Thoughts on 1UPT and possible improvements

Schetter

Chieftain
Joined
Nov 21, 2005
Messages
39
I'm working from the following premises:
  • - massive unlimited stacks have their own inherent gameplay problems. These were not really solved over multiple iterations of the game, and while they make sense in a larger context of a strictly strategic nation building game, it essentially reduces the military game to Alpha City with X power attacks Beta City with Y power. I don't wish to return there.
  • - strict 1UPT works better to a degree by forcing players to explode what would be a conventional stack of doom into larger formations; but it introduces its own set of inherent problems, and its implemention in Civ V, when considered with other corresponding gameplay rules, is half baked at best.
  • - A move towards xUPT, where x is some number, would devolve into a discussion over what arbitrary number is better, and so it's pretty useless to go there.
  • - The AI kinda sucks militarily and there's nothing to be done about it at this point. I'm approaching this mechanic from the standpoint that the AI should understand its proper use, as I'm assuming the proper mod tools will be released...eventually, or will be fixed via patch. Whether that actually happens or not, I'm not concerned with.

If you disagree with any of that, it's cool, and feel free to voice it, but it's my hope that I can sort of constrain what discussion there might be under this post to improve the 1UPT dynamic present in Civ V by those points. I'm already messing with some of the thoughts included below in my mod, and I was thinking it might be useful to others if I packaged these changes off to be used separately from the larger mod in progress (that I just haven't had time to work on as of late) so they could actually be used. I wanted some feedback to see if there were some other ideas that could be usefully incorporated.


So, here's my thoughts so far, problems first:

  • - There's a lot of mental hangups that the game encourages simply by how units are referenced and their range abilites (or lack thereof). It's difficult to maintain the two disparate notions of distance that the game creates between the strategic layout of cities, resources, improvements, continents, etc. and the tactical interactions between military units. Specifically, how archers can apparently fire off volleys that reach several miles, but modern tanks don't cover the same range. There seems to have been an overarching thought towards having different classes of units with different battlefield roles and positions in formations - recon, front line, ranged, siege, and fast moving - that definately showed promise, but wasn't followed through on completely. This thought either needs to be scrapped and return to a rock-paper-scissors system of power modifiers (problematic with the removal of separate attack/defense powers), or it needs to be refined and expanded so that it's consistently present across the entire length of a game, not just the early game (more realistic).
  • - Base movement of 2 basically nullifies any tactical interest that might have been created with 1UPT. I understand that it was an expedient change to make in order to try and alleviate some of the propensity for traffic jams that the AI tends to have, but I think there's better ways of trying to fix it that don't make the 'disparate distance' problem mentioned above even worse. Warriors moved one tile per turn for over a decade...why did it make sense to change that? Deficient AI doesn't validate it If the game's landscape becomes too choked with units each claiming whole tiles for themselves, the proper tweak would be in making it more difficult to produce units, not allowing all of them to move around faster so that they can play leapfrog all game. These decisions cheapened the feel of the military game quite a bit.
  • - Workers simply get in the way; particularly foreign workers of nations that you've entered into open borders with. I've yet to figure out a way to mod in the ability for two units of different types (civilian/military) and of different nations to be able to occupy the same tile; there's likely some checksum that we don't have exposure to yet that will need to be modified for that. But in the meantime, this is still a problem that creates traffic jams. I think that rethinking the way Workers are built and operate could have some beneficial effects on gameplay. (Who the hell enjoys managing workers anyway?)


Some definate ups that 1UPT brings to the table:

  • - Combined with the ability to buy tiles, forts actually become useful to build! Fighting defensive battles, in which you fortify a defensive line with the intent of actually making your stand there, is no longer a fool's errand, which I find particularly interesting.
  • - Larger armies must fan out into proper formations. Armies take up space; this is absolutely appropriate.
  • - It creates the potential for interesting symbiotic placement relationships between units of different classes, such as front-line w/ ranged, or ranged w/ recon.


And, my proposed changes, in general:

  • - Scale back the base movement of all units by one, keeping it elevated for those that make sense. Human units with feet on the ground should be one tile per turn. Mounted units and possibly recon should be 2. The most primitive ships should start at 2. And so on... Couple it with a reduction in visibility to a base of 1 tile as well, leaving the larger sight range to recon units. This would return a sense of a massive world that you're exploring and traversing, instead of some toy construct of a planet, and creates relationships between ranged units and front line units, where they rely on each other for sight lines and protection, respectively.
  • - Modify the worker mechanic so that (1) they don't cause traffic jams by just sitting around idle and (2) city-states can't be subject to stolen worker abuse quite so easily in the early game. The change goes something like this:
    • -- Workers should cost a fraction of default production (so by 50 turns into the game, you should be able to crank them out in one turn from a mid-sized city, or a pretty nominal gold fee.
    • -- All production times for improvements stay the same
    • -- On completion of an improvement, the worker is destroyed, the effect being that the worker is now 'working' the improved tile. (Note for self: Mission.EntityEventType is null for the 'build improvement' mission, possibly change this to 'ENTITY_EVENT_DIE' to achieve this?)
  • Thus, each worker should be looked at as a single improvement that hasn't been built yet, instead of a permanent construction crew. This will essentially turn the world map into more of a military domain, instead of a funky shared civilian/military hybrid, that's made all the more funkier with 1UPT and one-nation-per-tile.
  • - Reduce the combat power of recon units to 0. This makes the game engine treat them as civilian units (triangular unit icon instead of circular), allowing them to traverse freely through tiles that have combat units occupying them, but also makes them defenseless (which, in a way, makes sense). I have not tested yet how the engine reacts to a unit that has no base combat value but does have a ranged attack; if such units are also treated as civilian, that means that units such as archers, for instance, could be modified so that they could occupy the same tile as a warrior, but would not require lifting the 1UPT rule, and would still only allow one archer in that tile.


