MOO2-style unit design and combat

[Shyrramar]

It may be then, that I am just plain stupid?
I had this effect when nearing the end of the game, but that was because every strategic game in history has had that effect (not many care to play modern times in Civ3). I did get my butt kicked around in the beginning - and mostly because the AI had better ships than I! I was young and stupid back then, though, so as I am now merely stupid, I might see things differently. Anyway, after Doom Stars, things usually got quite easy.

Well, the main concern seems to be in the AI. I think that it should be the main concern of Civ4 anyway. IF (and only if) it is enhanced, this system should be given serious thought. I do wholly agree, that human players should not get too much advance by this, as humans are clearly superior to the AI as it is.

 

[unscratchedfoot]

The problem with MOO2 was not the milk... whoops... I mean not the game design, but the AI intelligence. I think the AI for that game would be relatively easy to enhance compared to civ because it just needs some code on how to build good ships with the tech it has available. The weak code in that game just put a little of every weapon on each ship instead of pumping it full of one miniturized weapon type with huge numbers of it like humans do. And then have a random generator where it sometimes uses another good build strategy for example an optimized defensive ship using alien tech or whatever so the players don't know what to expect everytime. Same for civ custom building - it just needs the good code written in.

 

[vesuvius_prime]

Originally posted by Arathorn
Did I play a different MoO2 than the rest of you?

Combat kill ratios of 100:1 or more were very common for me -- I essentially never lost ships, precisely because the AI was so bad at ship design. Interest in MoO2 can disappear very quickly when you realize just how overwhelming the advantage of building your own ships can be -- and just how bad the AI is at it.

Arathorn, you played the same MOO2. But I imagine you were not an expert from the very beginning. It must've taken you some time to find out the best designs. In the early game that would undoubtedly be missile cruisers with nukes (if MIRVed, even better). That alone guaranteed the destruction of at least the nearest rival. But the real problem was that the developers of MOO2 never addressed this in subsequent patches. Players found designs that worked flawlessly. That was bound to happen. But then there had to be some reaction from the developers. We're talking here that it would be great if the Civ developers monitor closely the players' tactics and designs and then provide counter-measures in the patches and, perhaps, expansions.

Many real classics provide armies and tactical combat (Heroes of Might and Magic, for example). I much prefer that over the random-number crunching right now. But I wouldn't deny that it requires programming effort. We assume by default that the Civ developers cannot do it. We appear to have given up and to have put up with dumb AIs. Well, I haven't played GalCiv but people say that its AI is good. And now I read that its developers enhance it non-stop. Cool! Civ3 is a very popular game. And one of the most important parts of it is the AI. There must be some very big budget for it.

 

[warpstorm]

AFAIK, the budget for Civ3's AI covered one person part-time (he had other major duties).

(Admittedly, Soren is a good programmer)

 

[Shyrramar]

@Warpstorm. That explains a lot. I just wished they could see that AI is the single most important aspect of the game.

 

[warpstorm]

oops, wrong button

 

[Commander Bello]

I liked the system of ship design in MoO as well, but that was MoO, wasn't it?
For sure, I would not like to have to make my thoughts about to create and recreate my swordsman with a bronce knife, a bronze sword, a bronze two-hander, a bronze sword and a bronze dagger, two bronze daggers, a bronze sword and a long bronze dagger, a bronze sword and a short bronze dagger, a bronze sword and a flintstone knife, a bronze sword and a small axe .....and after I've researched shields, I'll do it all over again? You get the picture....

I agree, though, that this *could* be done with ships...

 

[Shyrramar]

@Bello: You have a way of getting it sound like the worst idea ever! Well, luckily there is a more positive way of seeing this. Those choices you listed, well, I am with you: it would suck. The point here is simply that the choices would be something like "a sword" or "a sword and a shield", not much more. I agree that this system doesn't really make much difference in the early ages, but could be more interesting later.

