Attack and Defense VS 1 Number

[V. Soma]

I am for units to have defense and attack value separately:
(I believe it was like that in civ3)

If you move to a tile and thus enter a fight, you attack.
If you stay on a tile and thus enter a fight, you defend.
The two is different, and a unit should have differentiated abilities for the two.

 

[snowlyon]

I am for units to have defense and attack value separately:
(I believe it was like that in civ3)

If you move to a tile and thus enter a fight, you attack.
If you stay on a tile and thus enter a fight, you defend.
The two is different, and a unit should have differentiated abilities for the two.

agreed.~

 

[Shurdus]

I am for units to have defense and attack value separately:
(I believe it was like that in civ3)

If you move to a tile and thus enter a fight, you attack.
If you stay on a tile and thus enter a fight, you defend.
The two is different, and a unit should have differentiated abilities for the two.

Some units do indeed get bonuses in civ4, like archers having a bonus when defending a city. Longbows get bonusses while defending on a hill. One single number + bonusses can account for this too, right? Why would we need two numbers?

By adding bonusses rather than static defensive and attacking values it becomes possible to make distinctions in unit types. For example, spears may be better in defending vs horses than axes are, but why would they be better than axes against archers? With the civ4 system there is the distincion like this. with the two number system, the spears would be better defenders just because they get a higher defense number. It is a less sophisticated system imo.

 

[Ahriman]

I'm unsure which side of this I fall on, but having different attack/defense values and having specific bonuses vs different types are not mutually exclusive.

Its entirely possible to have a 3/4 unit that is +25% vs mounted units.

I think I lean towards a single value, given 1upt, because of the relatively low value of primarily defensive units. 1upt also has *much* more scope for making units have interesting specialized roles. Specialists are finally useful once you remove "best unit defends the stack".

 

[Thyrwyn]

I think I am with Ahriman on this one. Especially given that supposedly units in Civ V will always defend at full strength. I would rather use bonuses (for attacking, defending, terrain, etc...), the ability (or not) to take advantage of defensive terrain, etc...

Why is a spearman or a paratrooper inherently better at one or the other? Most units should not be inherently better at one than other all else being equal; against certain units or in certain situations, yes, but not as a general rule.

 

[Schuesseled]

this is the way i think it should be done:

Attack applies to causing damage and defence to resisting damage, units also have health values. In simplest terms units take damage to thier health from combat equal to defence minus attack. The defence and attack values for units do not change when injured so a tank will still be a tank if damaged just with less health.

Melee Attack/Defence - Applies to the capability of commencing and resisting melee attacks.

Units like a swordsman have high melee attack and defence values, where as archers have lower values.

Missile Attack/Defence - Applies to the capabilities of commencing and resisting missile attacks (arrows, bullets)

Melee orientated units such as swordsmen have no missile attack and some missile defence, depending on the units. Archer units have high missile attack but only a little missile defence, as they don't have much protective gear. Gunpowder units also use this missile rating however they will often get a modifier bonus against old tech like melee units.

Explosive Attack/Defence - Applies to the capabilities of commencing and resisiting explosive attacks. (Grenades, Shells, Rockets)

Mostly only modern units use explosive attacks, and most old tech units will have very little resistance to explosive attacks.

Some units will only be able to perform one type of attack, others will have multiple attacks they will use simultaenous in battle. For example, an infantry unit will have access to grenades, bayonets and thier rifle and will use all three in combat. For example when a swordsman unit attacks the infantry unit, the swordsman are only able to rush at the infantry and engage in melee combat. But the infantry are more verstaile, as the swordsman approach the have the oppurtunity to use thier rifles and grenades, and possibly again later in the fight. (this all represented graphically for the player as well as hard-coded ) which deal high damage to the swordsman, then they also engage in melee combat during the battle. Swordsmen are more resistant to that type of attack and so dont take too much damage from the periods of melee fighting, in fact they inflict more damage on the infantry than they take from melee however the infantry have more health. In the end the swordsman gets its butt kicked but you can see how these values can play into the fight and make it more realistic and enable the graphical represntation of the battle to be more intense and accurate of what is going on.