What it would achieve, I think:

  • - Richer tactical battles, and more defensible borders.
  • - Traffic jam alleviation
  • - 0 need for worker automation
  • - More potential for interesting relationships between units, particularly in the later game, where each class has particular strengths that fill in the gaps for others (recon->sightlines, archers->range, siege->extreme range, infantry->real estate control, etc..), allowing a more carefully constructed diverse force do better against a one-trick-army.
  • - A step towards bringing back the 'epic' feel that previous titles achieved.


Your thoughts are appreciated. If you have no interest in any of this, that's ok, please don't flame and just walk away from the thread. I'm prepared for the though that not everyone will be set on fire by this proposed modification...thanks.
 
I don't like the idea about workers only being used once. It makes the game a little bit more boring IMHO as you have to keep on building new workers, and it also ignores the fact that citizens work tiles. I think that it would lose the game to feeling of you building something and expanding. It also isn't very fun for builders. If you like warring, it would be equivalent to making it so that each military unit can only be used once.

I also don't like the idea of the scout (recon) not having any strength. I think it should go back to civ 4 where is cannot attack but can defend, so it should be classed as a civilian and so can avoid 1UPT issues.

When it is possible to mod, I would find that allowing a 1UPTPN (per nation) would help a lot.

Apart from that, I like it.
 
Scale back the base movement of all units by one, keeping it elevated for those that make sense. Human units with feet on the ground should be one tile per turn. Mounted units and possibly recon should be 2. The most primitive ships should start at 2. And so on... Couple it with a reduction in visibility to a base of 1 tile as well, leaving the larger sight range to recon units. This would return a sense of a massive world that you're exploring and traversing, instead of some toy construct of a planet, and creates relationships between ranged units and front line units, where they rely on each other for sight lines and protection, respectively.

Doesn't this just make hills and forests redundant? After all, it doesn't make sense that an infantry unit moving through a hill or forest moves at the same rate as marching across plains. I bring this up because if the point is to make 1UPT try to fit with scale, then that should be accounted for.

I like visibility reduction though.

Modify the worker mechanic so that (1) they don't cause traffic jams by just sitting around idle and (2) city-states can't be subject to stolen worker abuse quite so easily in the early game. The change goes something like this:

Workers should cost a fraction of default production (so by 50 turns into the game, you should be able to crank them out in one turn from a mid-sized city, or a pretty nominal gold fee.

Sounds alright I suppose.
All production times for improvements stay the same

Good. I like the improvement build times on marathon.

On completion of an improvement, the worker is destroyed, the effect being that the worker is now 'working' the improved tile. (Note for self: Mission.EntityEventType is null for the 'build improvement' mission, possibly change this to 'ENTITY_EVENT_DIE' to achieve this?)

I toyed around with this idea when there was a problem with the worker AI slowing the mid-game to the crawl. The only problem is that there's no production overflow (if I'm correct) and 1 turn producing a worker is still 1 turn. I think a better solution to traffic jams would simply allow civilian units stack but not improve the same tile.

Thus, each worker should be looked at as a single improvement that hasn't been built yet, instead of a permanent construction crew. This will essentially turn the world map into more of a military domain, instead of a funky shared civilian/military hybrid, that's made all the more funkier with 1UPT and one-nation-per-tile.

I still think unlimited civilian stacking is the way to go.

Reduce the combat power of recon units to 0. This makes the game engine treat them as civilian units (triangular unit icon instead of circular), allowing them to traverse freely through tiles that have combat units occupying them, but also makes them defenseless (which, in a way, makes sense). I have not tested yet how the engine reacts to a unit that has no base combat value but does have a ranged attack; if such units are also treated as civilian, that means that units such as archers, for instance, could be modified so that they could occupy the same tile as a warrior, but would not require lifting the 1UPT rule, and would still only allow one archer in that tile.

Sounds interesting.

May want to apply this stuff to AA guns and SAMs, making them anti-aircraft their main function (not clean up like they are now).
 
Nice approach I love your methodology.

I feel like Civ 5 was moving in the right direction, but was just really poorly thought through/implemented.

Personally I think you ditch workers all together and just let cities buy the tile improvements for gold or production. Makes a lot more sense, saves everyone some headaches, and is less needless micro management. It also keeps the map clean. Then you could even drop the civilian military distinction and just have the settler be able to stack.

The movement rate stuff is very tricky, not sure if 2 or 1 is a better base. Personally I would ditch ranged units all together, the AI has no idea how to use them properly.

If I was making my own mod (which I may yet some day), i would go to a completely different (an more appropriate) level of abstraction with the units.

So like in the first couple tiers you might have:

Stone Age Army STR2
Bronze Age Army STR3
Broze Age Offensive Army STR3+50%vs cities
Iron Age Army STR4
Iron Age Offensive Army STR4+50% vs cities
Classical Age Army STR6
Classical Age Offensive Army STR6+50%vs cities
Classical Age Mounted Army STR5+Move 3
Dark Age(Early Medieval) Army STR8

et cetera
et cetera

The AI would have a hard time screwing that up.

Offensive armies cost more, and are good at breaking cities. Makes warmonger work for it, which is good since war is so strong in the game. And mounted armies are weaker than unmounted ones, but get that extra move.

I think this was the big problem with Civ 5. Too much going on and too many elements when the basic underlying structure doesn't even work. Why have 15 different types of units when the AI can only functionally use 1 or 2 of them?

