Civ V (or VI ?) - The Solution to all Problems - Simultaneous Turns

Ryoko

Chieftain
Joined
Dec 29, 2003
Messages
14
Might want to sticky this, it is the truth.

The solution to the problems with Civ these days is making the turns simultaneous. Each player (or AI) would plan their moves for a turn. Once everyone is done, the game figures out what happens.

Obviously this would require a new combat system, but it would make the game way more strategic. You and your enemies would simultaneously issue commands to your troops. At end of turn, units would move and/or attack as directed.

I think units should be split between military and civilian, also. That way workers and settlers could move independently of army units without interfering with them.

This would also solve the performance issues in the game. As it stands now, each AI and player moves in turn, each unit one after another. My mp3 playing software uses more threads than Civ 5! We have multi-core CPU's out the wazoo and this game is still doing things in one long line.

The game should be made so each player and AI plans their turn simultaneously. First the civilian layer would happen, each worker would do its thing and each city would complete its current construction. Then all the planned military moves would occur, and the game would sort it all out. It could all happen in parallel.

As a bonus, the computations for the enemy AIs could all occur in the background as the player uses the interface to plan their turn. Let the AI think while the player thinks. There is no need to have all the AI planning in between turns while the user waits patiently.

The game would become more strategic and also flow better. Hooray!
 
Moderator Action: Moved to Ideas and Suggestions. :)
 
I have thought about a system like the one you are suggesting. I'm not sure how to solve various movement issues it creates. Let's say we have the following scenario:


On an island of ten tiles two units are located (shown as the red and the blue tile). Each unit can move two tiles per turn.

Now let's say that the blue unit cannot loose to the red unit. The blue player therefor wants kill the red unit with his next move. He should be able to do this since there are no tiles that the blue unit cannot reach in one turn (so there's no tile the red unit can escape to).

Now how would you construct a system in which this is possible if both players select a tile to move to at the same time, without knowing which tile the other player will move to?

Let's say you add a unit order "attack unit" in which the unit will always move towards a targeted unit. Basically the red unit would be moved by the game logic first and after that the blue unit would be moved towards it.

This might work when only two units are on the field. But what if the red player has a second unit on tile 4 that is stronger than the blue unit, and orders it to move in between the red and the blue unit. Should the blue unit still try to attack the unit on tile 2? Should it attack the second unit if it blocks the targeted unit?

Imagine another scenario where you have a cavalry unit. It can move three tiles per turn and have a sight radius of 2 tiles. You wish to scout with the cavalry, but not move close enough to any enemy unit for it to be able to attack you. Now, if you tell your cavalry to charge out into the fog of war you might end up next to an enemy unit. If you do this in a regular turn-based system you can always change your movement order after each step. In the system you suggest it would not be possible. How could this problem be solved?
 
It is entirely possibly to implement. I'm not saying it would be easy to work out, but it's probably even been done before.

You would tell your cavalry unit what to do. At end of turn, the action would take place. Perhaps your opponent has moved a defensive unit into the path of the cavalry. A set of game rules would be needed to figure out what happens. If you are at war with that tribe then obviously the units would battle. Maybe during peace-time and in neutral territory, the units could pass without fighting.

If two players each tried to move to the exact same hex during peace, the game decides who gets there first, based on things like unit speed, roads, terrain, randomness, etc. The 'slower' unit doesn't make it all the way to the hex, but instead gets put as close as possible on its path.
 
pileofnuts:

Turn 1: Blue moves to tile 6 and red then has to move to 1,4 or 8 to escape blue's ZoC.
Turn 2: Blue moves to tile 5. No matter what red does he'll end up in blue's ZoC, but let's say he moves to tile 9.
Turn 3: Blue can hit red(rule: if unit is within your ZoC you are guaranteed a hit). Red has to choose between retreat and fight and at least do some damage.

An alternative solution is one where units consist of several regiments. So it's natural to assume that the stronger blue can split and move into both tile 5 and 6 on turn 1. Maybe even units should be able to attack units in adjacent tiles?

It's really not a problem if one unit is chasing another. That's retreating and your losing ground. Usually there will be a city or maybe a goldmine behind red that he wants to defend. Should he withdraw and wait for backup? Or maybe make a stand and fight, and at least do some damage(and just as a notice, unit healing based upon simply skipping a few turns should go away)? I call that interesting decisions.