As an example of unit statistics:

Swordsman

Health 20
Melee Attack 12
Melee Defence 10
Missile Defence 6
Explosive Defence 2

Infantryman

Health 80
Melee Attack 10
Melee Defence 8
Missile Attack 24
Missile Defence 12
Explosive Attack 16
Explosive Defence 6

Tank

Health: 160
Melee Defence: 30
Missile Attack: 45
Missile Defence: 28
Explosive Attack: 75
Explosive Defence 40

unit promotions:

Offensive Maneuver- Increases all attack values by a certain percentage, further upgrades increase further as well.

Defensive Maneuver - Increases all defence values, increasing further for multiple upgrades.

Strength - Increases unit's health, further upgrades increase more as well.

Stamina - Increase movement range (requires Offensive and Defensive Maneuver 1)

Blitz - Allows for multiple attacks (requires Offensive Manuever 3)

Anti-Melee/Archery/Mounted - Percentage bonus against unit type (requires Offensive or Defensive Maneuver 1)

Anti-Gunpowder/Armour/Air - Percentage bonus against unit type (requires Offensive or Defensive Maneuver 3 (and relevant techs))

Withdrawl - Retreat out of combat if being overwhelmed, and move away, % chance (requires Stamina)

City Attack - Increase attack values when assualting city or fort (requires Offensive Maneuvers 1)

Units should also have a percentage bonus added to thier attack values when engaging in combat with technologically inferior units.

So for example a Infantryman should get a 100% bonus swordsman and the like, and a 50% bonus versus Musketman and Rifleman. Marines/modern infantry should recieve the same plus 25% versus infantry. And so on and so forth.

Making it harder for technologically inferior units to get lucky and kill a superior unit, they should only be able to do that with a much larger amount of troops.

 

[Ahriman]

others will have multiple attack options maybe using more than one in one battle

So you have to manually select what attack mode to use in each fight? a la Battle for Wesnoth?
That sounds like too much micromanagement.

Its much simpler if there is just normal combat with a single strength number, and then bombardment for artillery, siege units and archers.

I also don't see a need for separate defensive values vs melee, ranged attacks, etc.
This kind of thing is handled much better using unit classes like Civ4, where axemen are +X% vs melee units, or pikemen that are +Y% vs mounted units.

 

[Schuesseled]

So you have to manually select what attack mode to use in each fight? a la Battle for Wesnoth?
That sounds like too much micromanagement.

Its much simpler if there is just normal combat with a single strength number, and then bombardment for artillery, siege units and archers.

I also don't see a need for separate defensive values vs melee, ranged attacks, etc.
This kind of thing is handled much better using unit classes like Civ4, where axemen are +X% vs melee units, or pikemen that are +Y% vs mounted units.

sorry ill edit that, i meant all values used in the one attack, representing the different weapons and skills troops can employ in battle. Swordsman just have meelee attack but a infantryman can shoot/stab or nade. (and will do them all)

 

[Schuesseled]

basically it allows for units to be used properly in battle, tanks aren't gonna get stabbed to death they have to be blown up. Infantrymen can be injured by swords and arrows but its harder, so much easier to send in a tank or another infantryman. etc.

 

[Ahriman]

sorry ill edit that, i meant all values used in the one attack, representing the different weapons and skills troops can employ in battle.

This just makes it much more difficult to evaluate how one unit will do vs another unit, because you have to look at all its different attack types, vs all the differen defense types of the other unit, and then again vice versa. Much easier to just compare a single strength number for each, and then any modifiers vs particular unit types.

tanks aren't gonna get stabbed to death they have to be blown up

You model this by giving them a high unit strength, and only having particular other units like AT troopers and AT guns and gunships, having bonuses vs tanks.

Infantrymen can be injured by swords and arrows but its harder

You model this by giving modern infantry a much higher strength value than medieval swordsmen/bowmen units.

So we can get all of the scope for unit specialization without needing to explicitly model the level of detail you have in mind.

No need for multiple strength and defense values.

Which is easier:

Tank strength 24 vs rocket trooper strength 18, +50% vs armored units.