I know people love flavor, but I will take functionality over flavor any day, and once you get the game functional it is a lot easier to add the flavor back in.
 
About the workers:
"Killing" workers on completion of the improvement doesn't make sense, as this leads to the necessity of constantly re-producing them. This means that the player has to do the same action over and over again, which especially on bigger maps (more cities, larger distances) will become a nuisance.

Regarding unit movement:
The main problem of Shafer_5 is that units don't behave in a logical way.
You can easily move a unit, enter the line of sight of another unit and still have the original unit moving on, that way exposing it to the other unit's "range of action".
This is especially problematic when moving units by a "go to" command.

What is needed is to make units stop and ask for new orders in case of hostile units being spotted.

Furthermore, as long as "1upt" is still in place, the basic 2 movement points do more damage than anything else.
Your idea of reducing this basic movement to 1 seems to be appealing at first glance, yet it makes plains (where you will face a defense penalty) a "no-go" area.

Actually, I think the movement rates in Shafer_5 are too LOW, not too high.
I agree though that this conflicts with the "density" of the terrain.

Under the premise of staying with "1upt" the current movement rate seems to be better than 1, although weaker than 3 or 4.
The later two ones would require a different set-up of maps, though; leading to all kind of other issues like distance between cities, cultural borders, resource placement and whatnotever.
So, finally, if you want to stay with "1upt" and the current map setup, stay with the movement rates, too.
 
Thanks for the earnest feedback thus far.

I don't like the idea about workers only being used once. It makes the game a little bit more boring IMHO as you have to keep on building new workers, and it also ignores the fact that citizens work tiles. I think that it would lose the game to feeling of you building something and expanding. It also isn't very fun for builders. If you like warring, it would be equivalent to making it so that each military unit can only be used once.

I also don't like the idea of the scout (recon) not having any strength. I think it should go back to civ 4 where is cannot attack but can defend, so it should be classed as a civilian and so can avoid 1UPT issues.

When it is possible to mod, I would find that allowing a 1UPTPN (per nation) would help a lot.

Apart from that, I like it.

I think the point about having to build workers a lot is also made by another poster further down the thread. I'm not sure if it will actually be that burdensome, but you might be right. I think the only way to find the answer is to actually playtest it (soon). My thought about the Missions table didn't work...but (!) in the Builds table, there's a boolean column named 'Kill'...which appropriately determines whether the unit during the building 'disappears' after it's finished building. I set this to 'true' for everything except for roads, railroads, forts, and removing stuff like forests and jungles. I also put a hard limit per player on workers at 5; this will likely require some tweaking, and maybe some additional worker-equivalent units for later eras that have progressively larger caps. I did a quick test last night just to confirm that this proof of concept is working, with the worker killing, and will dig in on the details of the mechanic further after unwrapping some gifts.

Regarding the recon and offense/defense...that's currently impossible given the way the vanilla game is written, and the lack of a proper SDK. Besides, the point was really to allow recon units to be able to move around amongst combat units and provide lines of sight from the 'front line' for any ranged units that might be behind the front line. Yes, it makes them a bit vulnerable to barbarian units in the early game if you're using them for long range exploration, but you'll simply have to be smarter in maneuvering them. Another weird consequence of this change, I found, is that you no longer have the option to have them explore automatically, which is kinda weird to say the least;at least this was the case before the mega patch, I'll have to check up on this again.

Doesn't this just make hills and forests redundant? After all, it doesn't make sense that an infantry unit moving through a hill or forest moves at the same rate as marching across plains. I bring this up because if the point is to make 1UPT try to fit with scale, then that should be accounted for.

I like visibility reduction though.



Sounds alright I suppose.


Good. I like the improvement build times on marathon.



I toyed around with this idea when there was a problem with the worker AI slowing the mid-game to the crawl. The only problem is that there's no production overflow (if I'm correct) and 1 turn producing a worker is still 1 turn. I think a better solution to traffic jams would simply allow civilian units stack but not improve the same tile.


I still think unlimited civilian stacking is the way to go.



Sounds interesting.

May want to apply this stuff to AA guns and SAMs, making them anti-aircraft their main function (not clean up like they are now).

I think that hills/forests still have a meaningful impact on combat in the defensive bonuses they provide to units that occupy them, so I wouldn't call them redundant; I always figured the defensive bonuses were far more meaningful than movement penalties anyway. I guess the thought is that it's far easier to bring the military game further towards a proper hex-and-turn-based strategy feel than it would be to try and make a proper tactical military simulation out of Civ. More on this later, though.

While I agree with you in principle on the 'unlimited civilian stacking,' I'm not sure if that's possible with the current modding tools. There's a variable in the GlobalDefines that's 'PLOT_UNIT_LIMIT', set to 1 in vanilla, and seems to apply to each type of unit separately; meaning that if it were set to 2, the game would allow two combat units and two civilian units on each tile. Which...come to think of it...might not be a terrible idea...contrary to my initial post. Anyway...

I will certainly look into applying the zero combat, non-zero range idea to AA guns as you suggested, and will probably do the same with siege. That would allow you to manage 'combat' units and 'support' units separately, since they'd be able to intermingle on the same squares. I'll try taking this thought to it's absolute maximum with ground units, giving anything with ranged attacks 0 direct combat power, and see how that plays, at least up until modern era units like tanks.

Nice approach I love your methodology.

I feel like Civ 5 was moving in the right direction, but was just really poorly thought through/implemented.

