If You Could Change One Thing About Civ3...

[Sirian]

We've never done a thread like this at RB. There's a lot happening now, though. C3C patching is nearing its end. The MOO Imperia have been launched and are taking up some of the attention that would have gone to Civ3. Epics rules are getting an overdue tuneup. Civ4 is in development. And I have not been following all of the succession game chatter the way I used to. So I'm wondering what's on the minds of RB players.

If you could change one (and only one) thing about Civ3, with everything on the table -- removing something you don't like, fixing something you wish were different, or adding something that isn't there, without regard to whether you think it's reasonable, likely, or even possible -- what would you change? And why would you change that thing as opposed to something else? Why is that item the most important to you?

- Sirian

[Kylearan]

Difficult question - there's so much I would like to have changed (several aspects of the AI, trading, ...) that I wish they would publish the source code for us to change. But if I had to single out one thing to change, then it would be rail movement. Don't give units unlimited movements on rails! Give them a movement of 6, or maybe even more, but not unlimited. Why? It kills the idea of multiple fronts, it kills the need to position your troops with care and strategy, the need for zone defense, city garrisons etc etc.

Why would I choose this and not something else? Close contenders are several aspects of the AI: A different aggression model, a different trading behavior, and better micromanagement.
With a different aggression model I refer to the fact that at least some civs (those favoring certain governments, for example) should declare war and peace based on might, like in GalCiv or MoO, and not only based on needs, on a die roll or on damage done. But that would totally change the philosophy of Sid's whole civ series.
With different trading behavior I mean the fact that every AI is willing to trade away everything all the time, if only the price is right, which is stupid. If an AI is close to winning the space race, it shouldn't sell aluminium or uranium, for example. If it's the only one with a military tech and is currently waging war - don't sell the tech, etc. That would take away some of the predictability of the AIs, but if I wanted to have a more difficult game, I could play variant games or Sid, so I chose the rail change over that.
Better micromanagement means that the AI should have an eye on its cities each turn, and if a city is about to grow, rework tiles to max shields (and vice versa). Additionally, it shouldn't build settlers/workers in cities that are too small, and improve tiles better. While these are changes that could easily be done, they wouldn't alter gameplay for the human much, only make the game more challenging, so I didn't choose that one.

-Kylearan

[LKendter]

#1 is the trade rep issue - it breaks even once from something absurd, and the game gets much hard.

#2 is better calculation of trade routes. The reason huge maps take so long is the recalculate of all possible trade routes, and what is broken. This is why taking a city, destroying a harbor, etc. can take several minutes.

#3 is an indication on the cities what cities are ready to riot. This would make the early game easier to manage.

[DaveV]

Oooh, hot button. Civ 3 has just never grabbed me the way 1 and 2 did. Civ 3 has a lot of good ideas, and addressed a lot of the things I didn't like in Civ 2 (I don't play Civ 2 since Civ 3 came out), but the whole is somehow less than the sum of its parts. Typically, I play a game to about the start of the Middle Ages and lose interest. I'm usually winning at that point, but it's such a grind to keep going that I just move on to a game of MoM or MoO or something. Maybe it's all the micromanagement that bogs me down. So I guess I'll vote for a dramatic reduction of micromanagement:

1. The diplomacy screen: aarrrgggghhhh! To play the game optimally, I need to check *every* AI on *every* *stinking* *turn*. If they have something new, it takes a couple dozen mouse clicks to barter the best price. If I have something new, I need to check what each of the AIs will pay for it. I've never played with more than 7 AIs, but I can just imagine how painful *that* would be. My proposed solution: take tech trading off the table completely. Drastic, yes, but there will always be some way to abuse the AI.

2. Stack bombard. [b] <Click>, [b] <Click>, [b] <Click>: hard on my poor mouse hand, and *boring*. Even [b] <Enter> would be a lot better for a good typist. As a side issue, it would be nice if the AI could use bombard units so I didn't feel guilty about using them.

3. City micromanagement: I don't enjoy constantly juggling the worked squares. Easy solution here: have extra shields/food/beakers carry over.

4. Workers: they made a real effort to add extra commands here, but the basic automated terraforming needs work. When one of the first pieces of advice to newbies is "don't automate your workers", there's something wrong. The ideal solution here would be a modifiable script system for the worker AI.

