No 1UPT

[Gort]

Yet with unit stacking warfare is even more dull - tactical maneuvering practically dies as wars turn into "who has the bigger army?"

Not if you discourage stacking (as opposed to banning it). You want a middle ground between "put all your units on one tile" and "you can only put one unit on a tile". I believe that middle ground can be achieved either through stacking penalties (loads of wargames use these, imagine a 10% strength penalty per unit after the first on the tile - two units sharing a tile are 10% worse, three units sharing are 20% worse etc) or through collateral damage (where units sharing a tile all get damaged or destroyed when one unit gets damaged - Civs 1,2 and 4 used flavours of these).

Basically you want a system where you can get the job done by ramming 100 units down someone's throat if you have overwhelming production superiority, but the efficient way is to use only as many units as the terrain allows. That way production gives you an advantage, but using your army efficiently might let you defeat that advantage.

That way you get the tactics side of the Civ 5 1UPT system, as well as some of the AI-comprehensibility and general usability of the Civ 4 stacking system.

sorry.

No worries.

 

[Ryika]

So, AI can't handle 1UPT properly... and the solution to that is... to add a stacking system that penalizes too big or too small stacks? How the hell does that work? It would only make it even harder to get the AI to use that system properly.

 

[Acken]

I've always been a proponent of "fix the balance first" when trying to make an AI be good at something.

One of the biggest issue AI programmers face is that they have no idea a priori of what is good and what is bad. It's very easy to criticize some stuff a posteriori like: the AI make too many melee/siege, the AI doesn't use scientist slots etc etc.

I have more experience with civ5 so I'll talk about that for an example. The AI makes too many melees that get mawed down by the player mass of range units. While making the AI move units around more efficiently is also a goal in itself (and many modders have criticized things like the pathway algorithm) the AI would be more threatening if Melee were better or range less effective.

 

[Galgus]

I've always been a proponent of "fix the balance first" when trying to make an AI be good at something.

One of the biggest issue AI programmers face is that they have no idea a priori of what is good and what is bad. It's very easy to criticize some stuff a posteriori like: the AI make too many melee/siege, the AI doesn't use scientist slots etc etc.

I have more experience with civ5 so I'll talk about that for an example. The AI makes too many melees that get mawed down by the player mass of range units. While making the AI move units around more efficiently is also a goal in itself (and many modders have criticized things like the pathway algorithm) the AI would be more threatening if Melee were better or range less effective.

Agreed, even in Beyond Earth ranged units feel too durable when met with melee combat.

Generally I'd like to see ranged units die in one attack for minimal damage to a melee unit to force players to protect them and position them well - with exceptions like the epic units.

 

[Hail]

Not if you discourage stacking (as opposed to banning it). You want a middle ground between "put all your units on one tile" and "you can only put one unit on a tile". I believe that middle ground can be achieved either through stacking penalties (loads of wargames use these, imagine a 10% strength penalty per unit after the first on the tile - two units sharing a tile are 10% worse, three units sharing are 20% worse etc) or through collateral damage (where units sharing a tile all get damaged or destroyed when one unit gets damaged - Civs 1,2 and 4 used flavours of these).

I favor a "capacity approach":

• each unit type has a capacity parameter.
• each terrain type has a capacity parameter.
• coef = sum of units' capacities on a tile divided by tile's terrain type capacity parameter.
• ranged bombardment, collateral damage, etc. is multiplicatively increased by this coef.

So, AI can't handle 1UPT properly... and the solution to that is... to add a stacking system that penalizes too big or too small stacks? How the hell does that work? It would only make it even harder to get the AI to use that system properly.

However, symmetrical single-player games need to be designed as much for the artificial intelligence as for the humans themselves. Even if painful, designers must be willing to leave some of their most orthogonal – and often most creative – ideas off the table for the sake of the AI. Game design is a series of trade-offs, and empowering the AI is important for avoiding the downward slope of the Parabola.
Nonetheless, creative developers can solve this problem at the design stage before it even reaches some doomed AI programmer. One game mechanic that pushed Chick over the edge with Empire: Total War was amphibious invasion. The AI was simply incapable of coordinating its land and naval forces together to launch a coherent and effective invasion of an overseas target. Smart players would quickly learn that if the AI could not attack amphibiously, then the strategic balance can be gamed easily. Maybe England’s troops are not such a threat after all?

