Changing the Combat System for the Better

[Aussie_Lurker]

I am creating this thread, on Last_Conformists advice, to discuss major anomolies in the civ3 combat system-and how they could be solved for CivIV!
Now, though I have never seen a 'Spearman beats tank' scenario, I have personally seen a LOT of ridiculous outcomes for combat-both for AND against me! A major cause of these outcomes, IMHO, is the fact that the Civ3 combat system subsumes both units' ability to hit AND do damage into a single calculation-namely (AS (attacker)/(AS (attacker)+vDS (defender)).
Now, to use this in an example: if you have a tank attacking a pikeman fortified in a mountain, then the tank has a ((16/(16+6))*100)=72% chance of hitting AND doing damage to the pikeman. The problem with this, though, is that this means that the pikeman has a 28% chance of hitting and damaging the tank-which is totally ludicrous!!! So, how do we solve this situation? The answer, I feel, is 2 or even 3-fold:

1) Make combat simultaneous-this means that ALL combat occurs AFTER all units have been moved, and breaks down the distinction between a seperate attacker and defender (making it more of a melee). Instead each unit in the combat gets to attack AND defend. To use the example above, the tank will have a 72% chance of hitting the pikeman each phase, but the pikeman will also get an ATTACK, each phase, which gives it a ((1/(1+8))*100)=11% chance to hit (less than half the chance of the current system).

2) Seperate out the ability to HIT, from the ability to do damage, by reintroducing firepower and introducing armour. Firepower would be an integer multiplied by the success of the hit to yield damage potential, and armour would subtract from damage potential to yield actual damage. Again, to use the example above. Lets say that pikemen have a firepower of 2, and armour of 1, and tanks have firepower of 6, and an armour of 3. If the tank got a result of .20 (chance of .72), then it will do 0.52*6=3.1 HP damage, minus the pikeman's armour of 1 to do 2hp damage (rounded) on that hit. The pikeman, OTOH, gets a result of .05 (chance of .11) to give it a 0.06*2=.12 HP damage. Even rounded up, this will mean that any damage the pikeman does will bounce off the tanks significant armour! Even firepower alone would, along with solution (1), greatly improve the combat system!

3) Another way might be to alter attack and defense strength values (to stretch them out more) and to change the way combat odds are calculated-such as a direct Attack Strength (attacker)-Defense Strength (defender)/100 (to convert it to a value between 0.00 to 1.00) In such a system, the tank above might have a MUCH higher Attack and Defense Strength (say a 48 and a 24, respectively), as would the pikeman (say a 2 and a 6, respectivly). This would give the tank a (48-12)/100=36% of hitting the pikeman, and the pikeman a 2-24/100=-22% chance of hitting (i.e., can only hit on a 0).

Ultimately, the first two solutions could work quite effectively on their own, though a change to the values involved in the combat system could help as well!
Anyway, I would like to hear any thoughts, criticisms, suggestions??

Yours,
Aussie_Lurker.

 

[tig3001]

I like your first two points quite a bit, though stretching out the values would be better then the current system too.

 

[rhialto]

If you make combat simultaneous as in point 1, you immediately render all spearmen-type units obsolete. It might take a long time for any enemy to hurt them, thanks to their high defence, but if they can never effectively hit back, they are doomed to lose the majority of combats they face.

I'm in favour of firepower and armour, broadly as you describe, but I think your maths is a little weird - it seems to be based on averages, which beats teh point of a random number generator.

If we have firepower and armour, I don't think we'd need to stretch out the attack/defence factors. Consider that civ2 had much less 'stretching', but spearman kills tank was relatively rare. This might however have to do with the hit points stat. SOme playtesting will be needed.

 

[Aussie_Lurker]

I doubt it will render them obsolete, Rhialto. To prove this, consider the following example: two archer units go head to head with two spearmen units (unlikely, but still...) the spearmen have an AS of 1 and a defense strength of 2, and the archers have an AS of 2 and a DS of 1, thus the spearmen have a (1/(1+1))*100=50% chance of hitting the archers, wheras the archers have a (2/(2+2))*100=50% chance of hitting the spearmen-evenly matched.
Of course, due to their range, the archers WILL get a free hit against the spearmen before combat begins (yet I would think that archers would get some kind of AS penalty after this first hit, to reflect how poor they are at hand to hand combat, and reinforcing the need for mixed groupings of units to overcome this penalty). Even against swordsmen, with an AS of 3 and a DS of 2, the odds would be (1/(1+2))*100=33% for the spearmen, and (3/(3+2))*100=60% for the spearmen. Hardly brilliant, but by no means obsolete.

