Improving 1UPT

[Greizer85]

Damage to every stacked unit is certainly a large enough penalty to discourage the use of stacks in combat, provided that you program the AI to never use stacks (as it would just get its units killed in mind-boggling maneuvers). With unlimited stacks, you could still use stacks for 'storing' units close to the front though: simply switch out a badly beaten frontline defender for a fresh stack unit sitting a tile or two back from the front in a safely guarded stack. It would take out some of the tactial element in the form of much better 'battlefield mobility', as it were. Now if you'd also lose every movement point on entering and exiting a stack then that could make this tactic less effective. Whether it would totally prevent it I'm not sure. In chokepoints, where movement doesn't matter, you could presumably still use it to good effect.

Whatever approach is taken, basically stacks should be made such that their use in combat is either extremely situational or not worth it at all... Otherwise you end up making stack combat optimal and there goes the whole nice tactical puzzle of 1upt combat.

 

[kaspergm]

Damage to every stacked unit is certainly a large enough penalty to discourage the use of stacks in combat, provided that you program the AI to never use stacks (as it would just get its units killed in mind-boggling maneuvers). (...)

But then, given that at least half the incentive of (re)introducing stacks is to help the AI better cope with movement, doesn't the idea collapse on itself if we disallow AI to use stacks? And otherwise, if we code the AI to "only use stacks for transport", it will still fail, because it can't understand and predict the range of air units, ships, etc. Having five Battleships lurk out of view and then zip in and kill entire army in one turn will make the game pretty flat imo.

I'm not against re-thinking how units move, but I have a very hard time coming up with a goold solution without drawbacks. I think one could make some sort of fluid modelling where army is moved as an entity as long as not within enemy ZOC. Imagine AI has an army consisting of 4 Pikemen, 4 Composite Bows and 2 Horsemen - in total 10 units. When this army is moved, this should simply be calculated as a 10-tile "liquid" mass with a collective movement of 2 tiles per turn (i.e. we ignore extra movement of Horses to keep the army collected at all times) which needs to flow through the landscape best possible - but without necessarily calculating movement specific movement of each individual unit within this mass, but rather move the mass as an entity and then afterwards provide positions of each unit according to their role (we might put Pikes as front line, CBs in centre and Horses to guard the back). This would mean that army will slow down when it meets choke points like mountain gaps, but depending on the amount of "cheating" one will allow in not calculating individual units each turn but rather just flow as a collective mass, the AI might be able to better cope with this. The problem obviously would be that if one completely removes the individual unit identity during the movement and only assign it after the movement is done, we might end with some units potentially moving longer than their movement would allow - but if this method is not used during specific combat but only when transporting, and if overall army never progresses more than the minimum movement allowed, this might not be a real problem. Whether it would actually help the AI, I don't know.

 

[phoinix]

If you do it the civ 1 (or was it civ 2 or 3) way of letting the strongest unit defend a stack, the entire tactic side of 1upt is gone. Units don't die so easily in Civ 5 nowadays, it takes more than one attack to kill something and ooops, when my next attack comes, an other unit is defending the stack because that unit will be the strongest.

Also they will not do something drastic anyway because they won't re-program the AI and the AI is programmed to handle 1upt, not stacks, regardless of how you use the stack. They hopefully will finetune the 1upt AI so it works better, perhaps re-adjust some combat stats here and there and perhaps do something regarding stacking and civilians. Stacking of civilians is much easier to add since the AI doesn't have to be re-programmed to handle some new tactic situations, they just need to add a 'no no' for giving multiple work orders.

I'm treating the stack as 1 unit. That means that there's no healing for the rest of the units while 1 counterattacks. Also, as each attack on the stack affects all units it's natural to say that the toughest units just survive longer. The back damage that the attacker receives may be the average of the back damage from each unit in the stack.

I'm sure they won't do anything that drastic as to abandon 1upt in an expansion.

The more you suggest it the less sense it makes to me and the more silly it sounds compared to the current system invalidating territory, flanking tactics, false retreats, movement, partial retreats, stalls, etc. I'll agree to disagree, but I can honestly say I hope your system isn't ever considered, it would be 4 steps back and a fourth of a step forwards

Well it's fine by me that you don't agree but you're not giving any particular reason for it, so allow me to ignore you.

