I believe that most of what you are proposing will make the game funnier to play.

A few of my own ideas:

1) Culture. Do one of two things. Either tune the AI to make it understand what happens if it neglects culture. Or, tone it down, making defections less likely. And definately make it harder to achieve cultural victory.

2) Combat. What's wrong with firepower, for crying out loud. The combat system in Civ2 was perfect , why do we have to watch spearmen defeat armor all over again??

3) Terraforming. Make it possible with the discovery of explosives. It should at least be possible to terraform Hills, plains and grassland. I'm not too sure about desert, tundra and mountains.

Keep it up Vel - I'm looking forward to the Mod. Let's hope that Firaxis are willing to listen with regards to the rest.

And for the record: I love Civ3. I just happen to think that it can be made even better.

Asmodean

[Jumping Choya]

Well, I was thinking while I was cutting the grass today, and I came up with something that would appease both those who wanted SMAC's government options and those who see them as unrealistic in a history game like Civ III. So what I have decided is...

Split government into two sections: Economical, and The Other Stuff (sorry that I wasn't able to think of a better name for that).

The idea behind this is that not all democracies are alike, nor are all communisms alike. In fact, communism wasn't so much of a government model than an economic model (although when implemented, it did change other rights of citizens).

For the Economic section, options should include Frontier (nothing special, sortof like the Despot model), Capitalist/Free Market (like Democracy's current system), Communism (like it is now), and varying levels in between (something to fill in the gap between Frontier and Capitalism, for example). Your Economic system will determine in part how much corruption your cities experience, any bonuses, and worker speed.

The Other Stuff is more concerned with how the government is run. Options would be Frontier (again, Despotism), Republic, Democracy, Fundamentalism, Fascist, and Communist (even though I said it was an economic theory, I am at a lack to put a better term in for Soviet-style governments). This can also determine corruption, bonuses, and perhaps worker speed, as well as stuff like supported units and war weariness.

Using these two systems, you could mix and match. Each system could have some downfalls, so you might want to bring some balance to what you choose.

So, it is a rather rough plan, but I think it has some merit. What does everyone else think?

[notyoueither]

Economic: Barter, ..., Capitalist, Socialist.

Political: Despotic, Feudal, Democratic, Totalitarian, Fundamentalist (necessary?)

Rome was Despotic, ...
Elizabethan England was Feudal Capitalist
Nazis were Totalitarian Capitalist
USSR was Totalitarian Socialist
USA is Democratic Capitalist

What I am missing is the word for the proto-capitalism of the entire Roman Empire through early modern England (with the exception of the regressions after the fall of Rome, what used to be called the Dark Ages). Coinage is the the basis of much trade. Immense amount of goods are moved from place to place to satisfy local demands. Profits are earned by classes of individuals who are able to anticipate and supply these demands. Ahah! Mercantile.

Thus
Economic: Barter, Mercantile, Capitalist, Socialist.

and Elizabethan England was Feudal Mercantile.
They weren't quite forming corporations with share holders yet were they?
and Rome was Despotic, Mercantile.

Of course this still doesn't work, does it? Is Socialism immiscable with Capitalism? Ask the French, they'd say no.

But I'm done for now. Does anyone care to improve on it?

I just really wish a patch could implement such ideas. Now I'm wishful thinking, just like KM.

Salve

[Vercingettyrex]

Velociryx has got it spot on. I've play Civ 1, Civ II, SMAC plus a host of other stratgey 4x games like MORII and EU and I agree that the early to mid game is good but the mid to late game is unbalanced in some way.

I'm not sure whats causing it, but its definetly there. I think Firaxis were right to deliberately make conquest of the world a lot harder. I agree with this in principle, though the manual should have made this a lot clearer

On the other hand, I strongly agree with Velociryx that the player then has to be given some goals in the latter half to offset the inertia that comes with running a larger civilisation which has reached its optimum size. There are also other tweaks that could be made as have already been discussed making unit selection less of a chore.

The things I like in the game are Culture, Trade and Diplomacy. I think these are still in a raw state and could deal with further development. But the idea behind them is excellent.

Things I don't like so much are Espionage, and Unit handling. Corruption I'm still not sure about. It's there and maybe not much can be done about it.

However ... One possible solution is that if Corruption is considered lawlessness, the concept of law could be introduced and play something like Culture. Buildings and Wonders could promote law every turn and produce internal "Law" frontiers within your state. War of course, increases lawlessness, especially for border cities. The end result might be the same (ie expansion becomes harder), but at least it would allow players to feel as though they are tackling the problem. This concept would then allow certain diplomatic agreements to be setup to help with Corruption/Law such as Extradition treaties, Cross Border Agreements etc. One aspect of Civilisation in the broadest sense is law. Afterall Pax Romana meant Roman law. Global Capitalism relies on certain laws, ie property and contract, being consistent between countries etc. Of course this is not a patch issue but more of a CIV IV issue ?

