Why Sid Meier gave up on stacking units?

[Oleksandr Sereda]

Older versions of Civilizations allowed stacking units and it was possible to protect some units, for example, you could protect swordsmen from cavalry protection by locating spearmen at the same tile. As a drawback, artillery (and maybe archers?) could do collateral damage to units in the defending stacks. This allowed some interesting tactics.

Later, this was removed and a tile could only allow locating a single unit.

I'm curious, why this was done? To me, it was more fun to play with stacking.

 

[Birdjaguar]

Civ 3 Stacks of Doom!

Hello and welcome to Off topic.

 

[Bugfatty300]

Absurdly large stacks of units were overpowered and made AI turn speed excruciatingly slow.

Instead of limiting a single tile to a set number of units (the reasonable balancing fix) they went with 1 tile 1 unit for some reason.

I suppose making warfare less fun and more tedious you'd encourage players to put more focus on civics and diplomacy which were sorely neglected compared to later games.

 

[tjs282]

Civ 3 Stacks of Doom!

And also in Civ IV?

 

[mitsho]

Yeah, I feel that‘s the wrong forum to ask that question

Was expecting some philosophical on the side going up on what 1-tile stack implies for history or somesuch. Instead the answere is purely gameplay related. Many players complaining about the Stack of Dooms. The unintended consequence of that move was however not miltarily, but economically: It zoomed the whole map a lot closer. This was then made worse by the introduction of districts in civ6 - now instead of controlling an empire, you feel like you just control a few cities that cover the whole landscape. A similar problem by the way for Humankind which went in the other direction with it‘s stacking (limited, and battles are on a separate screen).

 

[H4run]

becauseeee it's not what the cool kids suppose to do. Stacking is something that's popularized by Civilization, Paradox perfected and balance it with attrition system, but Civ abandon it entirely try to make work on 1UPT system, plus the hexagonal mumble jumble, instead of perfecting what is already work, they create a new problem instead.

I don't know. They should keep the good old feature while adopting the progress of the new, but here the decision is just not satisfying at all. If tend right, the product quality should add up, not dumb down.

 

[PhroX]

Absurdly large stacks of units were overpowered and made AI turn speed excruciatingly slow.

Well, getting rid of stacking didn't seem to do much about AI turn speed. 5 and 6, the latter in particular, are much worse than 4 on that aspect.....

 

[Gorbles]

I mean, ultimately, we're unlikely to know why Firaxis opted to move away from unbounded MUPT unless it comes out in an interview, or a dev blog, etc.

• Maybe it was appealing as a reset strategy; to go all the way the other way, and sort things out from there.
• Maybe 1UPT was simply a part of a larger tactical combat overhaul.
• Maybe MUPT was proving too much hassle to design and balance systems around in the long run.

I prefer the direction they've chosen to go. They're slowly reintroducing layers, seeing what feels right (see: religious combat in Civ 6), but it's far more immersive for me personally to think about the logistical approach of laying out an attack, as supposed to stacking 20 Locusts of Chiron on a single square and reducing enemies to dust in a single turn (yes, SMAC reference). You now protect your units by scouting more effectively and being more judicious with your movement. You now actually have to think about where you put valuable, high-risk units (like artillery, for example) instead of just hiding them inside a bunch of other things.

Sure, MUPT had some upsides, but a large part of it was that people were simply used to it, in my opinion. The more common counterarguments are "look at how badly certain game systems have done since 1UPT", which isn't really an indictment of 1UPT itself. It's simply more pressure on Firaxis to get those systems (AI, pathfinding, etc) to work better in future iterations.

Personally, I've thought a lot about capping the amount of units per tile, and in my opinion it's too arbitrary. Units in Civ are already an abstraction; a bit like unit counters in Risk (a board game, for the uninitiated). One hex of Swordsmen doesn't represent literally seven Swordsmen or however many it is. It represents a legion's worth, or however many is a fair amount to consider attacking a whole city. Capping it at n units isn't organic - it's saying "well okay 1 is too limiting, but infinite is too exploitable, so let's shove an arbitrary constraint that requires constant adjustment to work out". But we already have unit stats for that. So the problem is chiefly logistics, and that's why I really appreciate moving religious combat to its own layer (for example).