Personally I think you ditch workers all together and just let cities buy the tile improvements for gold or production. Makes a lot more sense, saves everyone some headaches, and is less needless micro management. It also keeps the map clean. Then you could even drop the civilian military distinction and just have the settler be able to stack.

The movement rate stuff is very tricky, not sure if 2 or 1 is a better base. Personally I would ditch ranged units all together, the AI has no idea how to use them properly.

If I was making my own mod (which I may yet some day), i would go to a completely different (an more appropriate) level of abstraction with the units.

So like in the first couple tiers you might have:

Stone Age Army STR2
Bronze Age Army STR3
Broze Age Offensive Army STR3+50%vs cities
Iron Age Army STR4
Iron Age Offensive Army STR4+50% vs cities
Classical Age Army STR6
Classical Age Offensive Army STR6+50%vs cities
Classical Age Mounted Army STR5+Move 3
Dark Age(Early Medieval) Army STR8

et cetera
et cetera

The AI would have a hard time screwing that up.

Offensive armies cost more, and are good at breaking cities. Makes warmonger work for it, which is good since war is so strong in the game. And mounted armies are weaker than unmounted ones, but get that extra move.

I think this was the big problem with Civ 5. Too much going on and too many elements when the basic underlying structure doesn't even work. Why have 15 different types of units when the AI can only functionally use 1 or 2 of them?

I know people love flavor, but I will take functionality over flavor any day, and once you get the game functional it is a lot easier to add the flavor back in.

I agree with the thought of ditching workers altogether, but it's above and beyond my abilities to mod this in currently, although I'll stop short of saying that it's impossible given the current tools; it very well might be. I suppose the justification for killing off the workers, in my mind, is that permanent persistent workers are a blank check for production. It seems far more balanced to me that improvements to city squares should be included in the same production mechanism as other buildings that cities produce.

I kinda like the more generic approach to unit naming and function, even though I'm not sure I'm that hot on the idea of 'offensive' armies that have bonuses vs cities. I'd say that an offensive division would consist of a normal army with an accompanying siege brigade in the same tile - that's entirely possible if siege have 0 combat with range, and I think makes the most sense.

I also agree that unit diversity can become a major problem in a 1UPT system. Even with a base movement of 2, it can be far too difficult to juggle units properly to bring the unit you want to the front of a battle. I almost think that it would make more sense to bring a true rock-paper-scissors feel to the core combat units, where you have, for example in the Ancient era:
  • Army I - Warriors (bonus against Spearmen)
  • Army I - Spearmen (bonus against Horsemen)
  • Army I - Horsemen (bonus against Warriors)
All armies would either be the same production time and same power, or have different production times with different powers. Armies from the next era would be listed as 'Army II' and would be categorically more powerful than their predecessors. Toss in promotions for each that would allow you to upgrade the core bonuses as well as fill in the gaps (to a point) in the RPS model, and you have a combat system that has much more clarity and fits far better into the overall scale of the game. Unique units could still fit quite easily into this system as well.

About the workers:
"Killing" workers on completion of the improvement doesn't make sense, as this leads to the necessity of constantly re-producing them. This means that the player has to do the same action over and over again, which especially on bigger maps (more cities, larger distances) will become a nuisance.

Regarding unit movement:
The main problem of Shafer_5 is that units don't behave in a logical way.
You can easily move a unit, enter the line of sight of another unit and still have the original unit moving on, that way exposing it to the other unit's "range of action".
This is especially problematic when moving units by a "go to" command.

What is needed is to make units stop and ask for new orders in case of hostile units being spotted.

Furthermore, as long as "1upt" is still in place, the basic 2 movement points do more damage than anything else.
Your idea of reducing this basic movement to 1 seems to be appealing at first glance, yet it makes plains (where you will face a defense penalty) a "no-go" area.

Actually, I think the movement rates in Shafer_5 are too LOW, not too high.
I agree though that this conflicts with the "density" of the terrain.

Under the premise of staying with "1upt" the current movement rate seems to be better than 1, although weaker than 3 or 4.
The later two ones would require a different set-up of maps, though; leading to all kind of other issues like distance between cities, cultural borders, resource placement and whatnotever.
So, finally, if you want to stay with "1upt" and the current map setup, stay with the movement rates, too.

Not sure what the greater nuisance would be...waiting for the AI to figure out what its going to do with their collective 50 workers every turn, or having to spend a turn in a city to get a new worker later in the game, when improvement building is more rare...playtesting should solve it.

As far as the base 1 movement making plains too dangerous to enter, that penalty (which never made practical sense anyway) can easily be modded out. I think with other appropriate, minor tweaks to the combat system, base 1 movement could make the military game feel more like the hex-based military strategy games that it seems this was supposed to be modeled after.


Thanks again all, I'll post up an alpha build soon.
 
Ok, starting to really wonder about this game's construction.

So, if you give a unit zero combat, but a non-zero ranged combat, the unit's icon is rendered as a triangle, BUT it's still treated as a military unit when it comes to the plot unit limit checksums. Which means my hopes that a compromise could be made here where a single 'army division' unit and a single 'support' unit with range could occupy the same square are pretty much squashed.

It seems that the only way I might be able to achieve what I'm thinking here is to up the plot limit to 2...and just call it 2UPT. After numerous playtests and tweaks over the course of the last few days, I'm about ready to say that I've left the church of 1UPT. I wanted to stick to 1UPT mostly as an academic point, so it could be said that we left no stones unturned in trying to bring sanity to 1UPT...not sure it can be done.

The more I play with the 0 power Scout, however, the more I like it. Yea, you've gotta be a bit more careful in open territory, but they move through other's territory far easier.