[Willem]

Originally posted by Slax
The simplest end-game improvement: Add a great picture or movie for each type of victory (besides conquest and space race).

The current popup windows are such a huge let-down, especially after the late-game tedium.

No thanks. In fact I'd even like to turn off the pop-ups. How many times do I have to see a picture of a Pyramid? I don't even pay attention to it anymore. Besides, they cause problems for the Small Wonders I've created, the game keeps trying to crash as soon as those graphics pop up.

[Slax]

Originally posted by Willem


No thanks. In fact I'd even like to turn off the pop-ups. How many times do I have to see a picture of a Pyramid? I don't even pay attention to it anymore. Besides, they cause problems for the Small Wonders I've created, the game keeps trying to crash as soon as those graphics pop up.

Huh? You want to see nothing when you win the game?

[Skanky Burns]

Thats crazy talk!!

Well, i agree with Willem that another picture pop-up wouldn't add much, but a victory movie for each victory condition is nearly a must!!

Has anyone here played Carrier Command (old game, around the 386/486 era). Well, after months of playing, i finally won that game. I got a popup "You Win" and then it was back to the main menu... talk about a let-down.

While Civ 3 isnt quite as bad as that with the replay and the ranking, it still feels like an empty victory to win by any means other than spaceship. To me.

[Velociryx]

Ideas for the Editor

Everybody who's taken a peek at Civ3's editor knows it could use some enhancing. Off the top of my head, as I've been pondering ideas for the "Mod-with-no-Name" here are some things I'd really, really like to see:

1) An easy way to add new techs, units, etc, AND update the civlopedia right there on the screen.

2) A "this unit cannot attack cities" flag and/or a "this unit cannot leave your Civ's borders" flag (useful for creating Partisan/Militia defenders (especially colorless units) that could be used to whack settlers and other denizens from neighboring civs who REPEATEDLY violate your borders. If the AI will not let you be, there should be a way to keep all comers out without starting a world war. Such units, as I envision them, would be powerful enough to go head to head with whatever the standard defender is, assuming he's out in the open (but probably not have sufficient punch to actually attack a town, except in large numbers) to keep your rivals at bay. Trouble is, especially in the ancient era, there's simply no "room" given the default A/D/M values to do that (unless you wanna totally overhaul the entire system :: shivers:: ). So, a flag like this would allow me to create a colorless swordsman who could NOT leave my borders, or at the very least not capture enemy towns. (As it stands now, colorless units can attack and capture towns without causing a general war....the AI, in fact, does not even respond....not good! (as it stands now, I'm creating units that are x/x/1 - All Terrain as roads and requiring such units to cost one or more pop points as well, and even still, given their ability to capture cities without triggering a war, they're verging on being unbalancing). Such flags as mentioned above would open a whole HOST of interesting possibilities.

3) Working Gov-specific wonders/units (as it stands now, the "on" flag works fine for such units/wonders, but when you change governments, those units can still be built).

4) Age-specific HP bonuses (ie - a Veteran on an Industrial Age battlefield has experience that *should* count for more than a veteran of the Stone Age (reflected in higher HP's as the game progresses thru the ages).

Others? I've got more, and will post more later....right now, it's to bed with me!

-=Vel=-

[Willem]

Originally posted by Velociryx
Ideas for the Editor

Everybody who's taken a peek at Civ3's editor knows it could use some enhancing. Off the top of my head, as I've been pondering ideas for the "Mod-with-no-Name" here are some things I'd really, really like to see:

-=Vel=-

I'd like to see an option that ties units to certain improvements. I've never understood how a city can build an aircraft without an Airport, or a Battleship without a Shipyard.

[Libertarian]

Vel:

Your suggestions are good ones, of course. But how can they be implemented without massive coding changes on a scale that is likely prohibitive? New flags in the Editor, for instance. For every new flag, there must be new conditional branching algorithms hooked into the present code in every place that the flag might have any effect.

The musings are interesting, but how will they get done?

[notyoueither]

Good point Lib.

There are many great ideas for a new game being flung about, BUT...

I fear that our hopes for Civ3 must be more limited to fixes for current problems rather than major coding efforts. At least until an expansion that is. I just hope that stacked movement is not a *major coding effort*.

Salve

[player1]

Originally posted by notyoueither
Good point Lib.

There are many great ideas for a new game being flung about, BUT...

I fear that our hopes for Civ3 must be more limited to fixes for current problems rather than major coding efforts. At least until an expansion that is. I just hope that stacked movement is not a *major coding effort*.

Salve

I don't think it is. Look my post at start of this thread.

[player1]

Originally posted by player1
One programming advice to Firaxis: thing about stacking.

It is much easier to do this for human players only. Let AI use old ways, because if we force AI to use stacks then A LOT of code needs to be changed.