And what comes to that "doing it all over again"-part: it is a valid concern. One that they didn't take to notice well enough in MoO3... It could be done so that it caused no trouble: "add shields to these and these designs, take the needed space from..." But it is a strech, I agree.

EDIT: Nice to see you back, by the way. I missed your grumpy, albeit right to the point, comments

 

[Qpdaj]

I loved MoO2, also. Might have to go play a game or two, now (I think it's even still on this computer).

Not sure how such a system would work in Civ, though. It would be nice to be able to customize units in Civ (in the game, not just through modding), designing units that fit your play style better, etc.

 

[Commander Bello]

@ Shyrramar: thanks for liking grumpy remarks

Anyway, the system of creating own designs could be used for ships and armies, I guess.
For single units, it doesn't make much sense in my eyes.

But, if armies could be combined from single units more easily than under the current system, you might choose to have "light" armies consisting of 3 different units, "medium" armies of 4-6, and "heavy" armies of 7-x (maybe up to 10 or so) different units.
With your design, you could have offensive and defensive armies, defensive armies with good artillery support, slow but heavy offensive armies, fast offensive armies and almost everything in between.
An army would be the equivalent to a modern division then.

But still the problem would be - how to make the AI make correct use of this feature? How to make the AI counter the special designs of the bloody humans?

 

[vesuvius_prime]

Some ways to alleviate the micromanagement associated with the constant need for upgrades would be:

1. Upgrade ALL units of a given type to another given type.

2. Upgrade all units *within a given city* from a given type to a given type.

3. For ALL units that have a certain weapon, upgrade that weapon to another (more advanced) weapon (bronze swords/shields to iron swords/shields, for example).

4. For ALL units that wielding a certain weapon, add an additional weapon (like the pilum (the short throwing spear) of the Roman legionaries).

5. After you research some weapon, the game may ask you whether you want to redesign some unit(s) and would automatically take you to the Unit Design screen, if you choose to do so. And after you finish a design you may immediately order an automatic upgrade.

And, yes, there don't have to be hundreds of weapons (especially in the early ages). Swords, bows/slings, body armor, shields, bronze/iron (and the associated metalworking crafts), some special techniques (civ-dependent?) for edged weapons or spears/arrows, and the like would be enough for the foot soldiers, for example. That won't increase micro-management all that much.

There may be military sciences focusing on tactics, logistics, training. That would allow advanced army formations with attack/defense bonuses (i.e. pike wall - with shields that would become a phalanx; tortoise; etc.), supply lines (bigger armies operating deeper in enemy territory), more disciplined troops with higher morale, etc. But I'm not saying that all this should necessarily go into the game, neither am I suggesting that my ideas represent the ultimate wisdom in the Universe. Things should be thought out well. I just like more flexibility in my wars. Not just stacks of 20+ swordsmen which rely almost entirely on a random number generator.

Having the stacks of units grouped into armies would be great IMHO. And I definitely believe that armies should be upgradeable and it should be possible to remove units from an army and add other units into an army (as well as disbanding an army back into its individual units). What good is an army consisting of spearmen when there's infantry available? Why shouldn't we be allowed to upgrade an army? And why should we need a Great Leader to create the army? I will point out again that Great Leaders can improve an army but they should not be mandatory for its creation. After all, we know of many Bad (i.e. not Great) Leaders who lead armies and failed.

Also, I don't remember having problems with any game which offered me tactical battles. But that's me. There will be others out there who cannot stand tactical combat and prefer the current number crunching. And perhaps they won't like even the option of tactical combat, even if it can be turned off. Who knows. The battles don't have to be Shogun-Total-War-like. While I enjoyed Shogun & Medieval very much, I'm not pressing for a super realistic combat simulator. Some system like Heroes of Might and Magic 4 (without the magic ) would be sufficient.

It's a valid point, though, that it will take some effort to write a good AI which would create good armies. But is that so difficult compared to an AI that must formulate a global strategy? I believe a strategic AI is even more difficult to make than a tactical AI. And we already have a strategic AI, although admittedly not a superb one. If they really focus on AI for Civ4, they may hit the mark. I believe AI should be a top priority.