But finally, I've convinced myself that the one-use workers are, indeed, great. It opens up an expanse of new decisions to make, particularly in the early game. It sets up the building of a new improvement as another choice to make against extra early defense or scouting. Hopefully I can convert some of you folks soon; it does need some extra tweaking at this point.
 
Personally, I'd restore stacking but limit the number of units in a stack.

That would be completely arbitrary, and limit the players options. 1upt is the least arbitrary they could've gone for with the existence of some sort of limit. But simply not allowing stacks instead of nerfing them through a series of exponential penalties seems to be a lazy design decision. It's the easy way out, without providing the best solution.
 
Camikaze said:
That would be completely arbitrary, and limit the players options.

Arbitrary? No more than a vast amount of other stuff. Tech costs, unit costs, worker build times, tile yields - these are all arbitrary numbers. They're adjusted to balance relative to each other, but that's all. Similarly if you use a limited number of units per stack, that number would have to be balanced against existing arbitrary values too (to be precise against the average number of units per tile on the map).

One unit per turn simply cannot work in its current form - the map is too small relative to the number of units. For meaningful tactical combat in this kind of game most tiles on the map must be unoccupied at any given moment - otherwise it just turns into a traffic jam. With the existing style of maps you'd either have to have very few units (like less than one of each type per civ) or vastly increase the number of tiles, which would require a complete reworking of city tile use and improvement. If you want to get rid of SoDs, then a stack limit which keeps the map uncluttered, but still requires armies split across multiple tiles is a viable option.

But simply not allowing stacks instead of nerfing them through a series of exponential penalties seems to be a lazy design decision.

If you're trying to remove a particular style of play, adding a new one and then simply nerfing all the alternatives is not good design. Particularly if you want to offer players some choices to make, then it is better to take a carrot and stick approach, with the emphasis on the carrot.

For example - the problem with SoD was that it was always the best option. This is an inherent flaw with a rock/paper/scissors style game of units which counter other units since you can essentially choose all the options at the same time. Civ 5 simply bans the player from having more than 1 unit - there's still no incentive to spread units, it just doesn't give you any alternatives. In a sense SoD still exists, except it's a stack of promotions on an absurdly promoted unit that the AI can't kill.

So, what are some possible options:

1)Keep 1upt, but alter the map and/or unit numbers to reduce congestion. This will require rebuilding cities and tiles from scratch or limiting the number of units massively more than Civ 5 is currently doing (it's not just high difficultly levels - I was getting carpet of doom problems on my very first game at prince level).

2)Limited stacks. This is the obvious quick fix - that's not to say it's the best option, but it would massively reduce traffic jams to even allow 2upt. You could have the actual number increase via tech tree, policies etc. Just call it logistics. Probably would work best as a variable as the number of units on the map will increase throughout the game.

3)Split the map into strategic and tactical levels. Say, break down each tile into a number of smaller tiles. If each tile were composed of say 7 (tweak as necessary) smaller tiles that would greatly increase the amount of room for units. Cities and workers work on the normal tiles, but each tile is also a grid of smaller tiles on which 1upt holds true. Tactical combat can then occur as intended in Civ 5 without congesting the map, or requiring massive alteration of production and research level mechanics. How units would attack cities/workers would need some thought. Would work better with squares than hexagons because you can't make a perfect large hexagon tile out of small ones.

4)Remove the limit for unit stacking completely, but break out the carrot for splitting your units up. This would obviously remove traffic jam elements completely. What carrots could be offered? How about giving a bonus to a unit's combat strength based on the number of friendly units on adjacent tiles and/or the number of friendly units on tiles adjacent to the enemy unit. That would encourage spreading units out across multiple tiles to surround an enemy. How about another possiblity - make the experience gain of a unit dependent on the number of units in the stack. A unit alone gets a much larger bonus and hence more promotions than one stacked with twenty other units.

On the stick side a fairly simple option would be to remove the Civ 4 mechanic where the "best" defender is always presented to counter the attacker. Make it a random unit that's presented and the SoD becomes very unfavourable. Weak units like artillery would then need to travel in a separate stack behind a good all round defending stack or risk being easily destroyed.
 
Arbitrary? No more than a vast amount of other stuff. Tech costs, unit costs, worker build times, tile yields - these are all arbitrary numbers. They're adjusted to balance relative to each other, but that's all. Similarly if you use a limited number of units per stack, that number would have to be balanced against existing arbitrary values too (to be precise against the average number of units per tile on the map).

This is correct, but the specific number of units allowed per tile is less interconnected with other parts of the game than those arbitrary factors that you did list. I would say that the number of units per tile is only tenuously linked to other factors in the game (I'm guessing that other factors were worked around 1upt, rather than 1upt being decided upon because it turned out that was what balanced with other factors), rather than being heavily related. That is, the number of units per tile is much more a number pulled out of thin air than tech costs, production costs, etc.

One unit per turn simply cannot work in its current form - the map is too small relative to the number of units. For meaningful tactical combat in this kind of game most tiles on the map must be unoccupied at any given moment - otherwise it just turns into a traffic jam. With the existing style of maps you'd either have to have very few units (like less than one of each type per civ) or vastly increase the number of tiles, which would require a complete reworking of city tile use and improvement. If you want to get rid of SoDs, then a stack limit which keeps the map uncluttered, but still requires armies split across multiple tiles is a viable option.

The key difference is between 'limiting stacks' and 'stack limits'. They both achieve the same thing in terms of getting rid of SoDs, but the former doesn't arbitrarily prevent the player from exercising certain gameplay options, and completely gets rid of traffic jam situations.

