24 replies to this topic

[MustrumRidcully]

This is basically a brainstorming question:
Is there a way to balance Hit Boxes without changing Hit Boxes?

We know that some mechs are deemed ineffective or weak because they have "bad" hit boxes, like the Awesome (Torso too wide and too easy to hit), while others are considered having advantages for having "good" hit boxes, like the Catapult with a very small side torso (making carrying XL Engines less risky than on other mechs).

So, there is the option to change the hit boxes of mechs to "fix" such imbalances. But what if the devs are unable or unwilling to alter the models and hitboxes of mechs? What if the design of a mech just looks best if it's a wide target? And it's not just relative hit boxes, it can also be size - some mechs are large, possibly larger than is good for them, and if the devs won't shrink them, what could be done?

Would there be a way to balance hit boxes without changing them? Give a mech with a large side torso some kind of advantage to compensate for that weakness?

For example, one idea I came up with was to apply some kind of damage resistance to larger hit boxes. Another was to have a damage transfer mechanic that applies damage from a large hit box to a smaller hit box.

Is there a way to mathematically express such advantages? Say, calculate the area of a hitbox from a typical targeting position (e.g. side profile, front or back?)?

Or is that just impossible, and we can only hope some other "quirks" can compensate the drawbacks?

Edited by MustrumRidcully, 28 May 2013 - 09:30 AM.

[Deathlike]

Well, as you can see from the Raven hitbox nerf, everyone can shoot their legs now. So clearly it's possible to change the hitboxes to keep an oversized mech like the Awesome viable.

[TexAce]

I think you are confusing something.

Hitboxes are not the actual mesh of your mech, they are invisible and have lower details, sometimes are bigger or smaller than the actual section they are covering. It's like an invisible layer of lowpoly geometry engulfing your mech. So resizing them doesn't mean you are resizing anything you can see.

But seems like the only guy, who knows how to do this has left PGI months ago.

Edited by TexAss, 28 May 2013 - 09:22 AM.

[MustrumRidcully]

TexAss, on 28 May 2013 - 09:18 AM, said:

I think you are confusing something.

Hitboxes are not the actual mesh of your mech, they are invisible and have lower details, sometimes are bigger or smaller than the actual section they are covering. It's like an invisible layer of lowpoly geometry engulfing your mech. So resizing them doesn't mean you are resizing anything you can see.

But there are limits to what you can do. If your 50 Ton Centurion is as large as a 85 Ton Stalker, how could you adjust the hit boxes so they are balanced without misleading people. If, to hit the arm, I had to shoot between arm and torso, we'd have a problem, wouldn't we?

Quote

But seems like the only guy, who knows how to do this has left PGI.

Together with the guy that came up with the good idea of giving lasers a beam duration to avoid hit-scan alpha murder strikes?

Edited by MustrumRidcully, 28 May 2013 - 09:23 AM.

[TexAce]

MustrumRidcully, on 28 May 2013 - 09:23 AM, said:

But there are limits to what you can do. If your 50 Ton Centurion is as large as a 85 Ton Stalker, how could you adjust the hit boxes so they are balanced without misleading people. If, to hit the arm, I had to shoot between arm and torso, we'd have a problem, wouldn't we?

of course you have to keep the height of your mech. Because not doing so would mean to hit the head while aiming on the centurions CT.
But what can be done is for example make the hitbox of the Awesomes CT not stick out so much, or let his arm hitbox be part of his shoulder. Or make his arm hitbox slightly bigger than the arm so more damage is soaked up there.

Size differences between mechs is a completely different topic.

Edited by TexAss, 28 May 2013 - 09:28 AM.

[Hotthedd]

I always thought the different hitboxes WERE a balancing act.

Sure, the Catapult's tiny side torsos give it the advantage of being able to use XL engines, BUT the result is a bigger center torso, making it harder to dodge those hits and being more prone to being cored.

There is an advantage and a drawback either way.

[Deathlike]

TexAss, on 28 May 2013 - 09:18 AM, said:

But seems like the only guy, who knows how to do this has left PGI months ago.