Also, remember that under my simultaneous combat system, all units within a stack fight at the same time. To use the example I just gave, you would probably have each archer pairing up against each spearman. Within such a system, if you have one group of units larger than the other, then some units in the smaller stack will be forced to fend off more than one opponent. Under these circumstances, such units would be at an attack and defense strength penalty for each additional opponent, depending on its 'Blitz' value-of course, such numerical superiority is a double edged sword, as too many units will cause the larger stack to fight at a penalty to morale and AS/DS, due to poor command and control of such a large group! (not to mention potential for 'collateral damage')

As for the rounding in my armour and firepower example, that was just me being lazy. I agree that the numbers probably should be left as is, or only slightly rounded! You also are absolutely correct that with firepower and armour, AS and DS can be left close together, whilst still allowing for large tech level differences between units. Increasing the HP levels would also make this significantly better as well!

Anyway, Rhialto, thanks for supporting my ideas in the broadest context, I hope those areas of disagreement have been cleared up with my clarification!

Yours,
Aussie_Lurker.

 

[Jorgen_CAB]

I do agree on the simultaneous movement of units and combat, though I do think that attack and defence should be dropped altogether.

Instead all units should have a combat value, certain units will have certain advantages such as archer will have ranged combat. All factors should be considered in the fight and all units should fight together and the computer should calculate the result and present it in a small combat simulator (much like Europa Universalis from Paradox Interactive).

A light spearman unit might have combat value of 2, mobility 1, firepower 1, armour of 1 and bonus against mounted (+2vs mounted).
A heavy swordsman unit might have a combat value of 3, mobility 1, firepower 1 and armour of 2.
An archer unit might have a combat value of 1, mobility 2, firepower 1, armour of 1 and a ranged value of 2.
A light cavalry might have a combat value of 2, mobility 4, firepower 1 and armour of 1.
A heavy cavalry unit might have a combat value of 4, mobility 3, firepower 1 and armour of 3.


An army of 3 Light Spearman, 3 Heavy Swordsman and 3 Archers.
Combat value 18(24), Mobility 1.3, Firepower 1, Armour 1.3, Ranged 6

An army of 4 Light cavalry, 4 Heavy cavalry.
Combat value 24, Mobility 3.5, Firepower 1, Armour 2, Ranged 0

These values would be calculated and depending on terrain and if one are defender or even both might be attacker. Before a combat you should get the option of trying to retreat, the mobility factor will be used to se of the retreat is successful.

This would be a very good and realistic combat system, morale should also be added.

 

[The Last Conformist]

If we have firepower and armour, I don't think we'd need to stretch out the attack/defence factors. Consider that civ2 had much less 'stretching', but spearman kills tank was relatively rare. This might however have to do with the hit points stat. SOme playtesting will be needed.

The great predictability of CivII combat was almost entirely due to the higher HP counts. Pretty few units actually had Firepower higher than one.

 

[Khan Quest]

What an elegent solution Aussie. The fix is the best I've read and would be easy to impliment.

 

[Aussie_Lurker]

Actually I have been thinking that, were my system to be adopted, then the nature of Attack Strength and Defense Strength would ultimately change. For instance, Attack strength currently represents your ability to hit AND do damage, and your defense strength is both your ability to avoid being hit AND avoid damage. With Firepower and Armour, though, you are seperating things out. Now, AS is simply a measure of your 'accuracy' in hitting an opponent, wheras Defense strength represents your ability to avoid being hit altogether.

In addition to these stats, it would also make sense to give units a 'range'-from 0 to say 6, and a rate of fire between 1 and 3. Why? Well, in combat, 'ranged' units would get a number of free attacks equal to ((diff. in unit ranges)-(diff. of opponents movement rates))*Rate of Fire).
So, for example, if a stack of modern infantry (range 4) were in combat with a stack of longbowmen (range 2), then the infantrymen would effectively get 2 free attacks-without fear of a counterattack-before the longbowmen were within range to hit back. Against a mounted bowman, however (range 2, movement 2), the infantry would get ((4-2)-(2-1))=(2-1)=1 free attack-as the faster moving horse unit is able to close the distance quicker than his foot-bound counterpart (note, however, that this would all be abstracted)!! These two stats could help to REALLY seperate out units of varying tech levels-as you would assume that on average a unit from one age will possess a greater range than one from the age before it! This means that a single machine gunner unit (range 4, 'rof' 2), in the same tile as a stack of 6 speamen, is going to still have a reasonable chance of victory, as it would recieve ((4-0)-(1-1)*2)=EIGHT free attacks before the spearmen will have closed the distance. Of course, this is where the machine gunner will be in terrible trouble-as any surviving spearmen (yeah, RIGHT ) will have a reasonably good chance of overwhelming the lone machinegunner unit-as the machine-gunner would, in fact, probably have a relatively low Defense strength compared to, say, a standard infantry unit (think of a machine-gunner as a fairly 'static unit' and, therefore, easier to hit). Plus, because it might be fighting multiple opponents, then it will suffer a degredation to its defense strength and morale, thus making the likelihood of death and/or retreat much more likely! This was, out of neccessity, a very extreme example, but proves why in my system you would need to (a) upgrade your units as frequently as possible and (b) make sure you have a decent, MIXED stack of units!!