If you're trying to remove a particular style of play, adding a new one and then simply nerfing all the alternatives is not good design. Particularly if you want to offer players some choices to make, then it is better to take a carrot and stick approach, with the emphasis on the carrot.

For example - the problem with SoD was that it was always the best option. This is an inherent flaw with a rock/paper/scissors style game of units which counter other units since you can essentially choose all the options at the same time. Civ 5 simply bans the player from having more than 1 unit - there's still no incentive to spread units, it just doesn't give you any alternatives. In a sense SoD still exists, except it's a stack of promotions on an absurdly promoted unit that the AI can't kill.

This is precisely my issue with limits of any sort. You aren't allowing the player the option to do what they please. You are simply preventing them via the ruleset from doing something.

So, what are some possible options:

1)Keep 1upt, but alter the map and/or unit numbers to reduce congestion. This will require rebuilding cities and tiles from scratch or limiting the number of units massively more than Civ 5 is currently doing (it's not just high difficultly levels - I was getting carpet of doom problems on my very first game at prince level).

This seems the most likely solution; it wouldn't require a complete removal of 1upt. But it isn't the best one. You'd still be limiting the player. The only problem you'd eliminate is congestion.

2)Limited stacks. This is the obvious quick fix - that's not to say it's the best option, but it would massively reduce traffic jams to even allow 2upt. You could have the actual number increase via tech tree, policies etc. Just call it logistics. Probably would work best as a variable as the number of units on the map will increase throughout the game.

Well, I've said why I wouldn't like this solution, and I find it difficult to think it would be better than 1upt. What you'd gain in terms of lowered congestion you'd lose in terms of warfare gameplay. As someone pointed out in another thread recently (can't remember who or where exactly), allowing 2upt would remove any need to protect your weaker ranged units. The game would essentially be groups of 2 moving around together instead of 1's, but with less tactical variation.

3)Split the map into strategic and tactical levels. Say, break down each tile into a number of smaller tiles. If each tile were composed of say 7 (tweak as necessary) smaller tiles that would greatly increase the amount of room for units. Cities and workers work on the normal tiles, but each tile is also a grid of smaller tiles on which 1upt holds true. Tactical combat can then occur as intended in Civ 5 without congesting the map, or requiring massive alteration of production and research level mechanics. How units would attack cities/workers would need some thought. Would work better with squares than hexagons because you can't make a perfect large hexagon tile out of small ones.

Dear God please no.

4)Remove the limit for unit stacking completely, but break out the carrot for splitting your units up. This would obviously remove traffic jam elements completely. What carrots could be offered? How about giving a bonus to a unit's combat strength based on the number of friendly units on adjacent tiles and/or the number of friendly units on tiles adjacent to the enemy unit. That would encourage spreading units out across multiple tiles to surround an enemy. How about another possiblity - make the experience gain of a unit dependent on the number of units in the stack. A unit alone gets a much larger bonus and hence more promotions than one stacked with twenty other units.

On the stick side a fairly simple option would be to remove the Civ 4 mechanic where the "best" defender is always presented to counter the attacker. Make it a random unit that's presented and the SoD becomes very unfavourable. Weak units like artillery would then need to travel in a separate stack behind a good all round defending stack or risk being easily destroyed.

Exponential penalties FTW (or exponential bonuses, same thing). Simply apply maintenance and combat penalties to units in stacks of increasing sizes, and voila, SoDs solved. I don't like the idea of making it random as you suggest, as that would mean that large stacks would work sometimes, and not work sometimes, but wouldn't be predictable, either when attacked or in defence. Applying a penalty rather than changing the rules was all that was required.
 
Camikaze said:
This seems the most likely solution; it wouldn't require a complete removal of 1upt. But it isn't the best one. You'd still be limiting the player. The only problem you'd eliminate is congestion.

I agree it's the most likely solution to see in a patch for practical reasons, but it's a bad case of the tail wagging the dog. It's bending the whole game around making 1upt work, not considering whether 1upt is becoming more trouble than it's worth.

Well, I've said why I wouldn't like this solution, and I find it difficult to think it would be better than 1upt. What you'd gain in terms of lowered congestion you'd lose in terms of warfare gameplay. As someone pointed out in another thread recently (can't remember who or where exactly), allowing 2upt would remove any need to protect your weaker ranged units. The game would essentially be groups of 2 moving around together instead of 1's, but with less tactical variation.

I'd be thinking of more than 2upt. I was mainly highlighting that congestion gets exponentially worse the fewer units you allow per tile. 2upt would actually have worked better in Civ 4 than it would in Civ 5 due to the greater number of units needed in each stack to counter any possible attacker.

Dear God please no.

Just a suggestion. I was trying to come up with something that would allow tactical and strategic on the same map sensibly. The current set up doesn't work, and I know plenty of people wouldn't like a full Total War style separation.


Exponential penalties FTW (or exponential bonuses, same thing). Simply apply maintenance and combat penalties to units in stacks of increasing sizes, and voila, SoDs solved. I don't like the idea of making it random as you suggest, as that would mean that large stacks would work sometimes, and not work sometimes, but wouldn't be predictable, either when attacked or in defence. Applying a penalty rather than changing the rules was all that was required.

Actually exponential penalties are not the same thing as exponential bonuses at all. Firaxis realised that when they were considering Dark Ages back in Civ 3. The problem is when you're essentially punishing the player for doing well, which is distinctly unfun. In this particular case I really can't go with a mechanic which makes it actively negative to have more troops on hand - it removes the incentive to work on the unit building aspect. Better to regard SoD as the lazy baseline, and give rewards for avoiding it (bonuses for units on their own would need to be carefully thought out to allow for it being far more common in early stages of the game). Hence my suggestion of bonuses for units on neighbouring tiles, since it would introduce a tactical element, but more troops would still always be an asset, not a liability.

