Archived

This topic is now archived and is closed to further replies.

Sandman

Realistic Damage Models & Gameplay (RTS)

Recommended Posts

Counterstrike showed us that a realistic damage model works well in FPS - but what about RTS? First of all, lets scrap all this hit point crap. Instead, we''ll have two different types of damage: normal damage and critical damage. Normal damage represents damage to a units systems which, while not being fatal, will severely penalize it in some way - maybe a tank loses the ability to rotate its turret, or loses the turret altogether, or perhaps its gearbox is damaged causing it to move at half speed. None of this damage destroys the unit, but it can cripple it to the point of being useless. Critical damage represents a direct hit to some vital and highly volatile part of the unit, eg, the fuel tank, ammo stores, main reactor etc. A unit taking a critical hit is destryed instantly. The chances of doing critical damage will depend on a lot of factors - the weapon, the accuracy of the gunner (his ability to target vital spots and weak points) as well as the design of the unit itself. Taking lots of normal damage may also increase the chances of a critical hit. What does this mean for the gameplay? I think it makes an interesting tradeoff for otherwise dominant units - for example: Tank B is the second most powerful unit in the game. It is armed with a super deth laser which has a 100% chance of inflicting a critical hit on any unit in the game, except for tank A, against which it has a 25% chance. It costs 100 space macaroons to build. Tank A is the most powerful unit in the game. It is armed with the deadly Mega Ultra Doom Ray, which has a 100% chance of inflicting a critical on any target, and a 90% chance of inflicting a critical on units near its target. Tank A costs 200 space macaroons to build. Who wins in a fight of two tank B vs one tank A? It is difficult to say.... You could kill both of my tank B''s with a single shot in some circumstances, but my tank B''s could attack your tank A from two different directions - you would certainly kill one, but I might get several shots off before you can take out the other, almost certainly crippling you if not destroying you completely. Your thoughts?

Share this post


Link to post
Share on other sites
I like the idea in principle (very much actually) but am a little concerned about gameplay. On the face of it it sounds like one on one engagements would be over quickly, which is actually realistic (much like many Counterstrike firefights).

But having seen the first Dark Reign play out like this, I''m afraid gameplay would become VERY frenetic unless you tweaked some other factor, like move speed.

If your mechanism actually does allow for longer battles, then one of the things you lose with hitpoints is a progress indicator. Should you retreat yet? Should you add more units to the fight.

Another thing is the UI management element HP adds: Either you have to memorize the strengths of all units, or indicate it another way (unit size?) because you won''t have a number or bar helping you.

I still like the idea, though. I would go with some kind of strength indicator and a H#LL of a lot of defensive (berms) and offensive choices (manuevers) that affect whether or not a battle will quickly be over. This gives you more strategy, and since the units might die quickly, immediate feedback.

How about adding facing damage, like I''m thinking about it that space strategy post?

--------------------
Just waiting for the mothership...

Share this post


Link to post
Share on other sites
I like your idea, more interesting damage models would make tactical combat and operational strategy more interesting. I was a big fan of TA with the collateral damage and field resource angles.

Of course, what we really need is a game which uses your space macaroons as currency. Then you could choose to feed troops with them, or spend em for more units, or maybe use em as ammo.

Share this post


Link to post
Share on other sites
The thing you would need to do is generalize what Hit points actually are. If you used them internally as a measure of How alive a unit is (taking into account all factors, etc), and letting the unit "know" this value, you can build a system of "unit thought". They will know when they are winning, when they are losing, when to run, when to call for reinforcements. it would be the players job to send the reinforcements, but if the unit decides it''s time to run, they run, etc. Don''t let the player see the "HP" value in the game, so the player''s only judge is what the unit is yelling at him/her/it. Otherwise, I think this is a great idea. Depending on the Ammo the attacker is using, the armor on the defender, etc, you could use this model to great extent. In fact, I think I will >=).

Z.

Share this post


Link to post
Share on other sites
Sounds fun. However, I know that in that game I''d be the unfortunate guy who''s entire squad of Type A tanks gets destroyed by Cheap Throwaway infantry because of a rash of 1% critical chances coming through.

