Stacked vs Single Unit Combat - The Battle Continues

[Kuciwalker]

Originally posted by wrylachlan


I like ten to 15 or so based on the games of civ3 I've played. I think that would make enough stacks to allow you to have a more tactical game, but not so many that you're in micro hell dealing with a hundred stacks.

Assuming your choice of four is not arbitrary, why four?

Because it retains many of the qualities of C3 combat, and it means there is a lot less of a combat advantage to larger stacks. You won't have as many "useless" stacks that are too small to be effective.

[Merp]

Okily Dokily...

Assuming that stacked combat holds the day, I would propose the following features:

An optional tactical mini-game
Whereby the same ruleset is employed whether you show the mini-game or not. ie it can be automatically resolved.

No player input required or allowed once actual combat has commenced (not including setup)
There are number of benefits from this.

a) keeps the player from micromanaging *too* much , thus its simpler for the AI to compete.
b) removes any benefit from having lightning reflexes
c) allows the mini-game to remain optional

Army-centric rather than Unit-centric
This obviously applies to more areas than stacked vs. unstacked combat, but it relevant here. By this I mean that there are no unit stacks, only armies (which, I suppose, could be stacked in a limited fashion). The game Medieval: Total War is a good example of this.

Armies have unit limits preventing more than a certain number competing in a battle. This says nothing about tile limits, thus preventing AI pathing problems. Army unit limits could be affected by the leader (see below), the units involved and the possession of tech advances.

Note that when I use the term 'army' I am equally referring to naval or (possibly) civillian formations.

Armies have a general / leader
On its simplest level this affects morale and provides a rating for the quality of automated tactical decisions, a talent for commanding certain units, bonuses, etc. Once again Medieval: Total War provides a good example of this.

On a broader scale generals or leaders can be produced from academies or other city improvements, gain experience from battle to battle and have a quantified loyalty to the current regime. IMO a great way of personalising otherwise generic formations.

Battlefield Maneuvers should introduce variability
Battlefield maneuvers would be generic changes to the way the battle is resolved.

a) Battle field maneuvers should be defined extensibly (see combat AI, below) and apply to specific situations depending on the commanding general, the units involved, tech advances etc.
b) are not necessarily shown in any great detail graphically
c) are always defined in the battle setup or left to the general
d) should be define a simple exclusive list of options with minimal user input. Example: Pin Down and Flank, Fighting Retreat, and Charge. Combinations - if desired, can be defined by modders (see combat AI, below). The point being that the player (or leader AI or rival AI) doesn't need to spend five minutes setting battle options that aren't understood.
e) can effect the location that units are in at the end of the battle. e.g. a player can 'win' a fighting retreat and end up in the square behind him. A successful flank might leave the player behind the position of the enemy. This could be optional
f) can have outcomes that are not completely win or defeat. Both armies could survive in some form.

Combat AI is flexible and moddable
Obviously AI is a broad topic not limited to the discussion at hand. I would however like to see the following relevant portions of an over-arching AI setup though.

a) Combat (and all other) AI defined in user-editable rules in text files
b) These rules can both provide both additional flexibility and choices (define a new 'flanking' maneuver, for example) and define automated AI and player behaviour (e.g. Retreat if only one frontal unit remains and if number of ranged units is over 1)

Combat AI handling of this type would be good, so that players who have no wish to see the minigame could still ensure that 'under the hood' everything is still happening exactly as they want it to.

The key objectives of this system

1. Ensure that the minigame remains optional
2. Introduce more depth and realism
3. Reduce micromanagement - if desired
4. Making sure AI is moddable, for both the lazy player and the AI.

Thoughts?

[LDiCesare]

I don't like the idea of minigame. Even an optional one. Optional minigame means resources dedicated to coding it, its graphics and debugging it, which means either a more expensive product or less details in other areas.
All ai's in all games that have a TB tactical game with remotely complex rules (more complex than chess, and even worse, moddable) have always been way below the best players. Thus I'd feel forced to play the minigame for tough battles because I'd use a better setup than what the ai would chose for me. All this to say I will definitely not buy a civ game with a minigame combat.

Could you define what you call an army versus a unit stack?

Whatever the objectives of this system, the side effect is that it enhances the wargame aspect of the game even more, which is a thing I don't like.

Stacked combat streamlines combat. A minigame would only make combats more lengthy and complex.

[Fosse]

I don't know exactly what everybody means when they say Army or Stack... but in an earlier thread a few people were able to agree on the following definitions:

A stack is simply a group of units that move together, as implemented in PTW or Conquests. They don't fight together, or take orders together. They just move from point A to point B together.

An army is a stack of units that fights together.


I agree 100 per cent with having NO tactical minigame. Let the player craft his armies, and let the computer fight the battles.

[Kuciwalker]

How about this - "armies" work exactly like they do in C3 (or even better, C3C, except for the movement bonus), except you can load and unload units from them. Armies are really cheap to build (i.e. about 10 shields) and there is no limit to the number of armies you can have. Armies can be built in any city. C3C's army system rewards combined arms (if you don't trust me, ask Theseus about mixed-unit armies ).

[Fosse]

Because Civ 3 armies are still one unit at a time fighting another unit. My swordsman in the army walks up to your spearman in your army, and when one of them gets down to one HP he retreats, and somebody else steps up to the plate.

Blech.

Battles should not be relay races.

If all of the units are fighting each other at the same time, then battles are more realistic, interesting, and shorter. One army fighting another in Civ 3 takes forever.