Or:
AT infantry
Health 80
Melee Attack 10
Melee Defence 15
Missile Attack 24
Missile Defence 18
Explosive Attack 35
Explosive Defence 25

Tank
Health: 160
Melee Defence: 30
Missile Attack: 45
Missile Defence: 28
Explosive Attack: 75
Explosive Defence 40

Both achieve the same design goal.

 

[GhostWriter16]

No. Just, no.
If you want realistic combat, go look for a wargame, not Civ. And pick one that has a combat engine designed to model a particular era of warfare, not everything from the clubs to stealth bombers.


All get trumped by:
Is it simple, transparent, accessible?
Is it fun?

I'd say both need to be balanced. I, for instance, think the tiny chance of a spearman beating a tank is ok, as it is less than 1%, however, a longbowman garrisoned in a city having a 33% chance (This was based on a specific situation) to kill a tank is unnacceptable.

I don't want it to be a wargame obviously, but I like the steps that are being taken.

 

[Ahriman]

however, a longbowman garrisoned in a city having a 33% chance (This was based on a specific situation) to kill a tank is unnacceptable.

I don't want it to be a wargame obviously, but I like the steps that are being taken.

The steps that *are* being taken?
What steps are those? We have no idea.

Certainly, without having units sitting in cities with big defensive bonuses, then there will be much less scope for large stat boosts for lower tech units.

So you're not likely to have longbowmen beating tanks because you're fighting out in the field.

But there's really not much they've said that indicates that tanks in the field will do better vs longbowmen in civ5 than they would in civ4.

 

[Schuesseled]

Actually due to how Civ combat works, a much stronger unit say a tank versus a spearman even though the spear man only has a minute chance of winning, it has a much better chance of damaging the tank. Thus a completely insignificant spear man can be used as a suicidal meat stick, not to win against the tank but simply to damage it. This isn't 100% gaurenteed but it usually works, so if you throw enough units which didnt have a chance of winning into the fight first, pawns in chess if you will, you soften the defences and your stronger units say a marine who normally would die against a tank will now win. I believe Schuesseled's suggestion (This is his brother btw - I don't have my own account) elminates or atleast reduces this "bug" in the Civ 4 combat system, I will quote the bit I am talking about.

"In simplest terms units take damage to thier health from combat equal to defence minus attack. The defence and attack values for units do not change when injured so a tank will still be a tank if damaged just with less health."

This will mean a swordsman as given in the example will attack with 12 melee attack and the tank will defend with 30 melee defence, 12 - 30 = 0 damage. This thus removes the "bug" whereby a completely insignificant unit such as a man with a sword can kill a tank with 10 inch steel armour.

Also another benefit of the health system, all of a units health points might not be taken with one attack, this means combat between units on the frontline will last longer, as well only having 1 unit per hex. So what we will see is a "font line" like we had in the trenches of the world wars, with the line holding whilst enemies attack it, rather than a stronger enemy straight away killing the weaker one, they have to duke it out. This will give the weaker player time to perhaps re-inforce the front line with more troops to help the fight.

The point i'm trying to make is to ahriman here, whilst the end result would be the same with the old combat system and this proposed combat system, its how you get there thats important. Currently they take a shortcut, sure it is simpler to look at, but it is also "bugged" whereby somethings happen that really shouldnt happen, such as suicidal attacks of pointless units to get a win. The old system caters for this but the proposed system does not. A swordsman attacking a tank will just bounce off liek a fly against a windshield.
The fact that it is more complicated helps it be more accurate and a better combat system for gameplay and realism purposes. It doesn't need to be simple, the human player doesn't have to do any of the maths, the maths are displayed as they are currently along with the result of the fight, instead of in % to win, with this system it will know that it wins and does damage. Ofcourse if you didnt want all this certainty you could throw soem randomness into the mix as well, but this proposed system is still better than the old one.

12agnar0k.

 

[tom2050]

The steps that are being taken are:

A 99% dead longbowman defender will still defend at 100% strength. And a half-dead swordsman will attack at half strength.