So stacks should be programmed like some sort of batch procedures. Like players is binding several units toghter. And when player says: "move", then batch proc. is started and computer moves them one by one (with animations turned off for all exept first unit). Same for bombard & worker tasks.

At the end it could be included for some other orders: forify, activate, air superiority, etc...

This is really NOT DIFFICULT to programm.

P.P.S.
Dan, if you are actually reading this, and Firaxis intends to make stacking orders, then let Soren read this last advice (stacking).

[notyoueither]

I really hope so. It would do so much to sooth the wound that is workers in the late game. If only for that.

Salve

[Libertarian]

Yes. As I see it, the problem of unit coordination — manifested by unary movement and bizarre activation sequencing — is the chief game-killer. After a fun beginning, the encroaching quicksand of tedious micromanagement begins to smother me, and I lose interest in the game. The decisions are no longer interesting, but rather are boring.

Fix that single problem, and I might likely revisit the game.

[Zoid]

Hi´ya all!

I haven´t read the whole thread (I´m doing the laundry...) so pardon me if I "state the obvious". Has anyone tried the mod. by Dreifels? Get it! (The link doesn´t work for some reason. Use copy and paste) This solves the corruption problem and adds a few trinkets like explaining graphics to wonders and such. It also adds "regional palaces" (to further reduce the corruption). It also puts a stop to the rampant expansion by the AI by changing the amount of food you get from tiles, the amount increases when you switch from Despotism to a more advanced form of government (much like Civ2, as I recall) This mod. made Civ3 fun again, even if it doesn´t solve the "slowing down-problem" that Velociryx described in his first post. But I think that if you can incorporate parts of the mod. with Velociryx´suggestions you would have a killer game!!

I recently asked a friend to return SMAC that I borrowed him like a year ago, I was so disappointed with CIV3 that I felt like reliving a past glory

Velociryx, couldn´t you team up with some game programmers and crank out a Civ-clone? It would kick some serious butt, man Further more I love the stuff you write, you´re one hell of a strategy gamer. If I had half your talent I would be happy

Keep up the good work!

[moomin]

Originally posted by Libertarian
Vel:

Your suggestions are good ones, of course. But how can they be implemented without massive coding changes on a scale that is likely prohibitive? New flags in the Editor, for instance. For every new flag, there must be new conditional branching algorithms hooked into the present code in every place that the flag might have any effect.

Yeah, those suggestions are unrealistic. They'd require some serious rework of a _lot_ of things in the game. Just imagine rewriting the AI to take advanatge of this feature (by the way: has anybody seen AI privateers? I haven't).

But more to the point: I'm not even sure I think having an anonymous unit is good - the potential for abuse is infinite, while it tries to solve a problem that could probably be addressed better by a patch in the first place. Firaxis has to rip out the code that gives the AI the ability to "yeah, yeah, I will get out" while presenting the player with "withdraw or start a war" and have the the AI face the same choise. Furthermore, a withdrawal should be a deal under the same conditions as any deal - a twenty turn moratorium on doing it again, unless you want to take the diplomatic hit. Since both those bits of code are already in, it should be possible. (If there was a diplo penalty for reneging on a withdrawal, something would have to be done about troops standing where the AIs cultural border expands. Auto-awake and remove them, perhaps).

Also, there should be an two-turn automatic right-of-passage on peace treaties, so that you can withdraw your forces without that stupid "get out" message from the AI. You couldn't attack the same AI while this was in effect, of course.

[Dev]

Good post Vel but let me ask, what took you so long?

I ask because I only had to play halfway through one campaign before I realized what a player griefing letdown Civ3 is. :\



/dev

[nato]

Your suggestions are good ones, of course. But how can they be implemented without massive coding changes on a scale that is likely prohibitive? New flags in the Editor, for instance. For every new flag, there must be new conditional branching algorithms hooked into the present code in every place that the flag might have any effect.

The musings are interesting, but how will they get done?

This is such an important point.

Big changes like are being suggested are pure fantasy ... they won't happen. It is way too much to program for a game that is already mostly sold.

If I were to come up with ideas without having to take the difficulty of programming into account, I, and anyone else, could come up with lots of great ones ... but its just idle talk.

That is why I have always tried to keep my suggestions within the realm of the possible. They aren't near as fun, but they are realistic.

This has a second benefit. Making really big changes always has 2nd and 3rd order uninteded side effects. It is near impossible to guess what they will be. Making smaller changes is more predictable and less likely to ruin things somewhere else.

Seriously, who could not come up with brilliant ideas if the sky was the limit?