[Merp]

Originally posted by LDiCesare
I don't like the idea of minigame. Even an optional one. Optional minigame means resources dedicated to coding it, its graphics and debugging it, which means either a more expensive product or less details in other areas.

Well thats a fair enough comment.

All ai's in all games that have a TB tactical game with remotely complex rules (more complex than chess, and even worse, moddable) have always been way below the best players. Thus I'd feel forced to play the minigame for tough battles because I'd use a better setup than what the ai would chose for me. All this to say I will definitely not buy a civ game with a minigame combat.

Okay a few things here. I believe fervently that whatever AI is included at release, it will be suboptimal. Futhermore, having seen the amazing things modders at this and other sites have done given a basic infrastructure, I have great confidence in their ability to make challenging AI. In reality I doubt that players will apply all of their thinking power to every single engagement, and often the basic level of strategy is pretty consistent.

To clarify, when I say 'minigame' I mean something very similiar to CTP - if you wish you quickly review the setup, select any other options and go! Certainly quicker than many large unstacked battles.

Could you define what you call an army versus a unit stack?

I agree with the definitions above, with the following clarification: an army is a group of units semi-permanently moving and fighting together. A unit stack is collection of units that just happen to occupy a tile at a given time.

Whatever the objectives of this system, the side effect is that it enhances the wargame aspect of the game even more, which is a thing I don't like.

Once again, this comment is fair enough. But if it was optional, and you had faith that most battles could be left to the ai, would it still be a problem?

Stacked combat streamlines combat. A minigame would only make combats more lengthy and complex.

Complexity equals game depth, and not necessarily complication. As I alluded above, the idea I propose would certainly be shorter than the current unstacked system regardless.

Any thoughts on the ideas about army leaders or generals and their bonus, talents, loyalty etc?

[LDiCesare]

To me, CtP doesn't have a minigame but a battle report. Changing the setup by hnad would take a few seconds, and be fast, but I still don't think the ai could make it. If you make the setup more of a rock-paper-scissors setup than the CtP setup (where you realise that the ai put your artillery in front and the tanks behind it), then the ai could cope.

I'm not sure about leaders. Leaders who actually help the units in their stack (or army) would be good. CtP2 had this in the Alexander scenario, although veteran status didn't seem to do anything.

[Frozzy]

Well, CtP had the 3 tiered battle system, where the frontline was hand-to-hand, second was ranged, third was non-combat.

How about a leader adds one attack point to each unit, but is in the third tier. So (going with the 9 unit cap rom CTP1), the front 4 HTH units have +1 attack, the 4 ranged units have +1 "range" (attack), and the back guy does stuff all.

[MrBaggins]

Originally posted by LDiCesare
To me, CtP doesn't have a minigame but a battle report. Changing the setup by hnad would take a few seconds, and be fast, but I still don't think the ai could make it. If you make the setup more of a rock-paper-scissors setup than the CtP setup (where you realise that the ai put your artillery in front and the tanks behind it), then the ai could cope.

I'm not sure about leaders. Leaders who actually help the units in their stack (or army) would be good. CtP2 had this in the Alexander scenario, although veteran status didn't seem to do anything.

Veteran status just got fixed

[MrBaggins]

Originally posted by LDiCesare
I'm not sure about leaders. Leaders who actually help the units in their stack (or army) would be good. CtP2 had this in the Alexander scenario, although veteran status didn't seem to do anything.

Veteran status just got fixed

[LDiCesare]

Veteran status just got fixed

Well, CtP had the 3 tiered battle system, where the frontline was hand-to-hand, second was ranged, third was non-combat.

In CtP2 there was front row, flankers, and support. Some units like spies/slavers/settlers were non combat, but this was not because of their tier, but because they couldn't attack. Leaders could provide boosts like you suggest by being in the army/stack.

[ThePlagueRat]

I dont't think the creators of civ4 would even try to imitate ctp combat system. (even though they should) Deep inside they know is superior to their own system (civ3), and they will probably try to make their own system better instead. I strongly doubt they will make a combat system as good as the ctp system... if they do, they'll have to rip off a great deal.

[CerberusIV]

Having read the entire thread I feel entitled to post.

The unmodded CTP2 system wasn't good. With half a dozen tanks, a dozen bombers and then machine gunners dropped in to garrison from helicopters you could overrun every city on the map one by one.

Civ3 isn't brilliant and units don't really fight co-operatively in any meaningful way in the game.

Making it complicated with a mini battle view or lots more parameters isn't going to be popular with many people.

So, what is possible?

Civ4 won't feature CTP style combat, the best we can probably hope for is an improvement on the army system and maybe some changes in the way bombardment units work.

What would help IMHO is a stacking limit per tile, maybe dependent on terrain type, which would represent the difficulty in bringing together and supporting a number of units. That would limit the sheer number of units that could be brought to bear on the critical point.

It should be possible to unload units from an army. The 2 key changes I would like to see are integrating bombardment units into armies - maybe so they fire during the combat resolution - and giving units a support bonus depending on what other units are in the same tile. A naval example of this second point would be to give a defensive bonus to naval units if they are in the same tile as a carrier (to benefit from the radar and plotting teams on the carrier) when under air attack.

[MrBaggins]

Originally posted by CerberusIV
Having read the entire thread I feel entitled to post.

The unmodded CTP2 system wasn't good. With half a dozen tanks, a dozen bombers and then machine gunners dropped in to garrison from helicopters you could overrun every city on the map one by one.

Thats got more to do with the fact that the AI can't deal with concentrated attacks. This is demonstrated by the fact, that if you play vs. a human, the system works as advertised.