One, for me, strong argument about the stacking-with-collateral-damage idea is the AI. Imagine how "fun" it will be when the AI marches in all its units and fails to understand that your city or bombers or whatever will eleminate them in one strike. *shutters* Not that the AI currently offers much of a tactical opposition, but this could potentially make it much worse.

It will sure be difficult to program the AI not to do such stupid mistakes but there's no reason to believe that a mature patched AI with this system in place will behave worse than now.

But then, given that at least half the incentive of (re)introducing stacks is to help the AI better cope with movement, doesn't the idea collapse on itself if we disallow AI to use stacks? And otherwise, if we code the AI to "only use stacks for transport", it will still fail, because it can't understand and predict the range of air units, ships, etc. Having five Battleships lurk out of view and then zip in and kill entire army in one turn will make the game pretty flat imo.

I'm not against re-thinking how units move, but I have a very hard time coming up with a goold solution without drawbacks. I think one could make some sort of fluid modelling where army is moved as an entity as long as not within enemy ZOC. Imagine AI has an army consisting of 4 Pikemen, 4 Composite Bows and 2 Horsemen - in total 10 units. When this army is moved, this should simply be calculated as a 10-tile "liquid" mass with a collective movement of 2 tiles per turn (i.e. we ignore extra movement of Horses to keep the army collected at all times) which needs to flow through the landscape best possible - but without necessarily calculating movement specific movement of each individual unit within this mass, but rather move the mass as an entity and then afterwards provide positions of each unit according to their role (we might put Pikes as front line, CBs in centre and Horses to guard the back). This would mean that army will slow down when it meets choke points like mountain gaps, but depending on the amount of "cheating" one will allow in not calculating individual units each turn but rather just flow as a collective mass, the AI might be able to better cope with this. The problem obviously would be that if one completely removes the individual unit identity during the movement and only assign it after the movement is done, we might end with some units potentially moving longer than their movement would allow - but if this method is not used during specific combat but only when transporting, and if overall army never progresses more than the minimum movement allowed, this might not be a real problem. Whether it would actually help the AI, I don't know.

Such problems are inevitable to crop up and exist even now. For example why doesn't the AI cover its units with mobile SAM when it's proven that I have bombers? Same thing with a stack. Ideally there should be some code somewhere saying if unit/stack attackable or possibly attackable then split/move to safety. Even better if there is some risk number associated with each stack that triggers defensive measures easier or harder ( a big stack is more precious and thus should split at the first sight of danger). As a matter of fact even a human player can loose a stack to an unseen ranged unit, so I can live with an AI occasionally loosing a stack to unforeseen dangers. The problem would be to loose a stack to known dangers. Also in the sea, where there are more dangers, there is a lot more free space available so the stacking mechanic looses traction there.

As for movement, remember that the stacking mechanic is only meant to make life easier when mobility is hard! When land is abundant, there is no reason to hold the stack together at all times. Units can spread out or not. That's the point. There's no restriction holding the units apart all the time! This way a large force becomes more like a fluid like you said.

 

[gumbosis]

One thing I wish they would change is units freaking out that someone is occupying the tile you sent them too, even though they may be 3 or 4 turns away

This is the one thing that annoys me too.

 

[Sonereal]

One thing I wish they would change is units freaking out that someone is occupying the tile you sent them too, even though they may be 3 or 4 turns away

The crazy part is that when A New Dawn experimented with XUPT, as in you can set the number of units allowed on a tile (CivIV mod), the units handled this EVEN BETTER. They would keep moving to the tile even if the tile was full and only stop moving once the next move they make would be on the full tile.

 

[9gkj548hkm20vk79dj689bnj4]

If you don't want one unit-per-tile, then you probably don't want a hexagonal grid board, either.

Just think, how can two or more units theoretically occupy the same game space in a turn-based game? Does ANY other game do that successfully? Not even the king in checkers. Because what you're doing with stacks is not having several unique units in a space, but essentially one unit with multiple abilities instead, just like the checkers king can move backwards and must be jumped twice in order to liberate that game space. Because it is pretty much impossible for stacks to act (i.e. attack/defend) independently while occupying the same space. You can't do it. Anything resembling a compromise is going to invite a new series of rules to justify what is basically a single stronger game piece in that one space, nothing more. Okay, two units? Why not three? A hundred? How do they move? Who gets damaged? In what sequence? Do their attacks stack, too? Etc...