[Velociryx]

Hey guys! Hmmm....I'm not sure I agree 'bout the difficulty of adding new flags....(again, this from a guy with limited programming experience) is it not just a glorified "if/then" statement (If Flag=true, then follow this set of conditions without fail). I see what you're saying Lib, but I think it could be workable to approach it from the other side of that coin. Rather than have a branching altorhythim to try and figure out every possible permutation of the game, handle it at the individual unit level, flag on or off (1 or 0) and have rigid rules it MUST follow in all cases dependent on the flag status. (Especially in the case of attacking cities...I'm thinking there aren't too many ways to interpret this....the unit simply "looks at" the color of the civ controlling the city....if it's yours, he enters normally, if it's not, he gets a null...can't move into that tile at all).

I DO agree that pie-in-the sky enhancement requests are just so much wishful thinking for a game that's already on the market, but in reviewing the list of ideas posted here, very few seem to fall into that category. And, to a lesser extent, even pie-in-the-sky ideas are good things....perhaps not for civ3, it's true, but imagine what would have happened if the first guy who said "hey guys, rather than workers, how about we try something a little different....like...Public Works maybe?" had been shot down?

If nothing else, the ideas presented here serve one very important purpose....they demonstrate alternate ways of looking at the game system as a whole.....ways of stretching and growing the system into something greater than it currently is.....and that's a good thing....

-=Vel=-

[nato]

I love sharing and creating ideas too...

However there is another problem, besides the major one of programming difficulty.

I have proposed some pretty simple changes in other threads ... some people support them, but they also draw a lot of resistance.

I would be really surprised if big major changes could get majority support ...

On the other hand, I think pretty much anything Velociryx says will be accepted near unconditionally here, so maybe it is not an obstacle for him ... maybe he is the one person that could get changes to be accepted ...

Use that celebrity voice wisely please!

[Velociryx]

If only it were so, brother nato....if only it were so.

I'm no celeb....just a redneck who really, really enjoys strategy games and who likes to hear myself talk...lol. Oh, I might not have as *many* posts as some, but I like to think that when I do write something it's generally worth the read....maybe not, but I can hope!

And in truth, lots of my ideas get batted around, discussed, and frequently disagreed with....that's cool though, cos it helps me see things from the other guy's perspective too....keeps my brain turning and churning (and that's sometimes tough with all of six brain cells left functioning--courtesy of a bit *too* much fun in college!)

But I figure it like this: If, by my writing stuff down here and generating discussion on it, it brings about even a few changes to Civ3....if a couple of the ideas it's impractical to add in civ3 find their way into civ4, then I will have helped influence the growth and development of one of my all-time favorite lines of computer games, and what more can any gamer ask for than that? In a word, that'd ROCK!

-=Vel=-

[Velociryx]

Resistance to change....

Hmmm...yes, you're right about that. Living in the deep south, I am confronted by that on an almost daily basis. Fortunately, where I'm currently working, I'm surrounded by people who are a bit more open minded than most, so no one here has threatened to firebomb my house or burn me at the stake....although, working in places where that happens DOES have the advantage of making the day go by faster....still, talk about exhausting!

And how to counter such resistance? In the end, only with persistence to a given cause.

For example, the Mod I'm proposing has spawned a number of lively and often ongoing debates. There's stuff that ultimately made it IN the 1.0 version of the mod currently being constructed that drew critisism and draws it still.

But it's in the mod.

It'll be tested by the general public.

Maybe they'll hate it, BUT....those who choose to download it will at least give it a try before completely making up their minds, and if it comes out that a clear majority are unhappy with it....well, it is a game after all, and supposed to be fun, so nothing currently in the mod MUST remain forever....we'll see how it shakes out, and who knows? It's possible that with time, people will come to see some of the changes as part of a synergistic whole, and that the game is better for it.

My fingers are crossed....

-=Vel=-

[Libertarian]

I'm not sure I agree 'bout the difficulty of adding new flags....(again, this from a guy with limited programming experience) is it not just a glorified "if/then" statement (If Flag=true, then follow this set of conditions without fail). I see what you're saying Lib, but I think it could be workable to approach it from the other side of that coin. Rather than have a branching altorhythim to try and figure out every possible permutation of the game, handle it at the individual unit level, flag on or off (1 or 0) and have rigid rules it MUST follow in all cases dependent on the flag status. (Especially in the case of attacking cities...I'm thinking there aren't too many ways to interpret this....the unit simply "looks at" the color of the civ controlling the city....if it's yours, he enters normally, if it's not, he gets a null...can't move into that tile at all).

Oh, dear. Where to begin?

I thought for quite a while on how to respond to your question without dampening your enthusiasm, and then finally it dawned on me that that wouldn't happen. One advantage that your programming naivité gives you is that you never run into "the wall". While the rest of us might walk past a particular area entirely, you won't be afraid to dive right in. And while the vast majority of your suggestions will likely be impractical to implement, you will find that occasional gem that the rest of us would have overlooked — though it might be implemented in some variant method over what you had envisioned.