(By the way, I should be allowed to enter tiles occupied by my allies. Either to protect the allies, or simply to work the tile or pass by. I cannot stand it when an ally who has Right Of Passage steps and stops on my ONLY road to a given city, for example, and I have to waste whole turns to go around his units. In the case when I'm occupying a tile together with an ally and the tile gets attacked, it would be nice if my ally's units and my units get into the battle together, although his units may have to move automatically without me having control over them.)

 

[Shyrramar]

Good suggestions, vesuvius_prime! I think that your suggestions for upgrades are good, but may have to be tweaked a bit - there may be cases where your simple upgrades are not possible. I think you could do it quite easily, though. Just change the "upgrade all somethings to somethings", you could make it "upgrade all x:s, y:s and z:s to X:s Y:s and Z:s" or "Add x, remove y and z" or something like that, where the letters are assigned with a simple clicking system. It could be something like in the editor, where you choose the civs that the unit is available to, but without having to press down shift or ctrl, or something like that. It should be easy, IMHO.

 

[Crazy Jerome]

Actually, given a big gaming community, unit design is a lot easier to handle than AI or other things like that. You don't need code--just flexible game design. This isn't like AI scripting, because once a unit is designed, it's just a set of numbers in a file.

Specifically, design the game where the AIs don't come up with their own designs. Instead, they use "unit packs". These are modular files that include good designs for each era. Make sure that the game can handle multiple packs, random or preassigned packs, maybe even changing packs (reflecting a change in the CIV's attitude). The overall efficiency of the pack gets a rating. The player can change those if he wants, or he can depend on the judgement of the community as a whole and/or the Firaxis can periodically post the pack. Perhaps getting really crushed in a war causes a CIV to randomly change packs.

Firaxis and modders produce some really tough unit packs and also some deliberately substandard ones. Maybe playing on a lower difficulty level decreases the chance that the computer will use one of the higher rated packs. Best of all, there's the mysteries "things that they player has used in the past" pack, that randomly gets added to from the player's customized set (or maybe not so random--perhaps based on winning three games with a particular unit in the set). Occasionally, a CIV "designs" such a unit.

Maybe there's an option for those that don't like to design but do like the variety to have packs assigned to them, just like the AI. So with a few extra lines of code, suddenly you have the option to not know what unit you will get until you research the technology. It brings you down to the level of the AI for a more challenging game, but gives you something in return for handicapping you.

This is all ridiculously easy to code. There's no AI design there, except for some small unit pack pick routines to keep them from being completely predictable. It'd make for a much more exciting game without having to give the AI such an edge. Is Bismark kicking your tail in the Ancient era on Monarch setting? Is it because he randomly drew a great pack? Is it because he has a skewed pack that concentrates effective designs in the ancient era? If he gets ahead, will he get cocky and switch to a less effective pack in his complacency?

 

[warpstorm]

Crazy Jerome, that would work well. It's a shame that the previously mentioned games haven't made use of it.

One game that had an interesting twist on this was the RTS Kohan. In that game you never controlled individual units. You designed armies out of core units and support units and controlled the armies. Something like this could work.

 

[Arathorn]

I still don't like it.

The claim is made that a tactical AI should be easy to program -- easier than a strategic one. I have several responses....

- The strategic AI is so bad it needs incredible help to compete -- even at 40% cost factor, it loses.
- One of the main weaknesses of the AI is its tactics. Kill ratios of 10:1 are not too uncommon.
- I've never seen a reasonable tactical AI -- from Warcraft/Starcraft to the Civ series to MoO to any other game I've played (yes, limited sample size) -- they're all miserable and depend on overwhelming odds to have a chance. Kill ratios vary from 10:1 to 100:1 or more.
- Development time spent on a tactical AI is development time not spent elsewhere (like on a strategic AI). I'd rather have a well-rounded game than a purely tactical one (for the Civ series...I can have fun with a game that's mostly tactics, too, if that's what it's designed to be). If you claim the AI is trivial (and it's not, honestly), the time to set up screens and to code the interaction of different designs is very significant.
- Adding any in-depth tactical elements to the game will take it from a 30-hour experience to a 150-hour experience and I don't think that's a good thing.
- I'd rather see the game done right than reach for pie-in-the-sky asperations that aren't achievable and end up a horrible boondoggle. Tactical combat is a game in itself and either deserves full consideration or a simplistic approach. Full consideration would completely change (dare I say destroy?) the Civ spirit. Simplistic approach is the only reasonable option. Heck, they can't even get that done very well.