I don't think they should stop there, but I think any more layers are going to take a lot more thought into how it breaks up the existing combat, and its current strengths in terms of positioning (if you separate out ranged and melee, not only does it make little sense, but you lose the importance of keeping your ranged units back in tactical terms).

Maybe the solution is to bring back MUPT with the existing layers. A "best of both worlds" approach. I dunno. It's a bit of a segue, but I just want spycraft to be more exciting. Really go in on the layer approach. Give me Spy Wars!

 

[Birdjaguar]

Did 1UPT make the game less military oriented or at least point it in that direction?

 

[Gorbles]

It certainly gives it more room to breath, in my opinion, though early game rushes are just as prevalent. I don't see that as a bad thing tbh, even in the earlier Civ games (which was my main experience with MUPT), rushing and stacking units on tiles are pretty separate things a lot of the time. I believe (but can't remember / source) that the developers said it meant the terrain mattered more, with things like chokepoints actually having an impact. Of course, the flipside with that is you can be defeated by unfortunate terrain generation more than you could the other player.

It's also why I like (in Civ 6, which is my favourite of the two 1UPT titles in the franchise) the exploded city planning. I feel like it's a natural evolution of letting the actual civilising part of the game have some room to breath, and it gives some nice opportunities to slow enemies down by harassing their District output.

 

[warpus]

One of the factors has got to be that they never spend a lot of time on the "AI".. i.e. the computerized opponents you face in single player games. The "AI" is fairly dumb. It doesn't know how to react to changing circumstances, it only knows how to follow simple rules. When you increase the difficulty level, that doesn't make the "AI" any smarter, it just gives them more bonuses and gives you more handicaps.

One of the things that the "AI" is fairly bad at is combat. It's not easy to instruct a machine on how to fight a strategic war, especially if you insist on it following a set of simple rules. Strategy implies adapting to changing circumstances and at times throwing the unexpected at your opponent. If you only ever stick to rigid simple rules, you are never going to have a fun time fighting a war.

They got rid of stacks of death in order to make combat less complicated strategically. In part anyway. This way they don't have to even think about spending more $ or time on coding a good solid "AI", they can continue doing the bare minimum. Combat is now more complicated for the human player, so the studio can now celebrate and claim that combat is more "strategic", since humans are forced to be more strategic, instead of walking all over everyone with stacks of death.

So I mean, I don't hate this change really. Stacks of death should have had limits, like maybe 5-10 units per square? Building up an army of 100 units is way overpowered. So I don't mind the change really, it makes sense, but it's just annoying that you know that a big reason they did this was because they don't want to spend more time and $ improving the "AI". They're happy with the status quo there, and that's a shame because it's the bare minimum, if not worse.

 

[Gorbles]

I think I'd hesitate before throwing out assumptions on what the developers chose to do. It's certainly an explanation, but it hinges entirely on whether or not they did the thing to save themselves AI work. Which, of course, is impossible to prove. Folks either don't know, or are bound by NDA even if they do.

I disagree with it (predictably ), because, even with all the trouble they've had with it, they've rarely stopped working on it. And there have been noticeable gains.

What I think the problems with AI are, I'll save for another thread. It's a bit of a cop out, but the tl;dr is that AI programming is hard.

 

[GinandTonic]

It's kind of made the AI unable to pose any threat after the early game. In earlier iterations it was optimal to keep an eye on the sea, lest a doomstack appear from the fog. Now it's a joke. The AI is unable to launch any sort of offensive after the first sixty turns, perhaps 80 with a medieval UU they will actually build.

At deity, the AI is irrelevant once it's opening advantages are nullified. It cant cope with building districts, upgrading units or 1UPT. It just gets worse over time.

After a couple of thousand hours, I'm really ready for a sequel with functioning AI.

 

[Farm Boy]

That's exactly it. Achieving chokehold title lock is boring to do and tedious to maintain.

 

[TheMeInTeam]

Absurdly large stacks of units were overpowered and made AI turn speed excruciatingly slow.