Also in this model, cumulative chances of criticals based on damage seems to me to be a good way to go. Armor falling off, or unit fatigue leaving important systems unguarded.

Share this post


Link to post
Share on other sites
quote:

I like the idea in principle (very much actually) but am a little concerned about gameplay. On the face of it it sounds like one on one engagements would be over quickly, which is actually realistic (much like many Counterstrike firefights).

But having seen the first Dark Reign play out like this, I'm afraid gameplay would become VERY frenetic unless you tweaked some other factor, like move speed.



Fair point. Actually it could work both ways. If like in the example, critcal hits were very likely, engagments would be short and deadly. Entire forces could be wiped out before you have time to respond (you should have set up better defences shouldnt you?)
If the critical hit probability were very low, battles could drag on for ages, with units gradually getting more and more bits blown off them until the front line is cluttered with crippled tanks and incapacitated soldiers (probably resulting in a stalemate) I am aiming for something somewhere in between I think, set the critical level high enough to make one hit kills a serious possibility, but low enough to allow the odd unit to survive against all odds, albeit getting seriously beaten up in the process....

Another option is to tie the critical hit probability to a user option. They could choose "Invincible" mode, in which one shot kills are all but impossible, or "Sudden Death" mode, in which one shot kills are very likely. The players could then set it up how they like it.

quote:

If your mechanism actually does allow for longer battles, then one of the things you lose with hitpoints is a progress indicator. Should you retreat yet? Should you add more units to the fight.



Well, I was planning on allowing the player to set the morale of his own units (rather than any forced morale system). Unless they are set on suicidal mode, they will retreat when they are taking a beating automatically, how much of a beating they will take depends on how low their morale is set. As for conveying damage to the player, ideally, 'normal' damage would be reflected by the model, but a neat possibility would be something like in Starcraft - you get the wireframe image with parts in green/yellow/red depending on how knackered they are... only in this the red bits actually represent which parts of the unit are damaged . Critical damage is represented by replacing the unit with smoldering heap of junk

quote:

I still like the idea, though. I would go with some kind of strength indicator and a H#LL of a lot of defensive (berms) and offensive choices (manuevers) that affect whether or not a battle will quickly be over. This gives you more strategy, and since the units might die quickly, immediate feedback.



Duly noted. I have a fairly detailed order structure, which should allow a reasonable amount of control, *hopefully* without cluttering the UI too much (I posted a basic outline of it a while back). Players will have to learn to trade off between using small groups of units (and becoming vulnerable to having his forces wiped out before he can react by a well planned ambush) and having huge groups that can be spotted from a mile away and shelled into oblivion. I can also predict that stealthy units (eg infantry) will become quite important. In order to wipe out 5 soldiers, you need 5 shots (or a shell that can take the whole lot out in one)... and it only takes one to plant a charge on a tank and blow it to pieces... He just has to get close enough...

quote:

Of course, what we really need is a game which uses your space macaroons as currency. Then you could choose to feed troops with them, or spend em for more units, or maybe use em as ammo.



I have suspected this might be the case for some time. Space macaroons are underused in todays game, despite the fact that they are one of the most versatile units of currency.

Zaei: I think we are talking crossed purposes here - I am not using any hitpoints at all... Or am I misunderstanding your post?


Edited by - Sandman on July 26, 2001 9:00:53 PM

Share this post


Link to post
Share on other sites
I''ve been annoyed for quite some time by the massive innacuracies of a simple hit point system. I like your idea, but I must admit something similar to it has been floating around in my head for a while, although I haven''t given it much thought.

One thing that bugged me in many games (like C&C) was the abiltity to take out heavy tanks and even buildings with only a handful of infantry. It doesn''t take a moron to realize that a guy with a small-caliber firearm isn''t going to have snowball''s chance in hell of destroying a tank, even if it doesn''t shoot back.