At any rate, you're right that, at its root, all programming systems on von Neuman machines are reducible to boolean decisions: A, Not A. And in general, the higher the level of abstraction, the more the sky is the limit. Yes, everything is basically an "if/then" sort of thing when it comes to conditions and branching. You might be surprised to learn that Windows itself is, underneath it all, one glorified "if/then" (actually, a "switch" — same principle). But when I say glorified, I mean GLORIFIED. It is the switch statement from hell, consisting of thousands upon thousands of lines of code, conditions nested in levels so deep that only another computer program can find its way through the code, and branching so complex that it is nearly impossible to track every manifold of runtime sequencing.

Now, Civ3 is not as complex as Windows, but it's not as simple as chess, either (in terms of so-called "business rules".) Without posting a novel here, it would be problematic to explain why something that seems innocuous can become a monster. And that's talking AI only. Set aside the artistic, interface, and other considerations that must be brought to bear.

Even the very best computer program is like a house of cards: tenuously balanced, dependant on stability, and just a slight breezy surprise from tumbling down altogether. No one knows what sort of code Firaxis has written, whether it's a well encapsulated work of art or a hard-coded mess of spaghetti. But let's assume what is usually the case, i.e., it's somewhere between. Given that, even the innocent notion of checking a flag for a unit's status and execution might call into play code that you really aren't considering.

Suppose, for example, there is a function that examines unit status for purposes of updating a graphical element. That function must either examine your flag or else be sent a copy of or reference to the flag. Likely, there are many separate functions that move information through the various gates, all of which might likely be affected by the new flag — not necessarily because they have anything conceptually to do with your flag, per se, but because, say, there is a data structure in memory that does not accomodate and recognize the flag, and must be modified accordingly. Then, when that's done, the code that is entwined with it must be examined also.

It's not just the methods; it's the properties, too. And as Nato has pointed out, there are tertiary ancillations to consider as well. What I mean is that even when you have anticipated everything you can in terms of program modification, the real gotchas are always going to be the things you never thought of.

Anyway, keep suggesting things. It can't hurt. You seem to enjoy sharing your ideas, and we certainly enjoy hearing them. I just hate to see your talents wasted on futility. I agree that you would make a good consultant for a game from the ground up, before any coding is done. Just remember that it is a dark and foreboding cavern down in that code, and what might seem like simple changes can be downright daunting to implement.

Oh, and one other thing. Whether you realize it or not, you do indeed enjoy a status with Firaxis that is above the rest of us. (This is, in part, thanks to your naivité, which has thus far kept your criticisms at minimal levels.) But you must remember from whence you came. What would help us most is if you would use your status to help ensure that the game, essentially as is, is made more playable. Get Firaxis to work on late-game tedium and editor functionality. Get them to increase the interesting decisions and decrease the mundane ones. Do what is right, and not what is expedient.

As I said before, you impress me more as William Wallace than as Robert the Bruce. Don't disappont me.

[pg]

word lib. i really enjoy reading vel's ideas, but i can't think they will ever be much more than ideas for civ3. vel, you really should give up on civ3 and move on to open source civ games(clash, freeciv, etc). those guys will listen to you and do their best to make the ideas work if they are good, they will also talk back to you. btw, why should you start beta testing for firaxis for free? they are not worthy of you.

[nato]

Yeah Libertarian, that is what I was trying to get at!

Making large changes is really difficult ... and because the game is already sold, no longer in the process of being made, I don't think Firaxis can or will make large changes.

All they will want to make are small changes, so those are the only realistic ones to hope for I think. However, just because they have to be small doesn't mean they can't be real important.

Velociryx does have a special property ... enormously popular and recognized by Firaxis personally.

Therefore it would seem to me best for Velociryx to use that unique influence on coming up with and lobbying for smaller, realistic changes, because he will be listened to in a way few others would.

That of course is totally just how I see it, and of course Velociryx and everyone else should just do what makes them happy. I'm just saying a lot of suggestions I see on the forum are too big to be realistic, and in Velociryx's case its a waste of that magical influence.

I'll go mind my own business now...

edit: Sadly, pg might have the right idea...

[Analyst Redux]

Vel, I'm also going to throw a bit of cold water on many of the themes in this thread, though I certainly appreciate the basic POV. Here's the thing: what you (and most folks) find most praiseworthy about this new offering is the improved AI. What you find least praiseworthy is the lack of options for strategy. What I see very little of, in this or any other such thread, is an appreciation that these two things are inextricably linked.

Some of you might remember me from the SMAC boards. I’ve been a Civ addict since the introduction of the original game. I was a first adopter of Civ II and SMAC. Based on my experiences as a first adopter of products by this design team, I elected not to be a first adopter of Civ III. Based on the list of stuff the first patch to this product fixes and fiddles with, I’m comfortable with having waited. The one downside is that I’m late to the party. I fear mine may be a hoplessly redundant post and I can’t possibly adequately mine the old threads to determine if that’s true. Nevertheless, I plunge ahead, on the chance that my contribution will prove interesting to at least some.