Measured in actual real-world time, player and AI turn speed both got slower after Civ 4, not faster. Even if you play 4 on hardware contemporary to its time. For player turns, this difference is so large that I could manage 50 cities in Civ 4 faster than the fastest human beings on the planet could manage 30 in Civ 6. 4's UI is that much better, and since Civ 6 can't manage consistent input buffering, anybody trying to go faster with pure raw speed of inputs would be blocked by the game itself.

There are reasons one might prefer 5 or 6 to 4, but the claim "stacks/civ 4 is slower" is not valid.

~~~

Stack combat vs 1UPT at least has some tradeoffs where subjective preference is a factor. The AI is bad, in every civ game. Unless you had a significant tech lead, prosecuting offensive war in stack combat was incredibly dangerous, because of collateral damage to multiple units on the tile and opponent movement speed advantage letting them pick the engagement/guarantee access to that initiative (if they had an army).

Similarly, 1UPT allows for cluttering space around/between cities and having a significant firepower advantage despite lower troop count in a local area. This is partly because of the ridiculous introduction of cities getting their own attack and a ton of hit points. Regardless, the idea that it's impractical to fight anybody semi-competent at tech parity w/o catching them by surprise remains constant from Civs 4-6 (the AI does not fit the category "semi-competent" in any of these games).

The one advantage stacking had for the AI is that on deity, it could flood so many units onto a tile that it was difficult to damage them all with collateral. Similarly, its maintenance, upgrade, and tech boosts were enormous enough that it was playing a different game from the player. This isn't really a thing in Civ 6, but with enough bonuses to tech the AI in 6 could probably win space over most players, simply by having stupidly high city strength and completing all the stuff required at a discount faster than player can catch up in tech and win an offensive war. Would that make the game better/more fun? Don't know. Usually sheer brute force bonuses like "more health" or whatever aren't very engaging. But 6 could do it, if the devs wanted.

 

[Gorbles]

It's kind of made the AI unable to pose any threat after the early game. In earlier iterations it was optimal to keep an eye on the sea, lest a doomstack appear from the fog. Now it's a joke. The AI is unable to launch any sort of offensive after the first sixty turns, perhaps 80 with a medieval UU they will actually build.

At deity, the AI is irrelevant once it's opening advantages are nullified. It cant cope with building districts, upgrading units or 1UPT. It just gets worse over time.

After a couple of thousand hours, I'm really ready for a sequel with functioning AI.

While I'm fully onboard with future titles with improved AI, "a doomstack emerges from the fog" is not in of itself good AI

The problem here is often a bit semantic to be fair. Do people mean good AI, or do people mean challenging AI? Because good AI doesn't necessarily mean challenging, and challenging can often mean "cheating". The means justify the ends of the player experience, basically.

Or, to put another spin on it. Do players want an AI that can use features to the fullest, or do players want features that are enjoyable to use even if the AI doesn't use them (as / nearly as) well (if at all)? Because these are all product decisions that pull a game in different directions. It's hard to reconcile "good AI" with "advanced systems humans can use". The former will see more merit with people who play primarily against the AI, and the latter will see more merit for the (apparently small, but dedicated) player-vs-player crowd. But the former is actually a smaller demographic than it appears, because a lot of people who play against the AI won't be good enough to exploit its weaknesses like you and others can.

 

[warpus]

I think I'd hesitate before throwing out assumptions on what the developers chose to do. It's certainly an explanation, but it hinges entirely on whether or not they did the thing to save themselves AI work.

The devs never make decisions like that themselves, it's all passed down to them from the top. So I wouldn't blame the devs; the devs just do what they're told.

Given that the "AI" in Civ games has always sucked, and given that they never seem to care to improve the algorithms any.. and given that the "Higher difficulty? That just means more handicaps" method is still in place.. That all seems to point to the fact that they don't care to improve this part of the game. Removing the stack of death goes along too nicely with that approach. It makes things more complicated for the human player without them having to touch the "AI" code at all.

Let me also be slightly cynical and suggest that the newer iterations of Civ are also designed so that you don't expand a lot, like you could in previous games. They don't want you to conquer half the planet by 1500 A.D. They don't want to improve the "AI" algorithms to deal with human expansion, so instead they've put mechanisms in place to limit that somewhat.

 

[Gorbles]

The devs never make decisions like that themselves, it's all passed down to them from the top. So I wouldn't blame the devs; the devs just do what they're told.