This problem is not unusual; strategy games with transportation units almost always suffer from ineffective artificial intelligence. Coordinating land and naval units to be ready in the same place and at the same time – along with the necessary escort ships – is a non-trivial task.

Rise of Nations, Big Huge Games’s historical RTS, presented a blunt but effective solution to this problem; land forces which approach the shore simply turn into boats to carry themselves across the water. Once they reach their destination, the boats transform back into the original land units. No transportation ships ever needed to be built or managed at all.

With one simple stroke, Brian Reynolds, the game’s designer, removed a classic AI problem from the game, enabling water maps to remain interesting for veteran players. The design may have sacrificed the “realism” of requiring the player to build transport ships along with other naval units, but the upside was extending the game’s longevity significantly.

from here.

I've always been a proponent of "fix the balance first" when trying to make an AI be good at something.

One of the biggest issue AI programmers face is that they have no idea a priori of what is good and what is bad. It's very easy to criticize some stuff a posteriori like: the AI make too many melee/siege, the AI doesn't use scientist slots etc etc.

• focus on MP.
• put together a working prototype asap.
• 1. play it (devs + some selected individuals).
• 2. do a balance run.
• 3. add stuff.
• 4. go to 1.
• feature lock.
• write an AI.
• ???
• PROFIT

 

[SupremacyKing2]

Yet with unit stacking warfare is even more dull - tactical maneuvering practically dies as wars turn into "who has the bigger army?"

With unit stacking, I definitely think more should be done on the quality aspect of your units so that quantity is not always better. Ideally, the game should balance quality vs quantity so that the player has what Sid Meier calls an "interesting choice". For example, you could have more levels of training that the player could purchase with gold so that small but rich civs could have small but stronger armies. This way, the smaller stack of well trained units might actually be stronger than the larger stack of untrained units. You could also introduce a stacking penalty to each unit that is stacked (this would represent the challenges of managing, ordering, organizing a large army into a coherent fighting force). This stacking penalty could be reduced with a general or great general unit. This way the player with a very large stack would need to invest a general or great general unit. They could not just throw a large stack against an enemy and expect to always win since without a general unit, the large stack might get such a severe stacking penalty that a smaller stack might actually win. This would also encourage players to split up their stacks into several smaller stacks instead of having just one single huge stack.

 

[Gort]

So, AI can't handle 1UPT properly... and the solution to that is... to add a stacking system that penalizes too big or too small stacks? How the hell does that work? It would only make it even harder to get the AI to use that system properly.

OK, I'll explain.

Let's first define the problem.

The current issue the AI has is that it is very difficult for it to get units from A to B when they're all getting in each other's way, especially when terrain is restrictive. If they also have to worry about stuff like getting the pikemen in front of the archers when they arrive at their destination, or the fact that only melee units can capture cities, and that they're being shot at every turn they dither and try to rearrange their formation, and they don't know how to fire and move in the same turn, and your difficulty settings give them larger armies through low upkeep and bonus production, it quickly becomes a nightmare. The result is that the AI appears to be very stupid when it comes to warfare.

The obvious solution is to simply allow the AI to stack all its units in one stack and march them from A to B, since this requires the least pathfinding logic, has no risk of unescorted units being picked off, and interacts least with troublesome terrain. This would be the solution that the AI could handle most easily.

However, the obvious criticism of that solution is that it's boring when you just blob all your units together and run them into the blob the enemy have built, and whoever has the best blob wins.

Therefore the solution is to compromise somewhat. Allow the AI to blob up if it needs to, but put a penalty in place so that the optimal strategy is to spread out as much as possible. This retains the element of strategy in warfare so it's not just "best army wins", but also allows the AI to have an easier time of going from A to B.

And as a tiny aside, it also means you can trust the pathfinding to not do dumb stuff like go offroad the moment two units share a tile, and you don't get weirdness like how it's impossible for two archers to occupy a city, but an archer and a ranged boat is just fine, or how Isaac Newton and JS Bach are too big to both fit in London.

 

[Ryika]

Well, okay. I see that line of thought. But still.

How do you make sure the AI can't just be ambushed while it's in "Move units"-mode and lose its whole army in a single turn? How do you make it transition out of "Move units"-mode at the right time? Once again, how do you teach the AI to use the right stack sizes and how do you get it to find a midground between centralized and spread out army stacks?