MustrumRidcully, on 28 May 2013 - 09:23 AM, said:

Together with the guy that came up with the good idea of giving lasers a beam duration to avoid hit-scan alpha murder strikes?

Who is this guy, and why isn't he still @ PGI? The man deserves a medal.

The only way that would be fair, w/o changing the hitbox would be something along the line of the missile bay doors. Provide additional protection to a specific component with a tradeoff.

For instance, I like my own idea to provide the Hunchy's hunch like the HBK-4G or HBK-4H to receive 10% damage reduction for like a .5 to 1 second shot delay. Something similar could be applied to the HBK-4P, except increasing the heat by 10% than normal for such protection (well, we could always go back to the shot delay idea).

That's the only fair way that I can think of.

Edited by Deathlike, 28 May 2013 - 09:34 AM.

[MustrumRidcully]

TexAss, on 28 May 2013 - 09:25 AM, said:

Size differences between mechs is a completely different topic.

It was definitely intended to be part of this thread, however.

[Fooooo]

TexAss, on 28 May 2013 - 09:18 AM, said:

I think you are confusing something.

Hitboxes are not the actual mesh of your mech, they are invisible and have lower details, sometimes are bigger or smaller than the actual section they are covering. It's like an invisible layer of lowpoly geometry engulfing your mech. So resizing them doesn't mean you are resizing anything you can see.

But seems like the only guy, who knows how to do this has left PGI months ago.

Well the thing is I don't think cryengine exactly uses hitboxes in a traditional sense. It does some type of mesh detection or collision check with the model etc I think.

I'm not sure if it uses just the low poly version of the model or a specific geometry they setup. (much like how you setup the collision proxy)

From that, the actual part hit is based on textures or something. So they can paint this geometry with different textures to represent different parts.

So if they painted the texture for the CT on the legs, then hitting the legs would tell the game to send dmg to the CT etc......

I think the change they made not long ago was just adjusting this texture for the legs, or adjusting some of the geometry of the model / proxy model etc......


One of the devs confirming would be nice, but I cant seem to find a post about it............

thomas ? Omir ? Anyone ??

[Cyke]

Wait a second, are you guys ******** me?

PGI staff that are responsible for some good ideas (laser beam duration, hitbox tweaking) have left the company?

I'm just going to assume you guys are pulling legs here, because if you aren't kidding.. the sad reality of why MWO has so much awesomeness mixed in with the recent terrible development actually makes sense.


Edit: Don't get me wrong, I still love the game and I'm probably into this for the long run, I appreciate the recent technically complex improvements they've made like Host State Rewind, and I'm looking forward to eventually seeing a complete and well-made game that I'll be willing to spend oodles of money on.. but I notice that they've been floundering quite badly in terms of balancing weapon stats and behavior.

Right now I'd like to believe that they're waiting for Missile HSR to be completed before they make a balancing pass over the game's entire weapon arsenal and mechanics... but if what you guys say is true, it sort of provides an alternate explanation to the situation, and makes the game's future look a whole lot bleaker.

Edited by Cyke, 28 May 2013 - 09:38 AM.

[TexAce]

Fooooo, on 28 May 2013 - 09:32 AM, said:

Well the thing is I don't think cryengine exactly uses hitboxes in a traditional sense. It does some type of mesh detection or collision check with the model etc I think.

I'm not sure if it uses just the low poly version of the model or a specific geometry they setup. (much like how you setup the collision proxy)

From that, the actual part hit is based on textures or something. So they can paint this geometry with different textures to represent different parts.

So if they painted the texture for the CT on the legs, then hitting the legs would tell the game to send dmg to the CT etc......

I think the change they made not long ago was just adjusting this texture for the legs, or adjusting some of the geometry of the model / proxy model etc......


One of the devs confirming would be nice, but I cant seem to find a post about it............

thomas ? Omir ? Anyone ??