Attacking will be either very slow and painful to go through; or the only strategy will be to simultaneous attack a unit with AMUAP (as many units as possible) or Blobs of death; similar to Civ 3 and 4. There is no way to get rid of this aspect the game obviously.

 

[Ahriman]

Actually due to how Civ combat works, a much stronger unit say a tank versus a spearman even though the spear man only has a minute chance of winning, it has a much better chance of damaging the tank.

But you can still do this with a single strength statistic (without neednig melee, ranged, explosive, etc.).

Imagine that relative strength determines the likelihood of winning a combat round, and that winning a combat round does strength/3 damage.
Imagine that health/hit points fall as damage is taken, but that strength remains unchanged.

Then imagine, a strength 4 spearman with 6 hit points, and a strength 24 tank with 30 hit points.

The probability of the tank winning the first round is 24/(24+4) = 90%.
If it wins, the spearman is dead (takes 24/3=8 damage). If it loses, it takes 4/3 damage, leaving it with 28.67 health.
Second round is exactly the same as the first, except for hit point total.

So, in the probability of the tank taking at least 4 damage (ie barely 1/8 of its health) is the probability of it losing 3 times in a row - ie 0.1*0.1*0.1, or 1 in 1000.

The expected damage of the spearman is roughly ~= (4/3)*0.1 + 2*4/3*(0.1)^2 + 4*4/3*(0.1)^3+ .... ~= 0.133/0.8 = 0.2

So it would take, very roughly on average, 30/0.2 = 150 spearmen to kill the tank.

The problem is solved, without any need for having "armor-piercing" weapons or anythnig similar.

Also another benefit of the health system

We can have a health system (which I strongly suspect we will have) without needing multiple types of unit strengths.

A 99% dead longbowman defender will still defend at 100% strength. And a half-dead swordsman will attack at half strength.

Nope.

We don't know what "defend at 100% strength" means.

Attacking will be either very slow and painful to go through; or the only strategy will be to simultaneous attack a unit with AMUAP (as many units as possible)

Yes, you want to attack a unit with as many units as possible... but positioning stops you from doing that, because you can only have 2 units attacking a unit in a straight front line.

And we don't know that injured units won't attack at full strength.

 

[tom2050]

Nope.

We don't know what "defend at 100% strength" means.

How can you say Nope when you say no one knows? The only thing defend at 100% strength can mean, is that the units defensive strength value (2, 4, 10, w/e) will not change. Chances are the units 'health value' would be diminished according to it's damage level, making it still less likely to survive.

Whereas since a units Strength value is affected when attacking, if the strength is at 10, and a unit is 50% damaged, is strength value would be halved to 5.

Do we know for sure, of course not. But this is a plausible assumption so far according to how combat acted in previous iterations, and according to the interview.

Yes, you want to attack a unit with as many units as possible... but positioning stops you from doing that, because you can only have 2 units attacking a unit in a straight front line.

And we don't know that injured units won't attack at full strength.

You can still attack with more units than that. If there is 1 defender on a hill, that hex is surrounded by 6 tiles. Units can 'switch places' therefore you can use all hexes behind the front line as well (which is quite a few). So in 1 or 2 turns, you could amass and attack with quite a few units.

It is more similar to stacks than you realize, only that you have to bring them down in rows and lines instead. A single unit obviously will have no chance of surviving something as such. It will be interesting though using positioning against evenly matched armies, only if the AI can use it's forces effectively though; and no major exploitative strategies exist.

 

[Ahriman]

How can you say Nope when you say no one knows?

Because this is the result of a 2-step language translation from a reviewer who did not really sound like they understood what they were talking about, and the meaning is very unclear.
I say "nope" because you were implying something as fact, which we do not know to be fact.
This is how disinformation is spread, by people reading your post and thinking you knew what you were talknig about.
I just take on the noble and heroic task of setting the record straight

The only thing defend at 100% strength can mean, is that the units defensive strength value (2, 4, 10, w/e) will not change.

We don't know how the combat system works. We don't know that there *is* a "defensive strength value" or if there was what it would mean for the combat system.
We don't know whether or not it would equally apply to units when attacking.