As a baseline, let's keep in mind the self-proclaimed mission statement of the Civ III design team: to focus on making Civ III’s niche the best human vs. AI TBS gaming experience on the market. I don’t happen to agree with the Sid Meier/Brian Reynolds idea that sitting alone in a room trying to out-think a stranger’s algorithms is THE peak gaming experience, but I’ll be fair and stick with evaluating the result against the stated goal of the designers. Measured by that yardstick, this game stands tall, and stands as the dramatic improvement in AI performance over prior designs that Civ II and SMAC were not.

Of all the sequels to the series since the first Civilization, this is the first one that impresses me with the attention that’s been given to the issues of its fundamental rules design. From Civ I to Civ II, there was almost no tinkering with the internal rules of the game. Apart from tinkering with early unit strengths to eliminate the chariot rush, the patterns of development and winning strategies from the original game to Civ II were nearly identical—new strategies involving abusing the power of some new wonders being the only exception. The only learning curve involved formulating streamlined strategies for the added pressure of Deity level.

From Civ II to SMAC, it was not so much a matter of changing existing game rules, as it was laying a number of new rules, features and complexities on top of the old. But, as I detailed in a lengthy and meticulous post on the old SMAC boards a long, long time ago, adding layers of new rules, features and complexities to the design actually represented a step backwards IN AI PERFORMANCE (please forgive the shout), as it: (i) tended to create the AI equivalent of decision paralysis, and (ii) created multiple additional imbalancing points and features for human players to exploit, resulting in an overabundance of winning strategies. IOW, the game became too damn easy to win. I won my first game on Transcend level and never did lose a game. Learning curve, in terms of human vs. AI, of zero.

Civ III goes in an utterly different direction from SMAC, as many here are disheartened to realize, taking a "less is more" attitude towards design. But that design direction relates directly to the visible improvement in AI capability. My first blush reaction after completing a few games of Civ III, and having to drop down a couple difficulty levels from top in order to avoid a butt-kicking, was to be impressed with the capabilities of the new AI. But the more I thought about the patterns I was seeing, the more I realized that the “better” AI was mostly executing “new and improved” strategies that fell into two categories: (i) old strategies that would have been bad strategy in a prior version, but it would have done anyway, but which are now good strategies, or (ii) “new” strategies that were actually the result of eliminating its ability to engage in the old, bad strategies. Rules changes seem to explain as much or more of “better” AI than any dramatic change in AI behavior. In a follow-up post, I'll detail my personal top-ten things that are different (mostly simpler) that "improve" the AI by making the exact same behaviors "smarter".

In its own way, it’s a clever redesign, and certainly reflects more thought given to the rules and their effect on human vs. AI game play (consistent with the Firaxis mission statement) than was apparent either with Civ II or SMAC. If Mohammed won’t go to the mountain, you move the mountain to Mohammed. Making more intelligent AI is hard. Changing the rules to enhance the strategic value of persistent AI behaviors is rather easier. I think that the idea of making the AI "smarter" by making the game less complex was a deliberate strategy, and not merely accident. I don’t think the latter likely, as the ruthless emasculation and elimination of so many powerful and interesting, but ultimately badly balanced and badly used by the AI, features from previous versions of Civ/SMAC couldn’t have happened by accident. Ultimately, though, it leads to what many here are observing is true: it's a more difficult AI to beat, but a more boring game to play.

Implicit in the design decision, though, is an admission by Firaxis: they don't know how to make the AI appreciably smarter than it was in SMAC. That's why the game is dumbed down: not to win a broader human audience, but to create the illusion of an improvement in the AI capability.

So be careful what you ask for. You want all those neat toys and geegaws like social engineering, faction personalities and a unit design workshop? You're also going to get an AI that's pitifully easy to abuse on all those fronts, just as the SMAC AI was. You're choice: more options or tough AI. You can't have both--at least not at the Firaxis programers apparent present level of ability.

I'm going to try to post my top ten list now. It is a long long long document. We'll see if I can get it into one post.

[Analyst Redux]

Here’s a list of ten rule changes that, IMO, have the primary effect of making the AI look better for playing the same way it always has:

1. Unit stacking:

The AI constant: AI has, in all editions of the game including the present, brought deep stacks of units adjacent to your cities, then halted them, intending to attack.

The old rules: Allowed you to eliminate the whole stack, no matter how deep, with just a few units used in counter-attack.

The new rules: require you to engage each unit in the stack individually, and because multiple attacks are disallowed, requires counterattacks be carried out by units at least as numerous as the invaders.