The only alternative would be, as suggested, a different plane for each unit type with no interaction whatsoever (military, great person, religious, settler/worker) but keep the one unit-per-tile rule. And waypoints, too.

 

[phoinix]

If you don't want one unit-per-tile, then you probably don't want a hexagonal grid board, either.

Just think, how can two or more units theoretically occupy the same game space in a turn-based game? Does ANY other game do that successfully? Not even the king in checkers. Because what you're doing with stacks is not having several unique units in a space, but essentially one unit with multiple abilities instead, just like the checkers king can move backwards and must be jumped twice in order to liberate that game space. Because it is pretty much impossible for stacks to act (i.e. attack/defend) independently while occupying the same space. You can't do it. Anything resembling a compromise is going to invite a new series of rules to justify what is basically a single stronger game piece in that one space, nothing more. Okay, two units? Why not three? A hundred? How do they move? Who gets damaged? In what sequence? Do their attacks stack, too? Etc...

The only alternative would be, as suggested, a different plane for each unit type with no interaction whatsoever (military, great person, religious, settler/worker) but keep the one unit-per-tile rule. And waypoints, too.

Not sure if you're responding to me but,

Guys guys... I fixed it.
I'm a pro freedom and flexibility guy so my solution is this: ALLOW stacking. Any stacking like in Civ 4! Now before you go on and flame me let me explain a little bit more. Allow stacking, but attach penalties to it. Here's what I have in mind:
1) When a stack is attacked by a ranged any unit, all units in stack receive damage. Collateral. It's as if each of the units in the stack are attacked individually by the attacker, who (if melee) receives the average back damage from all units in the stack.
2) Only 1 unit can attack from a stack!
3) Keep the existing bonuses of having your units spread out: Flanking bonus, discipline bonus.

Now what does this solve?
The No 1 most important underlying problem of Civ5: Moving units around.

How is this not Civ4?
Now it actually makes sense to keep your forces spread out, to avoid collateral, spread your risk, get those combat bonuses and actually allow more units to attack simultaneously. It isn't strictly enforced! It gives you options:
-Do I just want to move my units around? Stack them! (and avoid the pain of meaningless coordinating movements between them)
-Do I want to fight? Spread them out!
A stack isn't a stack-o-doom now and you get all that tactical gameplay of Civ5! Plus, it'll be much easier for the AI to handle. It's WIN-WIN.

Your thoughts.

I think it's pretty clear, answering all your questions.

One more time: There is no designated stack defender as in Civ4, where the stronger unit defended the stack). All units receive the damage, as in, each unit in the stack was attacked individually by the attacker. There is also only 1 unit that can attack from the stack each turn. The stack acts as 1 unit. So it's basically not a single stronger game piece.

PS1: I love hex grid and it has nothing to do with 1upt.

 

[Apkungen]

I am not a fan of the 1 UPT. They way units block each other but some things like cities still have an out of scale way to damage units who have to crawl up to them is crap. I miss the real world operational actions of units defending and covering each other.

My biggest issue is naval convoys and amphibious operations where you have to spread out over a huge area and have AT LEAST twice the number of escorts and when you hit the beach with your invaders you can't just take a peninsula and build up from there, no you have to have an invasion area of 10 or 20 tiles and your first wave to hit the beach are destined to be wiped out. That is wonky even for a sandbox empire game.

Add the scale issues (archers fire longer than men armed with 20th century bolt action rifles), cities defence value mainly based on size and great lardasses blocking workers and eachother and you have a game I rarely bother to play beyond the ancient era.

A return to a stacking system with damage to all stacked units (percentile), better AI, building of bigger formations as eras comes (think the armies of Civ 3) and maybe a movement point and occupation system for tiles

 

[iconocast]

1upt is a massive improvement over tile stacking. as long as they dont go back to the insanity of stacks of 200 units in one tile, and instead add more unit/positional interactions eg. flanking.

 

[DemonMaster]

I am not a fan of the 1 UPT. They way units block each other but some things like cities still have an out of scale way to damage units who have to crawl up to them is crap. I miss the real world operational actions of units defending and covering each other.