Ryoko:
I find combat system where the actual movement/fighting happens after you hit the "end turn" button to be very interesting. Not only simultaneous turns, but that is one of the most interesting ways. It offers a few very clear advantages. You mentioned a few of them, but not multiplayer. At the moment Civ uses different combat rules in SP and MP which is far from optimal. The MP-experience would be a lot more fun if the combat model was specifically designed for simultaneous turns.

Be sure to check out Part_Time_Civers "Turns, Ticks and Theaters" thread in this sub-forum. It's not a combat system that has as strong support like the separate tactical map many people(like sullla) prefer, but at least some see the potential :)
 
Unit collisions could also take politics into consideration. Say two units move through the same tile. If their nations are trading partners and good friends, they pass without problem. If their nations have been at war in the past, maybe they skirmish a bit, like a limited form of combat. They take a couple shots at each other, and relations between the two are hurt a little bit, even though no war is declared.

Civilian units should be on a completely different system than the military units. You should be able to have unlimited non-combat units on any tile. If your worker is building a road in neutral territory, the Romans should be able to move their legion through that hex without there being incident. (Assuming you two are at peace.)

As for settlers, founding a city should take their entire turn. If two settlers are on the same hex, neither can found a city. However, if one player also had a military unit in that spot, they should be able to found the city and eject the enemy settler.

Perhaps it could be negotiable. You could contact the enemy king and bargain for the tile. After reaching an agreement, one nation would get to found their city there and the other settler would have to leave.
 
I forgot one thing. Download Castle Vox Axis vs Allies from here. I'm not saying Civ should have a rip-off like this, but it's an interesting thing to test. It even has a hex map :)

Perfecto! I figured someone out there had made a tile based war game with simultaneous turns. The 'Turns and Ticks' thread had some great ideas also.

The combat system would end up being quite different than civ as it is now, but done right, it would be better and fix so many problems. Multiplayer would be better, war more interesting, and it would help game performance too.
 
The more I think about this idea(and it has been in my head for a few months now) the more convinced I get that this is how combat should work in Civ.

In fact I'm so convinced that I've decided to try to schedule enough time to implement it. The SDK isn't out yet, but that doesn't matter since there's quite a lot of planning to do.

A simultaneous turns system can be quite complex if you want to. Supply, morale, attrition, Generals, ZoC, ranged combat, stationary, moving, facing, armor, promotions and really anything you can find in war games.

I'd like to keep it simple, at least in the beginning. A few reasons for that. I want a model where the AI doesn't suck, I don't want to turn Civ into a war game and it's a lot easier to make :) I also want to focus on the basics first rather than special abilities for individual units. Still, some of the more advanced rules might at least need to be thought of.

If you have any ideas, especially for the basic rules or just the feel and effects of combat, I'd like to hear them. Unfortunately I'm not that well into war gaming which I have to admit is a handicap for me.

Should units fight until one of the two sides are completely wiped out(like Castle Vox) or should a normal attack be like Civ5 where you only injure it? Does it even matter, giving that the number of dead units is going to be a lot higher.

Should units have 1 strength number(like Civ4 and 5) with bonuses against certain types or separate attack/defense-number? Maybe a combination? I don't know yet.

Stacking will of course be allowed, but I don't know yet how stacks should fight together. Penalties for large stacks? What unit types should there be?

I'll hope you have some ideas and suggestions :)
 
I honestly can't see how a system would work that would not result in more annoying movement. As it is, if you can see a tile within range, you can move to it (save 1upt restrictions). It's as simple as that. Requiring simultaneous moves would open up possibilities of conflicting moves (I'm thinking Diplomacy bounces), which simply makes movement more and more tedious.

If the issue you're attempting to resolve is one of not utilising computer potential, then is there a way to design it so that AIs plan their moves whilst you are moving, contingent on your moves, but still carry out those plans in a turn sequence?
 
Should units fight until one of the two sides are completely wiped out(like Castle Vox) or should a normal attack be like Civ5 where you only injure it? Does it even matter, giving that the number of dead units is going to be a lot higher.
Fight to the death would be a lot simpler. In particular it avoids trying to figure out where two enemy units would go if they try to move to the same spot, since one of them will die with a 100% chance.