Whereas since a units Strength value is affected when attacking, if the strength is at 10, and a unit is 50% damaged, is strength value would be halved to 5.

You assumed this, from over-parsing a very vague statement from a low quality source. You don't know it.

It might be true, but we can't be sure. Its a plausible interpretation, but it is not the only plausible interpretation.

For example, a single strength value that is unaffected by damage combined with a hit point value is also a plausible interpretation.

If there is 1 defender on a hill, that hex is surrounded by 6 tiles. Units can 'switch places' therefore you can use all hexes behind the front line as well (which is quite a few). So in 1 or 2 turns, you could amass and attack with quite a few units.

Not necessarily; it is perfectly possible (likely even) that a switch places command requires you to have movement points spare to accomplish the switch.
So you can't attack the hill tile, using up all your movement points, and then switch places with another unit.

The most likely interpretation of a "switch" command is that it lets a unit on tile A move to tile B while the unit in B simultaneously moves to tile A. Each consuming the appropriate movement points for the move.

Again, we just don't know.

I seriously *hope* we can't use switching to concentrate as much power as we like on a single tile, because that would eliminate much of the purpose of 1upt and positioning.
I also hope that we can't switch places with a tile that has no movement left, or I can imagine all kinds of movement exploits (eg using "chaining" to rapidly move a unit from one end of a line to the other in a single turn).

 

[tom2050]

Since most units will have 2 move points, switching may require both (or all) move points to do a switch, in which case the units in the back of the line would have to wait 1 or 2 turns after the initial attack to switch and have their turn.

 

[Schuesseled]

The steps that are being taken are:

A 99% dead longbowman defender will still defend at 100% strength. And a half-dead swordsman will attack at half strength.

Attacking will be either very slow and painful to go through; or the only strategy will be to simultaneous attack a unit with AMUAP (as many units as possible) or Blobs of death; similar to Civ 3 and 4. There is no way to get rid of this aspect the game obviously.

As ahriman said the review was translated twice and wasn't very coherent, plus defend at fulll strength could apply to multiple attack and defence values like in some civs or one value for attack and defence as in civ 4, or the reviewer could have been drooling on his keyboard as he wrote that. We cant tell.

But you can still do this with a single strength statistic (without neednig melee, ranged, explosive, etc.).

Imagine that relative strength determines the likelihood of winning a combat round, and that winning a combat round does strength/3 damage.
Imagine that health/hit points fall as damage is taken, but that strength remains unchanged.

Then imagine, a strength 4 spearman with 6 hit points, and a strength 24 tank with 30 hit points.

The probability of the tank winning the first round is 24/(24+4) = 90%.
If it wins, the spearman is dead (takes 24/3=8 damage). If it loses, it takes 4/3 damage, leaving it with 28.67 health.
Second round is exactly the same as the first, except for hit point total.

So, in the probability of the tank taking at least 4 damage (ie barely 1/8 of its health) is the probability of it losing 3 times in a row - ie 0.1*0.1*0.1, or 1 in 1000.

The expected damage of the spearman is roughly ~= (4/3)*0.1 + 2*4/3*(0.1)^2 + 4*4/3*(0.1)^3+ .... ~= 0.133/0.8 = 0.2

So it would take, very roughly on average, 30/0.2 = 150 spearmen to kill the tank.

The problem is solved, without any need for having "armor-piercing" weapons or anythnig similar.

Ah see the thing your forgetting there is that a spearman doesn't need to win a battle to damage a tank it just has to take part. Every combat will result in some damage to a tank. And so even if the damage is reduced it still has the desired effect.

With my system (it's schuesseled again) melee units have no capability of inflicting damage to a tank, or very almost no chance if considering randomn chance factored in.

This just makes it much more difficult to evaluate how one unit will do vs another unit, because you have to look at all its different attack types, vs all the differen defense types of the other unit, and then again vice versa. Much easier to just compare a single strength number for each, and then any modifiers vs particular unit types.



You model this by giving them a high unit strength, and only having particular other units like AT troopers and AT guns and gunships, having bonuses vs tanks.