I think Civ 3 could have been a great game, but fell short. For me, there's just too much work for the payoff. And maybe the payoff needs to be bigger: one of the things I like about MoO and MoM is that you gain completely new capabilities as the game goes on, but I just get better infantry/mounted/bombard units as I move through the Civ 3 tech tree. Although I hated camels in Civ 2, they at least provided some strategy, and something for a builder-style player to do other than sitting in his cities.

Sorry if I offended any Civ 3 lovers; maybe I've just been playing Civ-style games too long.

[barron of ideas]

What we need are more strategic options. Right now there are two, and mixed strategies that combine them, Building (aka the farmers gambit) and building improvements in your cities) and Attacking (building combat units: warriors/archers/horsemen/swordmen/immortals/etc and rushing your neigbors and their neigbors until you control the world). There could be trading, exploring, religious conversion strategies like now there are substrategies leading to UN victory or Spaceship victory, but to get them implemented I think what needs to be changed is the scoring system. And maybe a prohibition of attacking a power's last city. No genocide.

So I'd change the scoring system so having the biggest area and most people isn't the most important thing. Lets reward one city and five city challenges. Some of the scenerios are steps in this direction, but the victory points for killing units seems off to me, or set to be too rewarding, as it seems to end the game by a points victory before the most interesting approaches can be implemented.

As you pointed out, one of the best things about MOO is the lack of scoring. You can easily discuss who performed better, using how quick compared to what they faced, but there is not final official game derived score. I doubt that someone with 400 points clearly did better than someone else with 399, unless you review both games. Unscored games make it easier to be a fellowship rather than a stratified "I am better than you because I can score three point more" group.

Rambling, sorry. Short version. I would change or possibly eliminate the scoring system. If you had fun, why do you need to know you got 4,000 points?

[Sir Bugsy]

1. I would eliminate leaders of all sorts. They tend to unbalance the game. If you are ahead in tech, why do you need a SGL to get further ahead? A MGL generated army can make a average military very powerful in the hands of the human.

2. The trading interface is very time consuming and detracts from game play.

[kryszcztov]

OK, I could write pages of stuff if I was inspired and if I was in the mood for it (but I'm not ;-) ), but let's keep it like Sirian asked. One stuff, one that wasn't posted yet and that drew my attention these last days.

It's about great wonders, and it concerns Civ 1, 2, 3. I've always hated the wonders cascade, I think they are both unrealistic and a flaw in gameplay. A real flaw, I mean, what's the point in making odd plans to let that AI grab that wonder, so as we can get that other one, but then that tech must be discovered fast enough... A few days ago, in a PBEM that was set on a pelago map, where we both had one tiny island, I started on a wonder after only one settler (couldn't settle more before Map Making was around). The French got the Colossus, the Egyptians got the Pyramids, then the French again got the Oracle (same city), all of this on Deity. End of the wonders cascade, just as I got Writing (only). Around 200 shields lost for what I wanted to be the Great Lighthouse. Of course I could rant on the AI advantages, but here is my request : remove the ability to switch wonders. That way, a wonder would have to be built entirely from scratch, and not before the prerequisites are met.

10 years I have been waiting for this. Civ4 is fortunately written from scratch.

[T-hawk]

I'd change the fact that it's not open-source code. That'd cover everything else quite nicely. :D

Seriously, I agree with Kylearan that the most serious issue is the AI's willingness to trade anything. Stated more precisely, it's that the AI is completely unable to adjust its valuation of trade items based on circumstances. Exploiting that is where the human gains his tremendous diplomatic advantage over the AIs. ("You'll trade me Military Tradition for Music Theory? Sure!")

I know it isn't a simple problem; it can't be solved with a simple algorithm. Each circumstance would need to be identified, analyzed, and coded, and there'd always be subtle interplay between different factors that the code couldn't recognize but the human could leverage. But I think this is the one aspect where the game has a serious deficiency that could really be improved. It's hardly likely to come with Civ 3, but I gather that this thread is thrusting more towards Civ 4 and other games.