Should units have 1 strength number(like Civ4 and 5) with bonuses against certain types or separate attack/defense-number? Maybe a combination? I don't know yet.
One strength number would be easier. Moreover it has worked well for civ4 and 5.

Stacking will of course be allowed, but I don't know yet how stacks should fight together. Penalties for large stacks? What unit types should there be?

How combat involving multiple units is resolved is a good question. To further complicate the matter note that simultaneous turns imply the possibility of multiple players meeting in combat in a single tile simultaneously.
 
The system would have to make Land+Naval combat more like Air combat... Groups of units would be different defensive (non movement) mission[equivalent of ZoC/Interception]
"Intercept" would be max range, no bonus
"Defend" some range, some bonus
"Fortify" no range (one tile only), max bonus

The issue of Terrain would make it slightly more complicated. (since "operational range" would be in movement points so it wouldn't be a perfect circle)

Similarly you could have
"Move"... stop if intercepted by anything that could threaten you
"Attack".. go for the target area even if you erncounter resistance
"Sweep".. attack everything enemy in an area along the way to the primary target.


However... I would like to see units looking a lot more like aircraft, ie units have "operational ranges" and they conduct missions (for Footmen ranges of of 5-7.)

The issue would be that, just like aircraft can't operate outside of cities, units that are far away from a friendly [not recently conquered/in revolt] city/base have reduced operating capacity... lower range and worse combat.

This could overall make the system more Strategic. (you don't 'move units from one tile to another' you 'put an army of units in the general area and give it a mission')

This helps to make exploration more spread throuhout the game... ie scouts can be good because they have large operational ranges... but they won't be able to operate more than X total tiles from your city.
 
I agree with Camikaze that movement in a system like this will be more complicated and very likely somewhat annoying. The games I have played using this type of system it always annoys me how much of a pain it can be to block or catch an enemy running around in an open field. The reason why I'm still interested in looking at developing this system for Civ is that I imagine it could improve the multiplayer experience by removing the current implementation of simultaneous turns.

An inherent problem with this system when adopting if for Civ is that it is really developed to be used in games where the playing field consists of much fewer tiles and units as a result do not move many tiles per turn. To reduce annoying behavior during movement, I would take note if this and reduce how many movement points units have in the game, and if possible, reduce the number of tiles in a standard map. Regular infantry would have 1 movement point, cavalry 2 and so on. If the map size is not reduced this change would force the turn limit of the game to be expanded, since moving around the world would take much longer.

To reduce the added complexity additional types of orders should be kept at a minimum.

A general rule I believe will be needed: All units deal damage before damage is absorbed
This will resolve situations like the following: Three units controlled by three different players are all in tiles bordering each other. A attacks B, B attacks C and C attacks A. If one unit deals damage before any other (reducing the damaged unit's potential damage output) the system will not produce consistent results.


Combat Phases

I would first divide the combat resolution into several phases. Here's what I came up with:

1. Bombardment
Artillery, bombarding ships and airplanes on bombing run orders are resolved first. The reason why I put this first is to allow bombardment to deal damage to units in a tile before those units get a chance to move away from the targeted tile.

2. Unit Movement
Units with movement orders are moved to their targeted tile. How to deal with various situations that can occur (units bumping into each other etc.) is discussed further down. Units on intercept orders deal and receive damage in this phase.

3. Combat Resolution
Each contested tile (a tile containing opposing units) is resolved. The system for doing this, I imagine, should not need to affect the other phases. I see designing this system as a separate topic. I would suggest using a round-based system where the units take turn dealing damage, and after a set number of rounds combat is ended.

4. Combat Results
Retreating units are moved or destroyed. Cities taken switch hands. Forced unit movement, like units retreating, might be hard to make predictable. Generally units should retreat to the tile from where they attacked or if defending, in a straight line away from where the attacker came.
But when defending, a unit's natural retreat target might be occupied by an enemy unit. And when attacking, another unit might have moved into the tile from which it came. The unit should obviously not simply run into the enemy lines, so to which tile should it move if all other parameters are equal?
A related issue is that the player might feel a unit should retreat to a particular tile (a hill, a tile easier to reinforce etc.), something that the game logic cannot solve.
Because of these two issues it might be best to not allow units to retreat and instead use a system like it's currently implemented in Civ5, where tiles are only taken when the defender is destroyed.