Narrative: Because other units in a stack are no longer damaged or destroyed when a stacked unit loses a battle, now, when the AI brings a deep stack of units to your doorstep, that’s a serious threat, instead of an invitation to slaughter enemy units. Same behavior, dramatically different result.

2. Right of Passage Agreements:

The AI constant: AI has, in all editions of the game including the present, been inclined to declare war from remote locations. Provided that you and the enemy occupy the same land mass, the AI will never use amphibious means to wage the war, but will march overland, regardless of what is in between.

The old rules: Required the AI, in the absence of an alliance, to negotiate the march across third party territories while gradually antagonizing the third party by having military units in their territory.

The new rules: Allow third party access to territory at something less than alliance level activity, without incurring enmity.

Narrative: Remote AI’s can now effectively wage war on you absent overt alliances and without angering each other with military maneuvers through each other’s territory. This version of the game contains much less AI infighting over the issue of territory trespass, thanks to the inclusion of this intermediate level of cooperation. On a personal note, I was decimated in the early stages of a game by an American AI that had negotiated right of passage agreements with two of my neighbors, while those same neighbors refused to negotiate (at any price) the same with me. When I tried to march across the “neutral” territory (desperate, because I was losing the war of attrition at my gates), the nations between us declared war on me over the intrusions in their territory. Same AI behavior, but now much “smarter” under the new rules.

3. Terrain enhancement simultaneously simpler and more flexible.

The AI constant: AI tends to “max out” the improvement of tiles under its control, whether it’s using them or not, and whether the improvement is justified. AI also prefers shields to food more than humans.

Old rules: Featured increasingly complex systems for improving and transforming terrain.

New rules: Eliminate several types of terrain improvement and transformation, while making existing terrain improvable in more ways (e.g. mining grassland without food cost, and roads on all terrain adding trade).

Narrative: AI behavior in tile improvement has always been questionable even at its best. Under the old rules, AIs were constantly wasting enormous amounts of time with irrelevant improvements that added no value to tiles, or improved unused tiles. Most especially, the AI has always wasted enormous effort building roads and rails (or mag tubes) on tiles that add nothing to its productivity. Civ III’s rule changes regarding tile improvement allow you to improve the productivity of ANY tile with almost any improvement (most significantly, adding trade to hills and forests with roads) AND make it important to lay roads on tiles not being worked (i.e. owing to the addition of strategic resource tiles). This means that the inane tendency of the AI to lay down a road on every single tile it owns suddenly becomes an efficiency, or at least less of an inefficiency. It’s also true that the AI valued shields so highly in its terrain-improvement endeavors that it would stunt the growth of cities in preferring them. The new rules allow the AI to create shields it thinks it must have while stunting growth less badly than before.

The multilayered terraforming options in SMAC were terribly detrimental to the AI performance and re-engineering the tile improvement rules of this game in a completely opposite direction is entirely a sop to the AI IMO.

4. Monetary unit support.

The AI constant: All AIs, but especially the aggressive/militaristic ones, have a far greater unit-build to infrastructure-build ratio than human players.

The old rule: Units must be supported by individual city’s minerals.

The new rule: Units are supported by coin, on a civilization-wide basis.

Narrative: The AI was always great at generating coins and lousy about allowing cities to overbuild military units until individual cities were choked with the need to shield-support units. The unit-support rule change, from shields to coins, whatever else it does, plays to the advantage of the previously established patterns of the AI. Some have observed that the AI does a somewhat better job of upgrading obsolete units to better ones. I don’t think so. Rather, what used to happen was that the AI, running low on shields (owing to the support rule), but still “desiring” to build military units, would build the cheapest units it could, which would then simply lie around because they were actually not useful to the AI’s military efforts in any way. Eventually, you’d stumble upon such a unit-soaked, shield-choked city and kill, oh say, a dozen out of date defensive units and watercraft, sighing as you slaughter. That doesn’t happen so much now, not because the AI is too smart to do it, but because it doesn’t get backed into that pattern by the rules any more.

5. No ZOCs.

AI constant: The tendency towards wanting/needing to walk in a straight line and execute frontal assaults on choked terrain.

Old rule: Any unit could arrest the path of any other civ’s unit (friend or foe) with a zone of control.

New rule: No ZOC’s, absent fortresses or special unit characteristics.

Narrative: Civ unit pathfinding has never been anything to brag about, and AI combat units running in circles because they can’t figure out how to get from Point A to Point B was a common sight in prior versions. Pathfinding seems improved, but nothing succeeds like simplification. Taking away ZOCs improves AI warmongering by making it easier for the AI to accomplish unit pathfinding when operating in third party or enemy territory. Maintaining chokepoint defenses against the AI now requires a much larger expenditure of resources. This rule change also works hand-in-glove with the “right of passage” rule change to present more effective AI troop movements without any need to improve the AI.