That whole idea is, while certainly helping with the problem you want to solve, introducing SO MANY new variables that would hinder the AI, I really don't see how it's a good solution. Why not just make the grid smaller and give units more movement points to compensate? That way, setting up and maintaining a formation would by default become much easier.

 

[Gort]

Well, okay. I see that line of thought. But still.

How do you make sure the AI can't just be ambushed while it's in "Move units"-mode and lose its whole army in a single turn? How do you make it transition out of "Move units"-mode at the right time? Once again, how do you teach the AI to use the right stack sizes and how do you get it to find a midground between centralized and spread out army stacks?

Those are already problems with the current system - not new ones that unbanning stacking would introduce. 1UPT units are already ambushed enroute to places, embark next to cities and get one-shotted, all that stuff.

I would argue that the default state the AI is best at dealing with is limitless stacking with no stacking penalties, and every restriction you place on stacking will make the AI play worse. Enforcing 1UPT is clearly a bigger restriction on stacking than a stacking penalty, so the AI plays worse with 1UPT than it would with stacking penalties.

That whole idea is, while certainly helping with the problem you want to solve, introducing SO MANY new variables that would hinder the AI, I really don't see how it's a good solution.

I feel that removing the ban on stacking actually means one less variable for the AI to keep track of. There must be thousands of instances per game where an AI is prevented from moving the way it wants because of the stacking ban.

Why not just make the grid smaller and give units more movement points to compensate? That way, setting up and maintaining a formation would by default become much easier.

I have no problem with the idea of a higher tile-to-unit ratio and faster units - there are certainly different solutions that can be used - though faster units and more open terrain might actually make it easier to reach the vulnerable units that easier formation-rearrangement would help you protect, leaving you in the same situation.

 

[nimling]

The AI being unable to handle 1UPT is one thing, but my point was and is that 1UPT and ranged units are bad for human vs. human matches, and leads to playing the game in a very tedious way. This is especially apparent under the rules most MP matches play by, but even under ideal playing conditions, 1UPT fails in many, many ways.

I'd rather just have the old stacks of doom, and ranged combat that wasn't ridiculously unbalancing and boring, than the current system.

 

[Gorbles]

Removing a ban on stacking increases the variables to consider because you increase the available units in any consideration, Gort. Let's not cherrypick individual variables here.

nimling, you still haven't given any reasons. You've just said 'this is how it is'.

Also, stacks of doom were imbalanced, and please never use 'boring' in a serious balance / design debate. It's a valid personal feeling, but means nothing relevant in discussions like these.

 

[Gort]

The AI being unable to handle 1UPT is one thing, but my point was and is that 1UPT and ranged units are bad for human vs. human matches, and leads to playing the game in a very tedious way. This is especially apparent under the rules most MP matches play by, but even under ideal playing conditions, 1UPT fails in many, many ways.

I'd rather just have the old stacks of doom, and ranged combat that wasn't ridiculously unbalancing and boring, than the current system.

Civ 4's stacks still had a race to see who could collateral damage the other stack first. I don't think this problem will be properly solved until the game enforces turn-by-turn warfare - where one side gets to move all its units in the warzone first, then the other side goes. They went some way towards this in Civ 5 by introducing hybrid turns, but unfortunately the prohibition on doing anything while you waited for your turn was too crippling to be worth it.

The way I play Civ 5's multiplayer when at war with another human player is that the aggressor moves all his units, then says "Done" and then the defender moves all his. Both players agree not to do anything with the units involved in the war when it's not their turn, but can mess around with research trees, what they're building, and things like diplomacy while their opponent moves. Works out well - and it's what I'd like to see the game rules enforce on their own.

 

[TabeticClown]

Honestly, 1upt is my favorite change that Civ V implemented into the series. I simply cannot fathom how anyone would prefer stacks of death, it's tedious and anticlimactic and takes no skill at all besides what it took to build the units.

It sounds like the OP has a problem with 1upt purely because range units are too powerful, which makes absolutely no sense whatsoever. If range units were too powerful then why not buff melee so that they can range attack cities kind of like in AOE, or nerf range to do less damage or something. I think they're overstating the improper balance between the two classes. Melee units are still useful and necessary.