Increasing maintenance for stacked units is definitely worth considering, but I wouldn't really want to see a penalty for moving units in stacks far front enemy lines. One of the snags with 1upt is also that every unit has to be moved one at a time, which is just irritating when there's no tactics involved.
 
@Schetter - I guess you said you have tried this, but IMO it would work really well if it was possible to 1UPT for each unit class ( ie... Civilian, Melee, Range, Seige, Mounted ). So you could possibly have 5 units per tile. That would limit a lot of congestion, plus allow some protection for non-melee-combat units ( ie... Ranged & Seige units, plus give some defensive ability for weak defensive mounted units). If this was possible you could also drop range attack down my 1 tile, but if a range unit gets the extra range promotion, maybe making it have a penalty to fire from the second tile away ( ie.. the range unit only gets 75% or even 50% of its attack str ).

I use the Legions Mod ( unlimited units per tile ), but use those rules stated above for my game play, trying to stay in the spirit of 1UPT, but useing 1 unit TYPE per tile. With the Legions Mod you can have more then worker on a tile, but the tile can only be worked by 1 worker, so the second worker can only pass through the tile.

I like your ideas on the workers being consumed after completing an improvement. It would help a lot for game play, mostly with the AI making a ton of them then not knowing what to do with them. If you get something that works and is stable with that platform will you put a zip file for it on here so we could see how it works and then we can give more feed back. For me it would have to be downloadable from here though, the ingame MOD browser doesn't work for me - Thanks.
 
Actually exponential penalties are not the same thing as exponential bonuses at all.

Well, they are in the way I was meaning (apologies for not being clearer). Applying an exponentially increasing penalty to units in a stack of increasing size is the same as applying an exponentially increasing bonus to units in a stack of decreasing size. You're just attacking it from opposite ends.

Firaxis realised that when they were considering Dark Ages back in Civ 3. The problem is when you're essentially punishing the player for doing well, which is distinctly unfun. In this particular case I really can't go with a mechanic which makes it actively negative to have more troops on hand - it removes the incentive to work on the unit building aspect.

Better to regard SoD as the lazy baseline, and give rewards for avoiding it (bonuses for units on their own would need to be carefully thought out to allow for it being far more common in early stages of the game). Hence my suggestion of bonuses for units on neighbouring tiles, since it would introduce a tactical element, but more troops would still always be an asset, not a liability.

I wouldn't want a system that does that either. But if you apply the formula so that each individual unit in a stack becomes weaker when another is added, although the total strength continues to increase, then you are both disincentivising stacking and not penalising the player for having more units. The aim would not be to stop people from building more units, because that was not the problem. It would to encourage people to break up stacks. This would allow for tactical and strategic variation. For instance, if you were going through a chokepoint, you'd have to stack more, which would incur more losses (because each unit would be weaker when in a stack), but would allow you to take that chokepoint. You could then spread out when you have the opportunity. Weighing up the relative costs and benefits would offer more strategic and tactical variation than knowing that stacks are always best, or conversely that you can only have 1upt anyway.

Increasing maintenance for stacked units is definitely worth considering, but I wouldn't really want to see a penalty for moving units in stacks far front enemy lines. One of the snags with 1upt is also that every unit has to be moved one at a time, which is just irritating when there's no tactics involved.

Yeah, I guess there is the point of tedium in movement. But at the same time, such a penalty would be a good way of attacking the issue.
 
I'm really happy this discussion has continued rolling, thanks all. Was about to throw in the towel, but I'll continue trying to get 1UPT to make some modicum of sense...

@CBG - I'll post a link here when I've got something halfway presentable, I'll try to get things in a somewhat playable state this evening. I'll also crack open the Legions mod and see what they're up to.

I guess, as I think about it more, that the biggest problems with the combat system that need resolving first are:
  • Foot units being able to move and attack a unit two tiles away is simply too fluid a mechanic for a strategy game. I already tried to address this by dropping base movement and visibility down to one.
  • Ranged foot units, like Archers, are essentially completely out of balance with normal foot units, and are redundant along with siege and built-in city defenses. We should return such units to their traditional place as city dwellers by giving them a city defense bonus (if possible) and bringing their range and the defensive range of cities down to one tile, while making them less useful as offensive units; allow siege engines to fill that role, and remove the silly iron requirement for siege. That way we don't have to worry about trying to balance an awkward three-layer front line to armies. Either that, or we need to recognize that the archer and its derivative units are without a strong use case currently, and should just be taken out of the game.
  • To stack or not to stack. I'm going to give 1UPT one last college try, BUT 2UPT might be an inevitability here.
  • Unit diversity - is it worth trying to pare down the options? Right now it feels like the only thing differentiating most units is that they have different combat powers...and that's about it. We need to have more strict classifications of units and modify the properties of existing units so they fall squarely in a single bucket, with well defined strengths and weaknesses, so that's its obvious what units should be used for what. This would turn the question of 'what unit should I build' into something greater than a function of opportunity cost per combat power. It seems that the group of (melee, range, siege, mounted, recon/support) seems to be generally agreeable.

To be perfectly honest, 1UPT necessitates a more fine-grained grid, where cities might expand to occupy an eight or more tile radius. This would be a radical change that's honestly a bit far out of my reach considering my current schedule...but if someone else wants to try picking that up they should. Keeping the traditional grid scale was a design decision that clearly wasn't thought through in terms of the end game, directly resulting in the ridiculous carpets. Simply put, I think to tackle this issue of congestion, the only viable solution is going to be dramatic increases in unit upkeep costs.