Anyway, hope that made sense!

Yours,
Aussie_Lurker.

 

[bkwrm79]

Very interesting.

I'm not sure I agree regarding simultaneous combat.

1. The advantages of mobility would be hugely weakened. It would be much harder to skirmish with mounted units, or even to make a sweeping advance through open terrain with armored units.

2. I like the distinction between attacker and defender. Under this model, the AS and DS stats become almost interchangeable, except that defense gets boosted by terrain (I think?). I don't think this improves either realism or gameplay.

3. I just don't think simultaneous combat would 'feel' like civ.

I really like some of Aussie_lurker's other ideas, particularly the range/rate of fire one.

 

[Aussie_Lurker]

OK Bkwrm79, consider this: Under the current system, if you attack with a unit, and it has a 78% chance to hit-and damage-a defender, then it also follows that the defender has a 22% chance of hitting-irrespective of the defenders attack strength (which, as I see it, should be a major determinant of a units ability to hit another).

In my system (simultaneous combat), a unit with a 78% chance to hit will be able to hit 78% of the time WITHOUT having to worry about the other 22% of the time-as those times simply represent a MISS. The chance of your unit being hit by an enemy will be entirely based on ITS AS/AS+your DS. See, each units DS still matters-very much-as every ATTACK denotes a counterattack. Also holding a position will be important under my system, as a unit which occupies a tile FIRST will get any defense bonuses from it, wheras a unit which enters it afterwards will get any penalties for attacking INTO that particular terrain type! In this regard, you would effectively still have a division between 'attacker' and 'defender'. Its only when both stacks arrive in a square at the same time that the distinction is lost. In addition, though, a unit which occupies a square can have a chance to set an 'ambush', by hiding within the tiles terrain (thus reducing its profile)-something which the current system doesn't allow for!

Also, in my system, the number and type of units in a stack is FAR more important (in terms of combined arms) as a stack that is archer-heavy, for instance, will be able to cut down a purely 'melee oriented' stack before it can even fight back!

Lastly, mobility would still be an issue. As I have stipulated, movement rates effect how quickly you can close with a ranged opponent. It also helps to catch and defeat slower opponents-particularly if you end up moving AFTER your opponent! Also, though, units with remaining movement points-after combat-would have the choice to continue moving up to their movement allowance (so, for instance, a tank unit uses 1 of its 2 MP's to move into the same square as an infantry unit-beating it-the tank now has the option to move into any square beyond, or BACK, EVEN if it ends up in a tile with another enemy stack-thus reflecting a tanks ability to make deep penetration attacks against multiple enemies!

As for the 'doesn't feel like civ' part-well I could use the same argument about culture and war weariness in the move from civ2 to civ3. That doesn't mean that they weren't fantastic additions to the civ series !

Yours,
Aussie_Lurker.

 

[Aussie_Lurker]

Consider the following 'in game' example: Its the movement phase of a turn. A stack consisting of 2 Longbowmen and 2 knights has just 'stumbled' into enemy territory. On their turn, said enemy moves a stack of 3 longbowmen, 2 light horsemen and 1 light medieval infantry to intercept them. After all movement has been determined, combat ensues. The 2 archers in stack A target the two light horsemen in Stack B, whilst said horsemen attempt to close the gap and engage the archers. Meanwhile, the two knights in stack A attempt to engage the 3 archers in stack B whilst, again, the archers counterattack. The LMI attempts to attack one of the knights also.

The LBM's (movement 1, range 2, armour 0, AS:10, DS:5 FP:1) in stack A launch an attack against the approaching horsemen (movement 3, range 0, armour 1, AS:8, DS: 12 FP:1) The horsemen are too fast, however, denying the LBM's their free attack (2-0)-(3-1)=(2-2)=0 free attacks. However, the LBM's have a 10/(10+12)=46% chance of hitting, wheras the light horsemen have a 8/(8+5)=61% chance of hitting. When the dust settles, the LBM's of stack A are eliminated, but not without killing one horseman, and damaging the other.

The LBM's of Stack B fare even worse as, even though they get a free attack against the knights (movement 2, range 0, armour 2, AS: 13, DS:7 FP:2), only one attack gets through the knights armour. Once they engage, though, both knights find themselves at a 2 to 1 disadvantage, giving them a -1 to their DS (and a 10% reduction to their morale). This means that the LBMs have a (10/10+6)=62.5% chance of hitting the knights, and the light infantryman (movement 1, range 0, armour 0, AS:7, DS:10 FP:1) has a 7/(7+6)=54% chance of hitting, with the knights having a (13/(13+5)=72% chance of hitting the LBM's. Unfortunately for stack B, the knights armour and firepower ultimately prove decisive-the LBM's are destroyed, at the cost of 1 knight!