My biggest issue is naval convoys and amphibious operations where you have to spread out over a huge area and have AT LEAST twice the number of escorts and when you hit the beach with your invaders you can't just take a peninsula and build up from there, no you have to have an invasion area of 10 or 20 tiles and your first wave to hit the beach are destined to be wiped out. That is wonky even for a sandbox empire game.

Add the scale issues (archers fire longer than men armed with 20th century bolt action rifles), cities defence value mainly based on size and great lardasses blocking workers and eachother and you have a game I rarely bother to play beyond the ancient era.

A return to a stacking system with damage to all stacked units (percentile), better AI, building of bigger formations as eras comes (think the armies of Civ 3) and maybe a movement point and occupation system for tiles

I have never had "twice the number of escorts" or had to invade "10 or 20 tiles" when I make amphibious assaults. Such exaggerations only looks ridiculous.

Civ is an abstract simulation. The combat is this way in an effort to simulate a battlefield for each era and sure it can be a little strange if units of different eras meet. At least not more strange than it take 40 years to travel 20 miles (or whatever length two tiles is) but in the year of 2975 BC it only takes 25 years or only 10 years by the time you hit 1450 AD.

 

[Ninakoru]

Guys guys... I fixed it.

...

Your thoughts.

The bad part.. the stupid AI. I foresee a half dozen AI units caught with ranged fire and thus killing 6 for the price of one. AI has the habit to mass and pack massive unit quantities. Not to mention charge units hit and run tactics would be way more powered.

I would enable unit stacking up to two combat units, with the backdraw of not being able to attack, but being able to defend with a percentage decrease to defense, like the -20% from attacking over rivers. You should be able to select witch of the two units to attack. the worst part of 1upt is the extra effort moving around groups of units using long road/railroad paths, this should mitigate that problem, and stupid AI wouldn't be severely punished with their stupid decisions.

 

[Eru Ilúvatar]

There's a serious bug. If you pass throw a terrain that already has units and your movement ends, you can't go on in the game. Terrible when you are exploring new terrain or entering a very militarized area.

 

[AriochIV]

There's a serious bug. If you pass throw a terrain that already has units and your movement ends, you can't go on in the game. Terrible when you are exploring new terrain or entering a very militarized area.

I believe you have to disband the unit in this case.

 

[Eru Ilúvatar]

Or return to a saved game. Indeed.

 

[TheHanzou]

Id like to see a return of the Army feature from Civ3, where we where able to combine units.

 

[Bevel Lemelisk]

The only possible change I could maybe accept is for Defense only Ranged units (read siege) units occupying another non ranged units tile. Though stacking of more than 1 civilian unit would also be acceptable.

 

[Jabberwockxeno]

The maps need to be bigger to allow for the lack of stacks. Simple as that. This would also mean scaling up the amount of movements units can make, and IMO, making cities be larger than 1 tile.

 

[Jabberwockxeno]

Guys guys... I fixed it.
I'm a pro freedom and flexibility guy so my solution is this: ALLOW stacking. Any stacking like in Civ 4! Now before you go on and flame me let me explain a little bit more. Allow stacking, but attach penalties to it. Here's what I have in mind:
1) When a stack is attacked by a ranged unit, all units in stack receive damage. Collateral.
2) Only 1 unit can attack from a stack!
3) Keep the existing bonuses of having your units spread out: Flanking bonus, discipline bonus.

Now what does this solve?
The No 1 most important underlying problem of Civ5: Moving units around.

How is this not Civ4?
Now it actually makes sense to keep your forces spread out, to avoid collateral, spread your risk, get those combat bonuses and actually allow more units to attack simultaneously. It isn't strictly enforced! It gives you options:
-Do I just want to move my units around? Stack them! (and avoid the pain of meaningless coordinating movements between them)
-Do I want to fight? Spread them out!
A stack isn't a stack-o-doom now and you get all that tactical gameplay of Civ5! Plus, it'll be much easier for the AI to handle. It's WIN-WIN.

Your thoughts.

I REALLY like this idea. Would love to see somebody mod it hinthintcoughcough.

 

[les berceaux]

Have you guys noticed how the AI often gets to cheat the 1UPT rule during wars

 

[DemonMaster]

Have you guys noticed how the AI often gets to cheat the 1UPT rule during wars

No ...............