5. Fortification and Healing
Units on fortify order will finish their fortification and units healing will get their HP. The reason why this is done last is so that a unit cannot move within range of the enemy units and fortify before they can be attack.


Unit Movement

Two important potential problems I believe must be addressed in the movement system are the following:

1. Catching enemy units that is fleeing that I wrote about in my first post. Windsor's zone of control solution sounds like a decent way of solving this. If two units are next to each other, if one player chooses to attack and the other to move away, the attacking unit will deal damage before the moving unit escapes. It maybe shouldn't deal full damage, but that's a minor detail.
To expand on this system, faster units perhaps should be able to escape zone of control. A horse archer would be able to move out of the way from an attacking swordsman. This of course introduces the annoying behavior that the player controlling the slower unit would have to guess where the faster unit will move. If the player guesses right they will end up in the same tile and combat will occur. Since this is all about luck, I believe this is a big issue that needs to de resolved.

2. Two units that are next to each other and attacking each other should not simply switch places and avoid combat. Having the opponent avoid combat like this without moving back would be annoying. I would suggest that both units are thought of as attacking and no one receives a defense bonus from either terrain or abilities.
I'm not sure how to deal with units moving into any of the two tiles the units are attacking from. A unit moved into the friendly tile was not expected to be used in the attack by the player. But at the same time it were most likely meant to be used as a defender of the tile. Having the game rules commit the unit to the attack can in some cases be what the player wants and in others not. I can't think of a solution to this problem.

Moving Units To Their Target
A system that is easy to understand and predict would be to to it in steps. all units with movement points left move one tile. Any opposing units that end up in the same tile are stopped and will not move any more during the movement phase. Repeat this until all units have spent their movement points, reached their target or are tied up in combat.
This system can also be reversed so that the unit/units with the most movement points are all moved one tile first. Then make a new list of the units with the most movement points left and move them. Repeat until all units have finished moving.

Issues might arise when unit movement paths cross. Unit paths can overlap perpendicular or parallel to each other.
1. Units may cross in such a way that they at some point will move into the same tile.
2. Units may move next to each other and then attempt to switch places.
3. Unit movement paths may cross, but the units will not enter any tile at the same time.
4. Units may move next to each other and then one unit moves to an unoccupied tile while the other moves to the tile which the first moved from.

The above system would create rules for this. But I'm not convinced it would always create the best results. Reducing the number of movement points units have would reduce these issues. But they would still occur.


More Issues

Aircraft Intercept
The problem arises when multiple aircraft enter the intercept area of an opposing aircraft. To which aircraft should the interceptor deal damage? How should damage dealing to the interceptor be handled and affect combat with other aircraft?

Possibility of Using Blitz
In my opinion, it should be possible punch holes in the enemy front and attack through the holes at units in the back on the same turn. To allow this, it should be possible to give certain units multiple attack commands. First the unit attacks its first target, and if successful, it continues and attacks its second target. Now I can imagine this not creating a number of new issues that needs to be dealt with. All combat would for example not be resolved in one go since the blitzing units would need to move to their second target.


Resources
An interesting game to look at while designing this system would be Gary Grigsby's World at War: A World Divided. It is a more detailed take on the Axis & Allies concept. Definitely a great game.
http://www.matrixgames.com/products/330/details/Gary.Grigsby's.World.at.War:.A.World.Divided

It uses a standard turn based system mixed with a system where the player can move multiple units before choosing to resolve contested territories. To explain: When playing Germany the player would on the first turn start by move several tanks, a few paratroopers, some infantry and aircraft into the low countries (one territory in the game). The player would then resolve the conflict in the low countries resulting in it falling into German control. Note that this does not end the turn. So before the Allied player have a chance to respond, the German player moves several units into the territory containing Paris, circumventing the Maginot Line, resolves that conflict and capturing Paris. More unit movement and combat would be carried out on that turn before the German player would choose to end the turn.
That system worked great in that game. But since it doesn't help improve the multiplayer experience in the ways I want I have not considered it for Civ. But some of you might find it interesting.
 
I would think that if the real issue is the multiplayer experience, then the solution could be to just change how simultaneous moves work in those games, without worrying about changing the single player system at all.
 
Hey nuts, great post, I'm happy to see other people are putting some serious thought into the idea of simultaneous turns.