Now the remaining survivors go head to head, with the last knight facing off against 2 light horsemen and 1 damaged infantryman. This gives the knight a 3-1 disadvantage, leading to a -2 to its DS, and a 20% loss to its morale!
This means that the horsemen have a (8/(8+5))=61.5% chance of hitting the knights, the light infantry having a (10/10+5)=67% chance of hitting, and the knight having a (13/13+12)=52% and (13/13+10)=56.5% chance of hitting respectively! The knights armour and firepower proves decisive once more, as it quickly kills off the infantry, then moves in to finish off the two horsemen-albiet at the cost of being redlined!

Having 1MP left at the end of combat, the player decides to move his unit back over the border, in order to ensure that it can heal! However, had his victory been more decisive, he/she could have easily moved the unit deeper into enemy territory-but not into a square containing another enemy stack/unit, as the knight lacks the blitz ability!
Of course, had the knights had the blitz ability, then they would NOT have suffered any DS loss due to being outnumbered, and the surviving knight also would have only suffered a -1 penalty to its DS!

Anyway, hope that made sense !

Yours,
Aussie_Lurker.

 

[Khan Quest]

I like it. Can you elaborate more on what affects morale and how morale affects combat? How about morale and troop skill (conscript/regular/veteran/elite).

 

[Khan Quest]

Nevermind. I found Sir_Schwick's post.

 

[Aussie_Lurker]

I'll try and answer it anyway, Khan.

First up, Morale is a factor of BOTH a units level AND its combat experience. That is a veteran unit will, obviously, have a greater morale than a conscript unit. Every X combat a unit fights, however, will also raise morale by Y%!

A unit loses morale (temporarily) after it is hit-even if said hit does no damage (consider it panic under fire). The amount of morale lost is dependant on how successful a hit is, but is increased if the hit ALSO caused damage!

Primarily, morale loss would reduce attack strength and firepower, as the unit becomes more devoted to survival than trying to dish out any serious damage.
Also, after all damage is inflicted in a combat 'pulse', a morale check is made to see if the unit retreats from combat. If the morale check result is more than double the current morale, however, the unit is considered routed. A routed unit is considered eliminated, and enhances the effect of any existing war weariness currently felt by your nation!

Yours,
Aussie_Lurker.

 

[GoodGame]

Firepower, Armor, Morale should be factors along with the base Attack/Defense stats. Also, it would allow the military techno-tweaking some people have been asking for, since with secondary stats, those stats could be augmented by research without fundamentally changing the nature of the unit (e.g. we could now get different models of tanks without resorting to UU's).

OTOH, if you want to keep it simple, but stretch out the scale of ATT/DEF, a simple solution would be logarithms. Add base ten for each epoch/era difference there is between units. E.g. a tank would be 1000x (+ATT/DEF differences) better than an ancient unit, but still be like 18 vs. 12 when attacking infantry (another industrial era unit). (actual % = 18 vs (18+12) ).

 

[GoodGame]

My own perspective on the meaning of the attributes:

ATT: Ability to engage the enemy in a decisive battle (underlying assumption is 'zone of control' (combination of mass and the range in which the unit can quickly apply it's firepower, and also the ability to detect/patrol objects within an area) and 'mobility'--overall speed of the unit, as well as mobility in particular terrain types (e.g. cities, mountains, coast, etc..).



DEF: Basically the compaion to ATT, the ability to avoid decisive engagement. Basically from the 'mobility' (same), and 'evasiveness' (ability to camoflauge its existence by some means, or against some means of detection).

Basically ATT/DEF determines if a round of combat actually results in a Firepower/Armor exchange, or not. And it can also decide if the combat ends early with a withdrawal, and even if the withdrawal is intercepted by a pursuit (feasibly combat could now migrate across different tiles, if desired in the model).

Firepower and Armor are then just the technological ends of physically fighting---weapon (type) vs. armor (type). Not that there'd be types, but the real world would be averaged across different types to give each unit one Firepower/Armor stat (but then again, it could be broadened to a a couple different meaningful stats).

Basically Firepower vs. Armor determines if/how much damage is done to the Hit Points---mainly as a direct numerical ratio, but it could also be a matrix if subtypes of Firepower/Armor are created.

Morale would be great as the Napoleonic concept ("Morale is to physical as 3:1"). Morale could randomally affect Hit Points (bonus/deletion), and primarily ATT/DEF (willingness to fight/run away/even do your job). Morale would be great for introducing minor leadership units, as another significant factor, and could probably be tied to average Civilization Happiness, War Weariness, and gov't type.