Another thing I think would be neat to add:
Suppose a unit is heavily damaged, but manages to make it back for repairs. Why not work in the possibility of repairman error?
Suppose it''s just a quick fix, and once it makes it back into the fray, a light hit cripples it?

Share this post


Link to post
Share on other sites
Way to get a clue on how the battle is going without hitpoints?

Well, have your troops give feedback. If your forces are small (say 10 units) they might each individually report back to you (the commander). If your forces are large (say 1000 units) individual units would report to your sergeants who would then report to you.

''Sir, division 10 is taking heavy losses.''

''Commander, platoon 5 is overpowering the enemy at location X. Soon we will have control of the area.''

Woohoo... I''m on day 4 on my C++ in 21 days course. %Another two weeks and I''ll be a master programmer!%

Share this post


Link to post
Share on other sites
I think this boils down to a simplicity vs. complexity thing.

I''ve never been a big fan of the "ablative" style of damage tracking. Just because a weapon hits a certain location on a unit doesn''t mean that that point has taken permanent damage.

I think the key factors here are a penetration class, and a damage class. For example, you could have a weapon with very high penetration, but a very small damage class for a ultra velocity round that was very small in diameter. Conversely, you could have a huge damage class weapon with very low penetration such as a flame thrower (maybe a plasma thrower would be high in both).

A long time ago, FASA put out a game called Centurion that was very innovative in one respect. Weapons had a damage template, and the units had a matrix of boxes representing its armor. Lasers for example did one column of damage straight down, whereas a HEAT round would have a conical shape. I thought this was very fascinating, as depending on the damage model of the weapon, you inflicted a differing amount of damage.

However, I still prefer the penetration vs. damage class thing. AS an example, if you can''t beat the armor rating of the defending unit...you do no (or extremely little)damage....period. That''s why certain units are rightfully feared, you simply can''t take them out without the right kinds of weapons. Infantry should not be able to take out a tank unless they have some form of AT weapon, otherwise they''re up the creek without a paddle.

But, when you get down to it, that''s ALOT of data tracking. And I''m not even getting into the actual damage modeling if a hit does breach the armor and damage the unit. I think that for a turn-based game, it could handle this attention to detail, but I don''t know if it will work for an RTS. The trouble with RTS''s is that....they are just too fast. I don''t think the player truly has the time (as he would in the real world) to react to certain events and make informed choices. While some would say this is realistic, RTS are simply too fast for my taste. With the amount of micro-management necessary, if youadded this level of detail, it will simply get lost along the way.

Share this post


Link to post
Share on other sites
Sounds cool to me. The game would end up being a lot more tatical (realistic damage tends to do that) then your average RTS. It worked for FPS, and it can probably work well with any genre.

Share this post


Link to post
Share on other sites
quote:

However, I still prefer the penetration vs. damage class thing. AS an example, if you can''t beat the armor rating of the defending unit...you do no (or extremely little)damage....period. That''s why certain units are rightfully feared, you simply can''t take them out without the right kinds of weapons. Infantry should not be able to take out a tank unless they have some form of AT weapon, otherwise they''re up the creek without a paddle.



I didnt get into details on how the damage would be calculated, but I was definitely thinking somewhere along those lines. You can throw anti-personnel grenades at a heavy tank all day, you won''t even scratch it. Conversely, a hit from a rail gun will easily penetrate the armour but the damage will be very localized - unless you hit a vital spot it wont do very much. It may take several hits from a railgun before a tank even takes normal damage... Most anti tank weapons would need to be high penetration and high damage, but there is nothing to stop an expert sniper with a railgun taking your tank down in one...
As for infantry not being able to take out tanks, I want to be careful about going down that route.... My logic is that while standard infantry weapons would be ineffective against an armoured tank, if they can get up close to it they can plant explosives, cut wires, throw grenades down the hatch etc.... infantry will nearly always beat the tank at close quarters.
quote:

Another thing I think would be neat to add:
Suppose a unit is heavily damaged, but manages to make it back for repairs. Why not work in the possibility of repairman error?
Suppose it''s just a quick fix, and once it makes it back into the fray, a light hit cripples it?