6. No more diplomats, spies and caravans (a process begun in SMAC).

AI constant: The AI was slow to build these units and rarely used them well. (Although SMAC AI was much better with probe teams).

Rules change: obvious and well documented.

Narrative: These were two units whose features (and abuses) the AI could never hope to plumb as deeply as human players did. Ridding the game of them not only eliminates points of imbalance and exploitation, but also makes the AI look better. Bonus points to the design team for recognizing from its tentative steps in this direction in SMAC that managing trade as a macro-function, rather than a micro-function added depth, as well as improving playability. Points off for being unable to implement the feature in a non-AI-abusable fashion without making the AI's trade proposals remarkably one-sided. As others have observed, it forces humans into strictly linear strategies. The AI, once again, made "intelligent" by reducing choices to one.

7. Barbarians no longer capture cities and no longer advance in unit technology (carryover from SMAC).

AI constant: Over-aggressive about expansion in very early game, would often fall prey to barbarian raids at raging hoards setting.

Rules changes: see title. Also, barbs no longer originate randomly, or from the seas. Barbarian threat virtually emasculated once the map fills with cities.

Narrative: Ever play Civ II on Deity level, using the “Raging Hoards” level of barbs and NOT see the barbs owning and operating at least one—and often several—former AI cities? Toning down the effect of barbarians means AI civs aren’t troubled by “Raging Hoards” the way they used to be because barbarians, themselves, aren’t the trouble they used to be. This was the great irony of the “raging hoards” setting. It made the game easier to win, while providing a scoring bonus for difficulty.

8. Guaranteed base tile production (carryover from SMAC).

AI constant: All civ games, including the present, have had this “feature” where the AI just HAD to found its next city on a certain tile.

Old rule: base tile production was dependent upon terrain.

New rule: base tile production is a constant.

Narrative: AI city/base placement was, and remains, monomaniacal. In the past, if that tile sucked, the base would never grow. If that tile needed to be cleared of forest, swamp, etc. first, then the AI would do that before founding more cities. And as I found out using the cheat feature of Civ II, the AI would do that EVEN IF it was the AI’s first or second city. This kind of monomania about city placement early in the game is death to the future of a civ. Can’t stop the AI from behaving that way? Then just tweak a small rule to allow the city placement without wasting time or penalizing base tile resources, preventing civs from crashing and burning like this before they get off the ground.

9. No such thing as useless city improvements.

AI constant: The AI tends to build what it builds, instead of what it needs.

Old rules: Buildings could become obsoleted by wonders (yet, the AI never sold them), or could be of minimal value, e.g. the courthouse or granary after the discovery of more advanced governments.

New rules: Civ III’s redesign makes just about any building useful for something at any stage in the game. Even civs that are maxing their research and wallowing in luxuries still need research and happiness improvements for culture points. No corruption-free governments, no population booms, and almost no more wonders that replace buildings. Even building useless wonders accomplishes culture generation.

Narrative: There are still sub-optimal building strategies and choices for building infrastructure, but there are no longer completely useless ones (well, ok, walls in most cities, but they are now incredibly cheap). This raises the floor level of civ performance. I daresay, the way the game is now, you could build improvements by rolling dice to decide what building goes up, and you could still manage a good performance in the game.

10. Automatic air missions.

AI constant: The Civ team programmers never have been able to make the AI use its air power in even a minimally effective or intelligent manner, in any edition of the game.

Old rules: Air units became the most powerful in the game, capable of wiping out city defenses completely, and on their own. Once introduced, air power was the decisive power. SMAC’s choppers, with their capacity for innumerable multiple attacks, were insanely powerful.

New rules: Air units are strictly support/bombardment units.

Narrative: The reduction of air power to support mission status—no more “chop and drop”—not only achieves better balance in the modern combat phase of the game, it also covers up what was previously a glaring weakness in AI capabilities—as in, the AI could never figure out the whole chop and drop thing, or even figure out that it needed to attack with air units first and ground units second, never mind use them consistently in combination. Similar, though less dramatic, AI deficiencies are covered up by making artillery and naval units into strictly support units as well.

The constant in this list: what's sauce for the goose is sauce for the gander. A simplified decision tree makes it easier to play well, but harder to distinguish your own play from anyone elses (including, of course, the AI's). It also definately makes the game more boring and repetitive. OTOH, as you reintroduce complexity, you also reintroduce AI idiocy. Therein lies the rub, and the problem that Firaxis is, quite apparently, not up to solving.

[Willem]

Originally posted by nato
Yeah Libertarian, that is what I was trying to get at!

Making large changes is really difficult ... and because the game is already sold, no longer in the process of being made, I don't think Firaxis can or will make large changes.

Well judging from the fact that at least a couple of members of Firaxis are posting on this forum, I'd say it's not a foregone conclusion yet that nothing is being done to improve the game.