Say you're designing a Civ game. How do you make the combat system?

[DaveV]

In terms of fighting tactical battles, we already know from past games that the AI stinks at this, and will take disproportionate losses to the player.

Actually, I think the AI could handle your tactical map very competently. It's similar to chess, with a small number of pieces on a finite board. There are well-known ways of optimizing AI performance in such a setup. Assuming a good AI could be programmed, that would take care of your three levels of tactical combat:

1) Fight it out yourself against the AI.
2) Watch the AI fight with your units against the other AI army.
3) AI runs both sides of the fight, but no graphical display. This should look to the player like instant resutls. I've never played Dominions, but I think it works something like this, with the players being able to replay the fight in mode 2 if they want to see what happened.

[Gaspar]

*Sullla's wall of text crits you for 436986930634 points of damage* I understand the lack of visuals, but probably some color or some such would have been useful.

Regardless, interesting thoughts. As someone who hasn't play MoO, I find a lot of that stuff interesting. I'm not sure if I like the tactical map concept, but I definitely like the idea of having an army of 50 spearmen, instead of 50 spearman units stacked on a tile. Its sort of how I always imagined SoD in my head anyway. I'm not sure if I like the combat system or not, but I am sure I'd rather play that than 1UPT nonsense from Civ5.

With regards to the rest, one of the things they removed from Civ5 that I really like in Civ4 is "first" benefits. While I agree with using the FfH model of "building" the new religion as opposed to getting to Code of Laws first, bam, I've got Confucianism, I always liked the ideas of Music first = Great Artist, etc. Its a good way of making some mediocre techs more desirable, and it creates an interesting decision, namely, do I beeline, say Music, so I can get the free GP, or do I focus on the more valuable techs and let someone else get there. I do think Liberalism is a bit too much of a "One right choice" thing, but regardless, the idea of benefits for first to a tech make the tech tree more interesting. I think your tech tree as you outlined it could use some real spicing up - it seems a little boring. That may be because I never played MoO though.

The stuff about tile yields and expansion - spot on. The worst consequence of 1UPT was the sacrificing of tile yields, it makes settling on resources a "one right choice" as opposed to an interesting decision, because there's little sacrifice. I love the idea of mine/farm yields increasing like cottage yields as time goes on.

I also don't really understand the idea that you or anyone else has to apologize for thinking Civ4 should be the basis for the next title in the series. While its a game with problems, like any other, its so clearly the best Civilization title to date, I really don't understand how anyone can come to any other conclusion. The first 3 Civs were plagued with not being able to come up with a way to deal with the biggest issue facing Civ games - how do you make expansion a choice rather than a "right now, right away, as soon as you can, as much as you can" decision, Civ4 fixed this, then Civ5 scrapped it and ruined it again. Can the Civ4 model be tweaked/improved? Sure. Is it the best one the series has found to date? Unquestionably.

Well, interesting read nonetheless, Sullla. I'm not sure if you'd make a great Civ6 designer or not, but I'd be more interested in playing your design than any of the recent ones.

[Sciz]

A problem I see with the combat is the way the combat would be "calculated" if you don't want to fight it out yourself. As not fighting it out yourself would probably be the most popular for MP, the way it's calculated has to be very precise and non-random. Maybe that's something very easy to do, but for me, someone who has no experience of any programming of any sort, it sounds like a hard thing to get right. And it's an integral part of the game. But as I said I have no idea what i'm talking about :P

[luddite]

I agree, I think that would be a great game. Civ 4 and MOO were both great games, no reason not to borrow from them heavily (and civ 5 had some good ideas too). I've always liked the design philosophy taken by Blizzard, which specifically doesn't try to innovate too much, but simply takes models that have been proven to work well, and polishes them to a high degree. Nothing wrong with innovating and experimenting, but a flagship series shouldn't be experimenting too much.

I'm not totally convinced about the necessity of having a separate tactical screen for battles. It might work well, but if you're already going to allow unlimited stacking of the same unit, then do you really need it? If you do have it, then I'd suggest that for MP all battles could get resolved at the end of the turn, and put a turn limit of like 10 turns on each battle. I'd be really annoyed at getting a message saying "the AI has calculated that you lose" if I knew that with proper micro my army would win.

I'm also worried about having all tile improvements and some civics get better over time. I like the contrast in civ 4 between cottages that gradually become towns, and other improvements that just start off strong. It gives you some freedom to change your tiles later in the game. If every tile needs a long time to become effective, you might be stuck forever in whatever configuration you choose early on.

Anyway, a lot of good ideas there. The basic feel of creating a huge empire with huge armies is exactly what I'd want from a 4x game. Too bad it'll probably never happen .

[luddite]

Oh and here's a question I thought of- would the city towers have unlimited range, like MOO? Or limited, like civ 5? If it's limited, they might be too weak, because seige units could outrange them and make them almost useless. But if it's unlimited, it might feel a bit strange to see a city tower shooting across the map. I guess it would make sense if they were pictured like defensive catapults or something.

[SevenSpirits]

Thread necro! If were designing a Civilization-style game, here's how I would do it: http://www.garath.net/Sullla/designingciv.html

To be honest, if I had any say in the matter (and it mattered), I'd keep you strictly on playtesting / development and far away from design.

* There are close to no new ideas in your proposal. It's mostly just a mash-up of Civ IV and MoO.