All in all, I still think this is a very bad idea.

Arathorn

 

[warpstorm]

Originally posted by Arathorn

- Adding any in-depth tactical elements to the game will take it from a 30-hour experience to a 150-hour experience and I don't think that's a good thing.
- I'd rather see the game done right than reach for pie-in-the-sky asperations that aren't achievable and end up a horrible boondoggle. Tactical combat is a game in itself and either deserves full consideration or a simplistic approach. Full consideration would completely change (dare I say destroy?) the Civ spirit. Simplistic approach is the only reasonable option.

You hit the nail on the head as to why I am against it.

 

[Padma]

You've got it, Arathorn and warpstorm.

BTW, I'm moving this thread to the Civ4 ideas forum. Makes more sense there.

 

[vesuvius_prime]

I agree that the tactical engine, if implemented, should be a simplistic one. I already said that it wouldn't make sense for Civilization to have a super-elaborate combat engine like, say, Shogun. But what's the problem with an engine like the one in HMM4? Turn-based, tile-based, initiative-based, with both missile and hand-to-hand combat units.

Besides, it does NOT have to take the game "from a 30-hour experience to a 150-hour experience". It has already been pointed out that normally it should be possible to turn tactical combat off. If you like random number generators, then okay - just switch to that. Then the overall combat time will be just like in Civ3. And you'll have a purely strategic game without a single piece of tactics. Just suit your taste.

It has also been pointed out already that if it's so difficult for the AI to create good packs, it can pick from a set of pre-defined packs which can be supplied both by Firaxis and by modders and can be constantly updated to reflect the players' discoveries. Also, after some tweaking, the obviously unbalances units will be edited and there won't be super-units or super-packs. Just pay attention to the balancing of StarCraft. The races are totally different but nobody can say that one of them is better than the others.

Arathorn, WarCraft/StarCraft are Real-Time Simulators. The AI required to create a good RTS is much more complicated than the code that moves units in tiles turn-by-turn in a grid of, say, 16x16 tiles. There are chess programs out there that can beat the world champion. It does take them quite some time on the hard levels to make a move but we're not talking about chess here either. We are talking of simple missile units and hand-to-hand units, mounted and on foot. That's it. I'm a programmer myself and I have a very good idea how to make a chess program. The AI there simply evaluates all possible moves (several turns ahead depending on the difficulty level), through a mini-max procedure that assigns a value to each encountered position and chooses the best one. The main difficulty is to create the function that will assess every situation and will give it a value.

Furthermore, RTSs are not only tactical engines. The tactical part only refers to the individual encounters -- what enemy to shoot at, where to move to, what weapon to use, etc. But RTS are also strategic engines because they must design an overall strategy from developing the economics, through developing the industry, through creating appropriate units and to sending these units to capture beneficial terrain and to destroy enemy bases. There are layers of AI there, and that AI is distributed. The general-level AI makes the strategic decisions and then leaves it to its subordinates to refine each task it gives them. And each of these subordinates defines a general plan to fulfil the task given to it by the general and then passes it on to its own subordinates. And so on till we reach the level of the individual soldier's AI which has no idea of the strategic plans made by the upper level. It receives very basic tasks and it solves them at its own level.