 

[nimling]

How many MORE reasons do I need to present? This IS how it is. Try playing MP matches in Civ5 or BE under the commonly accepted rule set and the issues are even more apparent, whether you play with simultaneous or sequential turns. It might be easier to not see or ignore the issue against the AI (quite frankly I don't bother to micromanage wars against the AI, because the AI makes so many blatant errors and the AI can be manipulated diplomatically anyway).

A large problem with 1UPT is that it was built on broken ranged units being a core part of the game, and the reasons why those ranged units are unbalancing should be obvious.

The major issue with Civ4 wasn't even stacks of doom, but that stack killers were so strong and city raider promotions made a city a death trap, which would often victimize the AI. A problem like this would be resolved by changing combat from unit vs. unit to army vs. army - many strategy games operate on this premise, the idea of each unit fighting one after another is kind of silly. Some games just abstract the stack, others use tactical combat like MoO (or simulated tactical combat, an option I used when I got bored of tearing apart MoO2 in tactical combat).

Yes, it means building a bigger army will win most of the time; but that is how the game should operate, and 1UPT in CivV doesn't really change anything except draw out the process with more micromanagement. Instead of a stack of doom, it's a carpet of doom and a matter of who can spam the brokenly powerful ranged units for risk-free damage the best, with a small percentage of mobile units and smaller percentage of melee blockers. The only difference is that someone who can micromanage units can win through cheese tactics, which isn't fun or a mark of strategy, but a mark of who can put up with needless frustration better.

I'm actually shocked that the majority of people here are defending 1UPT, when just about everyone who isn't a civ player looked at 1UPT and realized that it was no good just after a cursory look at the game. It is the primary reason why my brother (a long-time Civ player that skipped CivV but was really interested in BE) isn't touching BE.

I started this thread under the presumption that 1UPT being kind of or very crappy was a well-understood given, and that people would be interested in hypothetical alternatives. It quickly got derailed into a thread on the merits of 1UPT (there aren't any, really) and has gone nowhere. I'm still campaigning for a fix to 1UPT, whether that comes from Firaxis or someone in the modding community. So far it's hard to find a useful MUPT mod because of the design decisions mandated by 1UPT, and how the AI has had to adapt to them - trying to code AI to MUPT on top of a game designed for 1UPT is really hard for a modder, especially given that CivV is AFAIK a much harder game to mod than IV.

Most likely though, a fix will have to wait until Civ6, and if Firaxis holds on to 1UPT I'm not buying it until the price drops dramatically.

 

[Lord Tirian]

I'm actually shocked that the majority of people here are defending 1UPT, when just about everyone who isn't a civ player looked at 1UPT and realized that it was no good just after a cursory look at the game.

That is anecdotal, that is your experience. My girlfriend never played Civ before Civ5 and loves it and quite dislikes the stacking in Civ4 (which I preferred at the time, it was only G+K and BNW that brought me around).

1UPT has issues but stacking has its own. Different people have different trade-offs, different preferences.

This said, I would quite like to see Amplitude (Endless Space, Endless Legend) try to take on the historic 4X, they are trying out clever and interesting things with armies, i.e. limited stacking for combat purposes. I just can't see Firaxis doing it for Civ:BE or even Civ6 at the moment.

 

[Ryika]

A large problem with 1UPT is that it was built on broken ranged units being a core part of the game, and the reasons why those ranged units are unbalancing should be obvious.

Still a silly argument. "Unit X is op and therefore the whole system doesn't work." - the reasoning is just flawed. That's like saying a combat system with stacked units can't work because in Civ 4 Catapults were stupid. Not that I expect you to even give this an honest thought that at this point.

I'm actually shocked that the majority of people here are defending 1UPT, when just about everyone who isn't a civ player looked at 1UPT and realized that it was no good just after a cursory look at the game.

It's almost like this game is giving its audience what its audience wants instead of listening to the people who aren't even buying the game. You know... almost like they're doing what any developer should do.

I started this thread under the presumption that 1UPT being kind of or very crappy was a well-understood given, and that people would be interested in hypothetical alternatives. It quickly got derailed into a thread on the merits of 1UPT (there aren't any, really) and has gone nowhere.

See, that's really the core issue, isn't it? "This is supposed to be a discussion where you either agree with me or you're stupid." is bound to fail. You didn't honestly try to argue for anything, you just assumed that your opinion is "ultimate truth" and even now, that you've seen how many people disagree you've still not even considered the arguments that are made in any other way than "You're wrong, I'm right!".