* What things there are that are, in a sense, new, are really just heavy-handed tweaks to old. For example, your maintenance system (which is practically equivalent to Civ IV, but changed anyway in a manner which then needs to be counteracted), or the (admittedly somewhat of a placeholder) culture slider penalty once you reach a certain culture level. Really?

* I have to wonder if you've given much deep thought to these things at all. For example, using the reason of 21 BFC tiles instead of 19 as the main justification for square tiles instead of hexes?? Seriously, a two tile difference? (Which you miscounted as three tiles.) Or, the fact that you didn't mention how screwed up early game scouting is going to be with 1-turn units fanning out in all directions. The way you just ignored the main problem with big empires, which is that, barring some mitigating game mechanics, that player has already won! Is the design goal of this game to continue long after the winner is clear?

The whole things reads just like one of the many poorly-conceived mods for Civ IV, dreamed up by some random guy, which make a few drastic (and generally negative) changes while leaving most game mechanics precisely as they were for reasons of expediency and perhaps a lack of imagination.

That doesn't mean that a child of MoO and CIV is a bad idea, just that this one lacks vision and basic game design principles. Though, I admit that I probably wouldn't like it much anyway, since I prefer smaller-scale stuff, and my favorite part of CIV is the tech tree with multiple different things available at each tech.

I say this with love and the knowledge that you can take criticism well, and I recommend you stick to parts of game making that aren't design.

[luddite]

Yeah I have to agree with seven about the maintenence system, which seems needlessly complicated. Do we even need city maintenence at all? MOO doesn't have any, and that game never suffered for it. Just accept city spam as inevitable, and go with it. Maybe balance it by taking away the free center tile, so size 1 cities would be really weak.

for culture- i dunno. I agree with Sullla that it's BS to win a cultural victory by ignoring culture completely for most of the game, and then just pumping culture like crazy at the end. On the other hand I also agree with seven that I don't like the idea of penalizing the culture slider at arbitrary levels. To be honest, the whole idea of "cultural victory" always felt arbitrary and unfun to me, so I wouldn't be sad to see it scrapped completely. I liked using culture to unlock social policies in civ 5.

[zakalwe]

I disagree pretty strongly with the criticism posted by SevenSpirits and seconded by luddite here. My impression was quite the opposite; rather than inventing all sorts of crazy new game mechanics, Sullla seems pretty level-headed and keeps a lot of things unchanged that are already working well. And if he were the designer I think he'd be perfectly willing to drop things that just didn't turn out to work well in testing. I agree some of the design is a somewhat "obvious" mash-up of Civ4 and MOO, but I thought some of the ideas are genuinely good, such as lightbulbing holes in the tech tree, and selecting any 5 civics (maybe some world wonder could expand this to 6?). To me, a good sequel is instantly recognizable as the "same game", but with lots of clever tweaks like that, and a select few entirely new features to expand the scope. (I.e. new tech system and tactical combat, in this case).

The scouting issue isn't hard to solve. Just add a little more barbarians. Plus, what's so bad about scouting anyway? I actually think it's a bit "unfun" that you pay such a high opportunity cost in civ4 just to get going with the scouting a bit sooner. After all, revealing the map is a lot of the fun in the early game.

[T-hawk]

Or, the fact that you didn't mention how screwed up early game scouting is going to be with 1-turn units fanning out in all directions.

This is a specific outcome of a more general difference between MOO and Civ: the nature of the game board. The map in MOO comprises a set of nodes as stars. The map in Civ is a grid of tiles.

Stack-based movement works in MOO because the map contains such a limited set of locations where ships can exist. If you want your fleet to be at Altair, they simply go to Altair; there's no notion of existing in Altair's countryside or defending Altair's front. Combat happens between hostile stacks that are both trying to occupy the Altair node. Micromanagement is tolerable because the ships at node X are probably all moving together to node Y, or at least to a very small set of possible destinations.

But how MOO stacks are built and controlled doesn't translate to Civ's grid map so well. Scouting with cheap single units is just one ramification of that. I like what you're trying to do with say 1000 spearmen, but think of trying to defend a 20 tile front with them. Besides the infeasible micromanagement, it blows up the stack-of-doom problem even worse (defender must defend everywhere, attacker can attack one point), allowing the attacker to engage a very small fraction of defending forces at each step and sustain a dreadfully high kill ratio.

I see a solution to these problems in borrowing from Heroes of Might and Magic. HOMM has the same general idea as MOO and New Civ: large quantities of units (often 1000+) abstracted into stacks for tactical sub-combat. But HOMM has a tile based map like Civ. How does it do it? Units never move around on their own; they must be led by a hero. HOMM imposes a nodal strategic system on a tile based map; heroes (mobile nodes) are expensive and a limited resource. Civ could steal this concept, that armies are only formed and led by a general unit which would be moderately expensive in hammers. That solves the scouting problem, and also the zone-defense stack-of-doom problem: one hero/general could gather troops from many cities to meet an invading army.

Also, HOMM improves the micromanagement situation by producing units not every turn, but at fixed intervals of every 7 turns. So it's quite tolerable to send a hero around to your cities to collect troops at appropriate times. I'm not sure if that flavor maps perfectly to Civ (produce troops once per century?) but it may be workable.

[novice]

Sullla, reading your article reminded me what a great game Master of Orion is. A "merger" of Civilization and MoO sounds fun (although I guess that's what the later MoO versions attempted and got tragically wrong).

One aspect of MoO (and Alpha Centauri) that I really enjoyed was designing your own units based on the technology you had available. Granted, it's kind of a game within the game, so maybe "expansion pack material".

Well, we can dream, right?