When I propose tactical combat for Civilization, by no means do I mean an RTS. Although Shogun and Medieval are very good hybrids of turn-based strategic games and real-time tactical games, I do NOT propose that for Civ. I think more along the lines of HMM4 and even then that should be adapted for the specific requirements of Civ.

And mind you again that if you hate tactical combat, you can just turn it off. It will save you the time that you don't want to lose. But have you ever preferred to turn off tactical combat in MOO2? It was possible but people didn't turn it off. It was better to personally command some battles. I've skipped many battles I was certain I could win. I simply pressed the "Z" key and it all ended in a second. But there were many battles where I wanted to take the command. I decided when and where. Tactical combat and unit design were great features of MOO2. MOO2 won't be a classic without them. Just imagine sending individual ships to bump into each other and to have a random-number generator resolve the outcome.

I know I'm repeating a lot of what has already been said, but that's only because most of the arguments in the previous posts have been largely ignored and I feel I could turn your attention back to some of them. Civ3 already has armies. What I'm suggesting is to add more flexibility to the armies (and their comprising units) and to add an additional tactical feature which you will be allowed to turn off.

 

[Crazy Jerome]

Well first, I mostly agree with Arathorn and Warpstorm, but I think we should distinguish the (at least) two different ideas in this thread:

1. Having customizable units -- This is worth having provided that the units can have enough variety to make customizing interesting. At a minimum, you'd have to multiply the current A/D/M values by some factor (at least 3-5) to allow some tweaking at the edges (e.g. bronze swordsman versus iron swordsman, but both clearly behind medieval equivalents). I think it would take more than that, but I also think that getting such a system correct is the main impediment to making customizable units worthwhile. IF that can be achieved, then custom units is minor and low risk--as I argued in previous post. This is a nice feature to have with tactical combat, but also useful in the strategic game. Who wouldn't want the fun of the units being slightly different each game? (And the minority that didn't could easily be accommodated by picking a "standard unit pack" that worked the same for each civ and the human.)

2. Tactical combat -- We all agree that we aren't talking RTS, or even complex/time consuming turn-based. So that leaves simple approach, probably something similar to Master of Magic had. It worked pretty well in MoM (small grid with some interesting terrains), but even that broke down against mildly competent human play, especially in the late game.

So assuming a simple tactical system is being considered, and unit customizing is in, there are two further hurdles:

A. I disagree that the tactical AI is a minor coding matter. It is doable. But it had better be a really good AI, because an 80% decent AI isn't good enough. A 80% effective strategic AI coupled with 80% effective tactical AI gets whipped every time. A 97% strategic AI is much more interesting. And we all know that last 20% takes the bulk of the work. Since the strategic AI can obviously be improved some more, this is not a trivial concern. Now maybe if the AI programmers think that they can provide a scripting engine such that the mod community can gradually improve the AI with player tricks, the opportunity cost is worth it. But it is still a huge risk for the developers.

B. Is a tactical side in the spirit of Civ? Or more to the point, given recent Firaxis comments about the direction of Civ4, does a tactical side enhance or detract Civ4. I'd say it detracts. They are looking to take out, what did Soren call it, "Whack a mole" pollution control and other essentially tactical concerns. They are looking to add "religion" as a first-class element, on par with government choice, culture growth, diplomacy, and trade.

Culture was a great addition to Civ3 not only because of what it was itself, but because of the way it enhanced the rest of the strategic play. It made temples and libraries less one dimensional. It affects military choices for borders/flips. I would expect "religion" to have a similar affect. Tactical combat, however, is just a "game within a game." It affects whether you win or lose, but not how you play the strategic game.

So a better question, I think, is what can be done to enhance strategic choice in combat? I'm not sure, though I think it would involve more army options, more combined arms options, and better strategic AI. I use "more" advisedly there, because a few "elegant, flexible, and reinforcing" options would be even better. Perhaps stategic battles would involve less drastic results, giving a large chance that a portion of the losing forces would gain experience and escape. Then when El Presidente of the Roman Republic makes his military choices and sends out his legions, he can live with the results of a simple clash on the strategic map.