[Justus_II]

OK, I was just thinking of this the other day, but one of the "features" I miss most from Civ2 were the restrictions on swapping production, and the graduated rushing system. In Civ3, it is effortless to accumulate shields with various projects, having virtually no relationship to what you REALLY want to build, and then once you have the required tech/resource is available, BOOM, instant Wonder/Cavalry/Reserach Labs/whatever. It is maybe most noticible in the Wonder Cascade, as Kryszcztov mentioned in his post, but I think it needs to be more than the wonders.

In Civ2, first there was a 50% shield penalty for switching between categories (Wonders, Improvements, and Units). Now you still had "Wonder Cascades" as switching within a category didn't have a penalty. But at least you paid a price to do too much swapping, and if your border city had to swap from a tempe to hurry a defender, it was going to cost you. Made for more strategic decisions, as you had to be more confident of being able to complete what you started. I would probably also include a lesser penalty for swapping even within a category (25%), so Wonder Cascades would be reduced, and you wouldn't start quite so many Coal Plants that you never intended to finish. It that realistic, maybe a Coal Plant really does have the same foundation and infrastructure as every other modern building?? :P

A second, related rule, was how they handled rushing. Rather than a flat 4g=1s rate, it was graduated. IIRC, if you were at less than 10% of the shield total, it would cost like 8g=1s, then it would drop to maybe 6/1 until you got to 50%, then it would drop again to 4/1, and I think it dropped again once you hit 90%, but maybe not. Again, it's more realistic, and makes it worthwhile to invest some time and resources into a project before just "throwing money at the problem". Especially getting that first 10% in, which should be doable even in corrupted cities, would be more realistic than the instant cash rush. I also THINK, but am not sure, that once you rushed something, it could NOT be swapped. So no rush a worker, then swap to market (which is now at 10%) so you can rush the rest cheaper. Not sure how well this translates into pop-rushing, but I would think you could have it calculate a 'sheild-equivalent cost' and then require that many pop be available. The "No swapping after rushing" would also close the phased pop-rush tactic.

So, my 'one thing' would incorporate the following rules:
1. Whenever you change production, you pay a penalty of accumulated shields. If the swap is within a category, it's 25%, if it's to an entirely different category, it's 50%.
2. Rushing costs start out high (8/1) and decrease as you have a higher percentage of shields accumulated.
3. Once a project is rushed (by any means) it cannot be swapped.

I think this would require greater strategic planning in what you build and when, probably require a little more honesty on defense, as swapping/rushing defenders when threatened will be much more expensive, and should at least reduce the wonder cascade. It also just 'feels' more realistic to me, most of the time, you will actually build what you set out to build. (Whether that's an accurate simulation of government programs is an entirely different debate.) It also removes a whole host of potential exploits, from Big Picture/Scroll-Ahead swaps, to "short-rushing", to the Pop-Rush ladder. Not all of these are considered exploits, but they are all things the AI cannot do, and so give the human yet another advantage. Now it may just as well open up a another can of potential exploits, but you asked for what I would change!

(I actually had something like this in mind as a concept for an Epic, where we would just voluntarily agree to no production swaps, unless forced (I.E. new tech obsoletes an old unit, or a wonder is actually completed, and you HAVE to swap away to something else).

As far as why I would choose that over so many other things (AI diplomacy, AI Armyphobia, Rail teleport, interface issues) most have been discussed often enough that I assumed others would address them, and I guess I have also just gotten used to some of them. The ease of production swaps always has annoyed me. Especially the way you end up looking at everything as a potential pre-build. I remember in the Celt game, thinking the best advantage of the cheaper aqueducts was that there was now a 50-sheild prebuild, makes it easier to pop-rush knights. What's wrong with that?? Same with late-game, where a port city has more pre-build options than a landlocked one, that may have run out of things to build, no logic to that.

[Zed-F]

Less micromanagement, across the board. I don't like having to move hundreds of units. I don't like to have to optimize production in dozens of cities. I don't like having to check the trade screen every turn for every AI. Make the game simpler. Make the game more strategic. Abstract more. Whatever it takes. I want to be able to finish a game in a reasonable amount of time.

Why is MOO1 still so popular? Richness on the one hand and simplicity on the other. That is what I want from Civ4. Give a game MOO1's simplicity and richness, with an AI up to Civ3's standards, and because the game is so much simpler, you'd start to see an AI that looks REALLY strong.