I will likely drag over a few applicable changes from my larger mod that seemed to work quite well. Those being:
  • Mounted units require double the horses than they do in Vanilla. This was originally designed to combat Horsemen spam, but I think it really brings the number of mounted units down to an appropriate level, and allows them to be used in an appropriate manner - to hit the flanks of an opposing army.
  • Siege units require no iron. This was just plain dumb and not tested properly in vanilla; period.
 
On last note before I leave the thread until this evening and actually try to get some work done:

I generally agree with the notion that the military map and the economic map should really not be sharing the same domain. If they were considered wholly separate by the game engine, the whole issue of what to do with workers, and workers 'getting in the way' would essentially be solved. Poof. The two maps could have wholly separate tile scales without having wonky effects on each other. Unfortunately, it would seem to require modifying the actual source code of the game, and we haven't even gotten proper exposure to algorithmic functions yet...we can dream, though. :crazyeye:
 
Ok, here's the alpha build:

keekerdc's 1UPT (v1)
Extract to [My Documents]/My Games/Sid Meier's Civilization 5/MODS.

v.1 changelog

[Base movement should be one tile per turn for units with feet on the ground; with some exceptions.]
[Scale back the movement rates for all units to bring the military side-game back in pace with the strategic whole.]
* All units have their movement range per turn reduced by 1.
* Scouts, Workers, and Settlers remain at 2.

[Cities should not have such ridiculous ranges when most ground units will move one tile per turn, and should not be able to reach siege weapons firing from two tiles away. If a player brings siege to assault a city, they should be rewarded.]
* City attack range reduced to 1.

[An appropriate role for Archers are as city-dwelling reserve troops. Reduce the range to one tile, making siege a more obvious and appropriate choice for front-lines support[
* Attack ranges for archery units reduced to 1.

[Most vanilla units see way too far.]
* Base sight range for all units of 1 tile.
* Base sight range of 2 tiles for Scout, Settler, Gunship, AT Gun, Tank, Panzer, Modern Armor, Frigate, Caravel, Ship of the Line, Ironclad, Destroyer, Battleship, Submarine, Nuclear Submarine, Missile Cruiser, Carrier.
* Base sight range of 3 tiles for Fighter, Bomber, Jet Fighter, Stealth Bomber.

[Recon units should be able to move amongst combat units to provide extended lines of sight.]
* Scouts have 0 combat power; the engine treats them as civilian units.

[Iron requirements for siege weapons don't balance with the typical supply on any given map, and restrict their use significantly.]
* Iron no longer required to build Catapults, Trebuchets, or Cannons.

[Horsemen spam. That is all.]
* Mounted units require 2 Horses each.

[Implemented experimental 'disappearing workers' mechanic.]
* Workers are consumed after building a Farm, Mine, Trading Post, Lumbermill, Pasture, Camp, Plantation, Quarry, or Well.
* Workers cost 20 production.
* Workers have a instance limit of 5 per player.
* Workers have an instance cost modifier of 2 (unsure if the expected mechanic is actually implemented in the game's code).

[Units that are defending a ranged attack should not gain experience from it. An example is a barbarian ship attacking a fortified land unit turn after turn, allowing the land unit to promote rapidly while healing the very nominal damage it receives.]
* Units gain no experience after defending a ranged attack.
 
@schetter - I will download it this evening when I get home from work. I like the look of most of your changes. I have a few ?'s on some of the changes, but will hold off on any comments until I have at least play test it.
One quick question - You haven't changed anything yet on the 1UPT yet?? At least I didn't notice any mention of it in the change log.

Thanks for the download link here.
 
I'm going to leave strict 1UPT for now and try to get the military game to make sense around it; if that can't be done then I'll scrap it and dive into trying to justify a particular xUPT as making more sense.

I'll be looking next at using unit maintenance costs as leverage against the carpet of units effect.
 
Feedback - Played several turns tonight and I really like a lot of the changes that you have made, but of course there is a but, there is always a BUT.

Likes:
1. Unit movements and sight range is an improvement IMO. Took some getting use to the only 1 tile movement again. It doesn't take long for a person to get spoiled to moving 2 tiles at time.
2. City attack range is a lot better. It makes a lot more sense to me that it is possible for siege units to be able to set up and fire outside of city defensive range.
3. Removing iron as requirement for siege units is a great idea, didn't ever understand how a wooden catapult or trebuchets or the small amount to build a cannon had to do with iron.
4. I can live with the Horseman requirement.

Dislikes:
1. This one I thought I would like, but have found myself not liking it very much at all. A worker being consumed after building a improvement has become more of a hassal than an plus IMO.
2. Scouts not having at least defensive ability is not fun to game play either.
3. Units not able to get experience while defending is not favorable or realistic. A unit can learn from a experience of having to defend from an attack.

Suggestions:
1. Workers not consumed after improvement building, but only allow 1 worker per city owned. They can work anywhere ( ie.. If a player has 5 cities they can have 5 workers and all of them could work at one city).
2. Scouts stripped of attack combat abilities, but retain defensive abilities.
3. Units receiving defensive experience. I agree with the principle of what you was trying to accomplish, just not the approach. Maybe increase the attack damage of the barbarian attacking, so that the unit being attacked would die before it could even reach one promotion level. So the player has to retreat the unit or it is destroyed before it even gains one promotion, but has learned something from having to defend itself.

Thanks again for your work on this project. it has a lot of improvements to game play IMO.
 
Top Bottom