Given that the "AI" in Civ games has always sucked, and given that they never seem to care to improve the algorithms any.. and given that the "Higher difficulty? That just means more handicaps" method is still in place.. That all seems to point to the fact that they don't care to improve this part of the game. Removing the stack of death goes along too nicely with that approach. It makes things more complicated for the human player without them having to touch the "AI" code at all.

I mean, the same applies. While it's certainly a solid theory that the removal of MUPT was dictated by higher ups wanting to save time and effort on AI development, I don't think that's necessarily the explanation is all. Otherwise why bother improving the AI at all? Which they have, notably, throughout Civ 6's lifespan alone (I wasn't as invested in patch news back in the Civ 5 days).

EDIT

Bit of a tangent, but I was interested. Now, take all figures with a large pinch of salt, but this is a search for AI programmer in games development in the UK:

https://www.glassdoor.co.uk/Salaries/ai-programmer-salary-SRCH_KO0,13.htm?countryRedirect=true

The first thing you notice is "not enough reports to show salary distribution". Sure, this could be a shortcoming of data entry on the site. But Glassdoor is pretty well populated on average. This backs up my general experience that AI programming is pretty niche.

The second is the ranges. From the four studios available (it then goes into AI researcher roles, etc) . . . they're pretty high. For games development in the UK. Games development is notorious for paying less than software development on average, and these are salaries that are competitive with a lot of experienced software roles (on the low end). They're all more than I was making until a year ago or so.

I'll save the rest for another thread, but like I said: AI is hard. This also means it is proportionately expensive to recruit for, compared to other roles. Correlation not being causation, of course, but I have some experience with AI, both in games and as an actual CompSci discipline.

 

[warpus]

Otherwise why bother improving the AI at all? Which they have

Have they though? From where I'm sitting they only ever do the minimal amount of work on the "AI" possible. Instead of designing something that can react to situations the human player might throw at it, they slap together a simple set of if-then-else situations and call it a day.

The computer players in this game have never been good, and that has never been something that the studio has wanted to fix. From their pov it's all peachy on this front. That's from the perspective of a player to be sure, but I have seen no signs of them ever wanting to make any sort of comprehensive improvements to the "AI". They just slap a bandaid on what they already have and move on.

I agree AI is hard. What we have now is not AI in any sense of the word. My point is that the studio wants to keep things this way, because it allows them to focus on other aspects of the game, such as the graphics. Only those who have spent some number of hours playing the game realize that the computer opponents are stupid, and that the difficulty level only ever means handicaps.

If you admit that this is a part of the studio's direction, then the lack of stacks of death makes a lot of sense. It goes hand in hand with that approach, IMO. It might not have been the catalyst for the change, but it sure seems like a big plus if you don't care to improve the computer players any.

 

[Gorbles]

Have they though? From where I'm sitting they only ever do the minimal amount of work on the "AI" possible. Instead of designing something that can react to situations the human player might throw at it, they slap together a simple set of if-then-else situations and call it a day.

The computer players in this game have never been good, and that has never been something that the studio has wanted to fix. From their pov it's all peachy on this front. That's from the perspective of a player to be sure, but I have seen no signs of them ever wanting to make any sort of comprehensive improvements to the "AI". They just slap a bandaid on what they already have and move on.

I agree AI is hard. What we have now is not AI in any sense of the word. My point is that the studio wants to keep things this way, because it allows them to focus on other aspects of the game, such as the graphics. Only those who have spent some number of hours playing the game realize that the computer opponents are stupid, and that the difficulty level only ever means handicaps.

If you admit that this is a part of the studio's direction, then the lack of stacks of death makes a lot of sense. It goes hand in hand with that approach, IMO. It might not have been the catalyst for the change, but it sure seems like a big plus if you don't care to improve the computer players any.

shrugs

I can't estimate the amount of work required for them to make the changes they have. I don't know how you're evaluating "minimal" changes. I can say for a fact that the AI in Civ 6 is more complex than if / else logic, and unfortunately this complexity trends to bizarre decisions at time which is where a lot of the "lol, AI" can come from.

Send me a PM if you want to carry this on - I'd rather let this get back to MUPT / 1UPT for a bit.