So far it's hard to find a useful MUPT mod because of the design decisions mandated by 1UPT, and how the AI has had to adapt to them - trying to code AI to MUPT on top of a game designed for 1UPT is really hard for a modder, especially given that CivV is AFAIK a much harder game to mod than IV.

So what exactly is stopping you?

Most likely though, a fix will have to wait until Civ6, and if Firaxis holds on to 1UPT I'm not buying it until the price drops dramatically.

I'm glad to hear that, but I assume that it probably won't stop you from arguing the same issue in the Civ 6 forums, even a year after Civ 6 has been released while the majority of people rather prefer talking about how 1upt could be improved further, right?

 

[nimling]

I'm not responding to the above post with anything other than *sigh*.

Moderator Action: Please do not add post responses that include no real content.
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

 

[ZTZaorish]

1UPT is not inherently flawed. I can tell you one of mightiest TBS games of all times (and i mean _ALL_ times), which is
- highly praised for its replayability,
- complexity,
- general brilliance,
- and is - you got it - 1UPT.

You certainly know it: Chess!

The real problem with Civ:BE (and Civ5 alike) is that units lack range. You see, in any 1UPT game, for military excercise to be really fun, interesting and rich, most units _must_ have either the ability to make ranged attack in their turn for many tiles, or move during their turn for many tiles, or both.

You see, in chess, only pawns are 1-tile-range "attack" and either 1 or 2 tiles range movement per turn. All other "military" units of chess (which is everything else except the king) - move and attack 3 (horseman) or up to 8 (other figures) tiles per turn. In Civ:BE terms, all chess "units" are melee units, of course. No true ranged ones (which attack from afar and stay afar).

It has quite very simple explanation, this "for 1 UPT, most units must be able to attack and/or move many tiles per turn". You see, the more range per turn your units have, the more _choice_ there is; game becomes more and more tactical. The higher range (in tiles) the bulk of players' forces are able to move/attack per turn, the more variation and complexity can be created in any given particular battle.

On the other hand, make the range for units too high, and things get _too_ complicated, inconvinient, awkward.

This same mechanic of 1UPT is also true for RTS, and you can see how it works in Starcraft, for example, based on ranged units' performance there. Only on ranged, because it's impossible to define how far movement per turn there is - since there are no turns. But ranged units still present "how far we can fire at any given _instant_" ability of theirs - and while there are no "tiles" there either, you can still count "space which is occupied by a single ranged unit" as being a "tile in terms of this kind of units".

For example, when you make, say, a dozen hydralisks in Starcraft 1, you can clearly see that their attack range is more than a dozen times larger than width of the space a single hydralisk occupies. For dragoons, the ratio is lower; for marines, is higher. But overall, lots and lots of RTS/TBS/tactical games you check - you'll see the same picture: often 1UPT in some or other form is used, and often most units move/shoot over a dozen tiles (or "tiles" virtually).

In Civ:BE, max ratio you get is 4:1 with SABRs for ranged units, and up to ~8 with naval units and tacjets. Too low. The only exception to this is fighting within completely magrail-covered area - there, the ratio for melee units increases to appropriate values. However, ranged units remain unable to fire that far (they can _move_ that far, but can't fire from that many tiles afar). And anyhows, who would build magrails over whole continents?

So you see, both Civ5 and Civ:BE has that flaw; i call that not "1 UPT problem" - i call that "too slow / too short range units" problem.

The solution? Rather simple in theory, but of course i doubt Firaxis would implement it... Still, here goes.

It would simply be enough to introduce "combat tiles" into the game, each "combat tile" being akin to a simple square on a chess board, but in Civ:BE that's of course hexagonal combat tiles; and the size (area) of each combat tile would be exactly 1/4 (25%) of the current in-game tile. Normal game tiles would remain for means of tile improvements, citizens working them, satellite coverage, etc. But for means of keeping 1UPT rule operational, those new 4 times smaller combat tiles would be used, and all units would need to be rendered, correspondedly, 2 times "shorter" and "thinner" (2x2=4) to fit those tiles perfectly. The latter part, i find rather desirable, personally: when i see a Xeno Titan which is larger (literally) than a settlement which he attacks, i start suspecting something's not quite right there... %)

This new sub-grid of combat tiles fits old "normal tile" grid perfectly: you make a hexagon "combat tile" which has all 6 of its sides being exactly 2 times shorter than sides of a normal tile, and then you put this smaller "combat tile" exactly into the center of every large tile. The remaining space - it's easy to see if you draw few "big tiles" and then "combat tiles" on a piece of paper, - will easily form additional "combat tiles" of the same "4 times less space = 2 times shorter side" size. Each normal tile will end up having one smaller combat tile right in the middle, and 6 more "halves" of combat tiles at each side.