Not sure about that... I think that might undermine the value of preserving units - if repaired units are less reliable than brand spanking new ones, the player will be less inclined to let his damaged units retreat, in favour of letting them do as much damage as possible before they gp critical. I think repairs made back at the base would always be successful. On the other hand, what about repairs made on the field? Damaged units could be ordered to self repair (tank crews would have some engineering units, infantry would have some first aid training etc. The unit would still be penalised, but at least it would be vaguely functional...

quote:

But, when you get down to it, that''s ALOT of data tracking. And I''m not even getting into the actual damage modeling if a hit does breach the armor and damage the unit. I think that for a turn-based game, it could handle this attention to detail, but I don''t know if it will work for an RTS. The trouble with RTS''s is that....they are just too fast. I don''t think the player truly has the time (as he would in the real world) to react to certain events and make informed choices. While some would say this is realistic, RTS are simply too fast for my taste. With the amount of micro-management necessary, if you added this level of detail, it will simply get lost along the way.



I dont think tracking the data should be a problem, but the pace + micromanagement is definitely one of the things that concerns me. (in all RTS''s)
In fact, it is almost worth a separate thread...

Share this post


Link to post
Share on other sites
You misunderstood my post =). I am saying that, keep Hit Points as an internal value (the player doesnt know it exists), but let the atual unit see it. I mean, if you get shot in the gut, you know you are probably going to die. Same thing. The unit knows "I have x hitpoints left. I have been taking y damage every second. Time to run like a baby", and he runs. The player never sees this. It is all taken care of within the unit. You see? We remove it from the player''s view, but keep it around to speed up AI calculations. Example.

Infantry Man A just got his leg blown off by Tank B. If A had 10 HP to start with, he now has 1, from factors such as weaponry, mobility, etcetera. A knows that he is probably dead. If A has good morale to begin with, he pulls out an AT grenade, pulls the pin, and chucks it at the tank. Otherwise, he screams. The tank sees the grenade, but, ut oh, its too late. It takes normal damage, but its been beaten up pretty badly, so, say the tank started with 100 HP, but now it has 20. It should probably retreat.

The idea could be extended even more, by using the internal HP value to actually measure the base chance of a critical Hit, using the formula

percent chance = (maxHP - currHP) * 100;
if(percent chance == 100) percent chance = 99;

So, from the above example, the infantry guy would have a 90% chance of a critical hit, while the tank has 80%. This base would then be modified by the type of weapon the attacker is using (Super Tank A will always destroy Tank B if Tank B has been hit, etcetera).

And, again, the player never sees the magic HP value =).

Z.

Share this post


Link to post
Share on other sites
This is an interesting system, but I think for it to be effective you must absolutely always stick to armour and projectile types. I mean the now famed, almost dead infantry with a rifle is NEVER going to kill a tank with several inches of armour by shooting it. Enless of course the tank has physical holes in it, but then, the tank shouldn''t just blow up or something like that. It would be more likely that the infantry guy''s bullet would bounce off the tank and kill his friend. If the infantry is allowed to through grenades in the hatch (don''t they lock them anyway?) then you should allow the tanks to gain speed to avoid the infantry trying to climb on them.

You need to make rules and stick to them 100%. 50 infantry shouldn''t be able to kill a tank by shooting at it waiting for a critical hit.

On another note, I think each unit should have some special abilities (like throwing grenades down a tank''s hatch.) I mean every single unit having unique abilities. Not simply different damage types, but things like projectile from tank C has a chance of not going off when it hits a target with D type armor (a rather commonly used armour or something.) This way the units and their counters (multiple) are unique and useful. Of course you might have to spend forever balancing things...

Share this post


Link to post
Share on other sites
about the micromanagement thing. if controlling small squads requires such a level of control, why not make the option for multi-controlled teams, like they have in starcraft or age of empires. that way you can assign your offense, defense, and resource gathering to a particular team member, and they can specialize their micro-management from there. just a thought.

cyn

Share this post


Link to post
Share on other sites