Some ideas I have had about combat and/or unit collision:

High speed units go first. Say the fastest unit on the battlefield has 4 moves - all units of speed 4 will make their first movement. Maybe the unit moves into a grassland hex and uses 1 move point. It will continue its next action when the units with 3 movement points go. If instead the unit moves into a hills hex and uses 2 movement points, the unit will only continue on its path/mission when the units with 2 movement points left are moving.

if a hex is empty, the unit moves into it. (Obviously.) If another unit happens to get there first, the slower unit can either give up the hex, or fight for it. This would depend on your policies and what you commanded the unit to do.

If multiple units enter/attack the same hex in one move, combat occurs. Each unit will distribute its damage among all the available enemies. Perhaps civs might distribute their combat efforts more against another civ they dislike. You could also take into account which direction units were attacking from. (Units moving into the hex from the north are more likely to engage the units from the north-east than they are from a 3rd civ to the south. In the end, who gets control of the hex is the civ with the most merit, based on remaining unit strengths and movement speeds, etc.

Units would need a few options for their orders.

"Peace Move" - Move into a hex but never at the cost of starting a war.

"Opportunistic Move" - Move into a hex if you have a good chance of winning battle, halt movement if combat looks scary. (Like a 1/2 spearmen moving into a 2/4 musketmen unit - bad idea.)

"Forced Move" - Move into a hex no matter what, fighting anyone in your path.

"Attempt Attack" - If an enemy unit is in range but could possibly run away in time, you can attempt to fight it. If you intercept / collide with the unit you fight, but if they back away from you, no combat occurs.

"Force Attack" - If an enemy unit is in range with no chance of escape, your unit will force it to fight. Say two slow units are adjacent to each other. Even if Civ2 tries to run away from its hex, Civ1 is so close to it that they are forced into combat.
 
Perhaps it would be interesting if there were only two choices, "Move to" and "Attack to", and all the results of whether units would fight or not were based on a combination of their orders and your social and/or military policies.
 
I also think units should be able to fight in unclaimed territory without causing an all-out war. Say you and an enemy civ move into the same hex simulaneously, then fight, and you win. The enemy civ will remember this and hold it against you, but not necessarily declare war. They might even ask for reparations to forget about the incident entirely.

If it were military units, they could skirmish and hurt each other a bit. For non-military units, they could escape, or perhaps be killed or taken hostage - based on civ policies.
 
I'd like to give a few comments on how I think a few situations that have been mentioned.

Let's start by this:



This is problem in a 1UPT system during peace. I'm still thinking in stacks where this is not a problem at all. For 1UPT solution is either a regular bounce where no one gets to enter the tile, a 50/50 draw or a workaround with 1UPT per civ. My choice would be bounce but units within their own borders gets priority.

A variation of this problem is this one:


Same situation, but Red uses a fast horse rather than a slow moving sword. I would still vote for priority within own borders, but in neutral territory the horse would get the tile. I prefer a ticks system rather than a simple "speed/initative" number. Ticks/speed will also matter when calculating possible combat during war when two paths cross.

If horse starts two tiles away then the horse no longer would have a speed advantage and if bounced the horse would only move 1 tile.

3 units attacks same tile:


All units are at war with each other, what happens? I think this situation is easier to look at this way:


So what happens is that your army split in two(or more depending on number of hostile civs you encounter) and you get 3 battles. Once one side is killed the two remaining sides fight out like a regular battle.

Two adjacent units attack:


This makes a fight. Swapping might make sense for some games, but not Civ. The swords don't get defensive bonuses since they both attack.

Let's make it more complex:


The easiest solution is to first solve the sword-battle like in the previous situation and then treat the 2nd battle like a regular situation like this:



However, I'm tempted to let the archers help swords from the beginning. That would make moving in the right defender more attractive. But no unit is receiving any defensive bonuses and archers don't deal damage to each other.




So what happens here? Well, to me this situation is only 3 * this situation:


So whatever happens here should apply to the previous situation. As mentioned, I believe there probably should be a ZoC system where the attacking units gets a hit (but as pileofnuts said, not as much as a regular fight. This makes for interesting situations where you have to decide if you want to retreat or get some defensive bonuses and at least do some damage yourself.)