Obviously, refinement of ranges for units would likely be needed, but roughly, their ranges would remain of the same scale, but now expressed in terms of combat tiles. So, gunners' range would still be 2 normal tiles, but now it'd be expressed as "range = 4" with combat tiles in mid; and their movement would also be 4 (or may be turned down to 3 huh?). City attack range =4 (perhaps bumped up to 5); upgraded SABR range would become 8, and its movement would be 2, etc.

And if you still don't get it and the question "what's the difference then?" lures in your mind now - then just imagine this: your SABRs will become able to in fact _move_ and then _fire_ in a single turn and without no road underneath; your Supremacy 2 attacks-per-turn gunners can now move a whole normal tile (two combat tiles) in addition to those attacks; you can put a line of a _dozen_ missile rovers all hitting a same target, if you'd want; and of course, your enemies can do all that - and more, - as well!

And as for terrain, it would still remain defined by large "civilian" tiles we have in the game right now, while those smaller "combat" tiles - would have movement defined by either the single "civilian" tile they are in the middle of, or by two "civilian" tiles they are between of; in the latter case, the harsher of two terrain kinds (if they differ) would be taken into account, so that if a unit moves through a combat tile which is half forest, and half tundra - 2x movement penalty would be applied. Coasts, obviously, would have combat tiles which are 1/2 land and 1/2 water - and those should be made passable by land units.

This solution i have in mind for a long time; it is simple, it removes most of "1 UPT" trouble, virtually allowing to _quadriple_ number of units every player can have within a single "normal" tile, gives more tactical flexibility and variability, allows for more complex tactical formations, and more.


But, again, i guess we won't see this ever implemented. Sadly...

Wow, that is an amazing Idea, and the first one I have ever seen that seems like it would solve the 1 UPT issue once in for all. Thanks for showing that 1 UPT can be implemented properly.

 

[nimling]

I think it would be better to abstract combat than to implement sub-tiles - Civilization and BE aren't too well designed to be detailed war games, and adding too much unit micromanagement to combat would detract from the game more than add to it. There ought to be a balance between simplicity and strategy.

RTS games are RTS games - there aren't really tiles, there are coordinates which a unit occupies, and a lot more of coordinates than tiles. In Starcraft or Rise of Nations, ranged combat isn't brokenly overpowered.

Board games that use 1UPT are also a different beast, and often do so for technical limitations - it's a lot more cumbersome when players need to manually manage the math of large armies, and only so many game pieces.

In short, the argument against Stacks of Doom that they are "too easy" isn't really an argument - SoDs simplify unit movement and logistics, both for the AI and for human players, and they make sense.
The implementation in Civ4 and Civ3 leaves much to be desired, for various reasons I have mentioned earlier in this thread. The answer isn't 1UPT which has been a disaster, but smarter implementation of unit stacking.

 

[Galgus]

I think it would be better to abstract combat than to implement sub-tiles - Civilization and BE aren't too well designed to be detailed war games, and adding too much unit micromanagement to combat would detract from the game more than add to it. There ought to be a balance between simplicity and strategy.

RTS games are RTS games - there aren't really tiles, there are coordinates which a unit occupies, and a lot more of coordinates than tiles. In Starcraft or Rise of Nations, ranged combat isn't brokenly overpowered.

Board games that use 1UPT are also a different beast, and often do so for technical limitations - it's a lot more cumbersome when players need to manually manage the math of large armies, and only so many game pieces.

In short, the argument against Stacks of Doom that they are "too easy" isn't really an argument - SoDs simplify unit movement and logistics, both for the AI and for human players, and they make sense.
The implementation in Civ4 and Civ3 leaves much to be desired, for various reasons I have mentioned earlier in this thread. The answer isn't 1UPT which has been a disaster, but smarter implementation of unit stacking.

Unit stacking still cheapens the power of bottleneck positions and positioning soldiers on the field.

The reason stacks doom are inferior isn't that they are too easy, its that they water down strategy.