You model this by giving modern infantry a much higher strength value than medieval swordsmen/bowmen units.

So we can get all of the scope for unit specialization without needing to explicitly model the level of detail you have in mind.

No need for multiple strength and defense values.

Which is easier:

Tank strength 24 vs rocket trooper strength 18, +50% vs armored units.

Or:
AT infantry
Health 80
Melee Attack 10
Melee Defence 15
Missile Attack 24
Missile Defence 18
Explosive Attack 35
Explosive Defence 25

Tank
Health: 160
Melee Defence: 30
Missile Attack: 45
Missile Defence: 28
Explosive Attack: 75
Explosive Defence 40

Both achieve the same design goal.

the bold section is the problem i have with civ 4.

Tank vs Infantryman, not an AT guy just normal infantryman, from worldwar 2, armed with a bayonet, rifle and potato masher. They didn't even anti-tank weps for infantry for a little bit hence it coming later as an AT man.

Unless an infantryman were to throw a grenade into the cupola of a tank, he literally can't damage it. (let's assume all civ tanks have thier lids shut in battle, and thier co's arent dumb.) Nothing he can do will stop him from dying other than hiding in a forest and hope he doesn't get it by a stray bullet.

And yet in civ 4 it only takes one infantryman in a 50% defence tile to destroy a tank, hell i've even had riflemen do it. Having health values and damage reduction only helps so much, if tanks are designed to be almost completley immune to basically anything other than tanks and AT men, the game would be a lot better.

It makes more sense to actually code this into battles with damage types and resistances than to put every armoured unit into a god mode statis. (where there immune to damage from unit types.) this would be the only way to make tanks better if they go with the one strength statistic, where as in mine they wouldn't be completely immune to damage from other units to make it fairer jsut makes it a damn sight harder.

They way i'd do it is like this:

Infantry
Health 80
Melee Attack 10
Melee Defence 8
Missile Attack 24
Missile Defence 12
Explosive Attack 16
Explosive Defence 12

Tank
Health: 160
Melee Defence: 30
Missile Attack: 45
Missile Defence: 28
Explosive Attack: 75
Explosive Defence 40

AT Infantryman
Health 80
Melee Attack 6
Melee Defence 8
Missile Attack 14
Missile Defence 8
Explosive Attack 18
Explosive Defence 4
(Bonus versus tank 300%)

Doing it this way as you can see makes an AT infantry vulnerable to a regular infantry, it has higher explosive damage but it leaves itself vulenerable to both missile and explosive attacks. (this is because its hard to shoot a rocket launcher from a foxhole/dugout you have to stand up exposing yourself.) But when facing armoured units it basically triples all of its values giving it enough explosive power to damage, and its training comes in handy preventing most of the damage a tank would inflict on regular troops. This allows tanks to be much much better than regular troops and allows you to counter them without having to introduce another impossible to kill unit.

I agree this can be done with one strength values, but my way is more realisitc and sensible. If single values are used the scaling should still be like what i described to make it a good game.

such as:

Tank: 60 str , Inf: 18 str, AT inf: 15 str (x6 vs tanks) or whatever.

Main point is that tanks should not be able to be killed by a rifleman in a bush.

 

[Ahriman]

Unless an infantryman were to throw a grenade into the cupola of a tank, he literally can't damage it. (let's assume all civ tanks have thier lids shut in battle, and thier co's arent dumb.) Nothing he can do will stop him from dying other than hiding in a forest and hope he doesn't get it by a stray bullet.

Unless of course, like in reality, that Infantry unit actually represents an Infantry Division, complete with anti-tank guns, bazookas/panzershreks, panzerfausts and the like.

Why would you think that an Infantry unit would be armed onyl with rifles and grenades?

Even an infantry *company* IRL has a bunch of heavy weapons.

An Anti-Tank unit is just a division more heavily focused on AT weapons; it has more AT guns, fewer machine guns and snipers and mortars.

And yet in civ 4 it only takes one infantryman in a 50% defence tile to destroy a tank,

There are plenty of ww2 examples of infantry heavily in heavily fortified positions, with their AT-guns and other weapons, annihilating incoming armor.