Then we have withdrawal. I generally don't think withdrawal is needed, but if it is included, then the last priority(if two tiles are equal) will be home, like here where the blue units natural path is blocked by a peak:


Units moving several tiles also give some situations. Like what happens here?



Do they fight? Swap? Or maybe this situation should be seen simply as this one:



Here it's more obvious that the units do not fight. They swap.

I really don't mind the "how can I catch that guy"-problem. First of, this is not just a problem it's also a strength that a small force can break enemy lines and create some chaos and predicting enemy movement is a skill. Secondly, with infantry as 1:move: ZoC and units on crossing paths not swapping but fighting this really limits it. And we have forts for defensive structure which can intercept. Really, if a bunch of knights are pillaging your countryside you have done something wrong :p

I generally think it's not that hard to give the player easy rules that sounds logical and are easy to learn. Some situations requires some care in how you actually code the rules, but from a player perspective its usually pretty easy to understand what should happen.
 


Part of this type of conflict should consider unit orders. For simplicity let's say each horse has 2 moves. If each horse is ordered to 'move' to the other hex, maybe they cross paths but don't fight.

What if one or both of the horses is ordered to 'attack' the other? They each move one hex each initially. Then since both units are on adjacent hexes, they may be forced to fight each other since one of their leaders has ordered it. If both Civs have chosen 'attack', then it is a fair fight. If one has chosen 'attack' and the other 'move' - the attacker should get a slight bonus.

Maybe CivA has ordered its knight to 'move' while CivB has ordered its knight to 'attack'. The 'attacker' unit will fight the 'mover' unit when they are in range, with a bonus. If both have chosen 'attack', then they just duke it out without bonuses, aside from whatever comes from the terrain.
 
I like tile capacity limitations like 1UPT, but I think it should have some modifications. What if things like cites and/or forts increased the capacity of a hex? For example, 2 units could fit in a fort, 3 units into a city.

Players should be able to order unlimited units into a tile. If I have 10 tanks in range, they should all be able to attack a certain tile. Perhaps they battle 9 enemy tanks and one survives, then that unit gets possession of the tile. What if it is 10 tanks vs one rifleman? Then the strongest/fastest tanks of the battle who survive will move into the hex, the rest will end up as close as possible along their paths.

If CivA, CivB, and CivC all attack the same tile, the strongest faction remaining after battle gains possession of the hex. This could also be affected by things like unit speed and unit experience.

Say you choose the 'move' action for a unit to enter a tile. If there are enemy units who are also moving out of it, then no conflict occurs. If instead you choose the 'attack' option, if the enemy if holding their ground you fight all-out, but if you 'attack' while the enemy is 'moving' away, your forces could do a limited amount of damage to them while they retreat.

I'm on the fence about the idea of ticks. Could be really good, but this is a game so we don't want to make things more complicated. Obviously things like attacking across a river should incur attack power pentalties and/or speed penalties. Terrain changes could count into this too. Moving from a grassland to a jungle hex should slow the troops down some.

Regarding ticks vs no-ticks - even if the math is complicated, that is OK. At long as the basic results and principles are part of the battle are easy for the player to comprehend, then that is OK. As long as the player can understand combat variables like terrain, rivers, unit speed, and who got their first, then the player can make sense of things and have a fun game.

For example, a swordsman attacks an adjacent hex with an enemy swordsman in it, across a river. The river will slow down the attackers. If the enemy swordsman holds its ground and defends, the attackers will be penalized in combat for crossing the river. If the defending swordsman flees the tile, perhaps the river will give it enough extra time to get away without taking any damage.

If multiple civs enter a tile with their units simultaneously, things get complicated. I think each unit should deliver its damage spread out among all available enemies. Say CivA enters a tile with a few units, and CivB and CivC (allied) also are attacking the tile. Damage caused by CivA will be spread among the units of CivB and CivC. The strongest remaining force at the end of it all will get the tile. Say CivB is strongest at the end, it gets the tile. CivC's units won't make it into the tile, and their movement will be halted as close as possible.

It might be good to differentiate between units and hexes when making unit actions. Your swordsman should be able to attack the hex itself, or the units in it. If you attack a unit but it is moving out of the hex, maybe it gets away. If you attack a hex and the unit in it moves out, it gets away free, because you are focused on the hex and not the military unit itself.
 
Top Bottom