They are using a low poly model of the mech, very low poly though. The damage texture is just an overlay image that gets applied to the intial mech texture on top of it, depending on where the hit detection was made. But since it gets applied to the original mesh texture, not the low poly one, it always is sticking correctly to the mech and not floating around it.

Where the damage texture gets applied doesn't play any role when it comes to hitboxes, because they come after the hitbox check.

Cyke, on 28 May 2013 - 09:34 AM, said:

Wait a second, are you guys ******** me?

PGI staff that are responsible for some good ideas (laser beam duration, hitbox tweaking) have left the company?

I'm just going to assume you guys are pulling legs here, because if you aren't kidding.. the sad reality of why MWO has so much awesomeness mixed in with the recent terrible development actually makes sense.


Edit: Don't get me wrong, I still love the game and I'm probably into this for the long run, I appreciate the recent technically complex improvements they've made like Host State Rewind, and I'm looking forward to eventually seeing a complete and well-made game that I'll be willing to spend oodles of money on.. but I notice that they've been floundering quite badly in terms of balancing weapon stats and behavior.

Right now I'd like to believe that they're waiting for Missile HSR to be completed before they make a balancing pass over the game's entire weapon arsenal and mechanics... but if what you guys say is true, it sort of provides an alternate explanation to the situation, and makes the game's future look a whole lot bleaker.

Yes we are kidding. Relax.

Edited by TexAss, 28 May 2013 - 09:41 AM.

[Deathlike]

TexAss, on 28 May 2013 - 09:40 AM, said:

Yes we are kidding. Relax.

I couldn't tell if that was sarcasm or serious. It worked too well!

[MustrumRidcully]

Cyke, on 28 May 2013 - 09:34 AM, said:

Wait a second, are you guys ******** me?

PGI staff that are responsible for some good ideas (laser beam duration, hitbox tweaking) have left the company?

No, we're just starring with you. At least we wouldn't know if anyone left.

[scJazz]

Is adding "fake doors" to some of these mechs a stupid plan? Doors that don't open or close but do provide damage reduction. Just asking cause it seems to me that...

1) It would be easy
2) Door protection value doesn't have to be a constant so it would also be scalable
3) Doesn't require the Artsy dudes to do as much work in terms of balancing

Totally dumb?

[MustrumRidcully]

scJazz, on 28 May 2013 - 10:46 AM, said:

Is adding "fake doors" to some of these mechs a stupid plan? Doors that don't open or close but do provide damage reduction. Just asking cause it seems to me that...

1) It would be easy
2) Door protection value doesn't have to be a constant so it would also be scalable
3) Doesn't require the Artsy dudes to do as much work in terms of balancing

Totally dumb?

Basically, a damage reduction value. Yes, perhaps. But how could you arrive at a balanced one? Is there a way to do it with a mathematical model, or is it all gut feeling.

[OneEyed Jack]

Possible? Probably. Balanced? No

The mechs that have missile bay doors give up something to use them. Applying the same thing to other mechs with no downside would be exactly the opposite of balance.

[Deathlike]

MustrumRidcully, on 28 May 2013 - 10:48 AM, said:

Basically, a damage reduction value. Yes, perhaps. But how could you arrive at a balanced one? Is there a way to do it with a mathematical model, or is it all gut feeling.

I'm not sure if you could get to a balanced one.. the value would have to be BOTH mathematical and feel (and one will eventually contradict the other).

For instance, the Catapult's arms are LITERALLY its ears (outside of the K2, which was simply guns). So a damage reduction overall is sufficient.

The Stalker's arms under most variants (outside of the half-armed 4N and Misery) get the same protection that missiles do, even if it had no missiles in it.

For the Hunchy, if the Hunch on a Hunchback is 15% of the side torso's profile, perhaps the protection would be 15% less damage to that section. Mind you, I would still prefer a fire delay for ballistic weapons and a heat "penalty" for energy weapons in the "Hunch" area.

I didn't like what they did in MW4 in terms of how the armor values were generated, especially for the specialized slots, but that's definitely not a route that should be taken.

Edited by Deathlike, 28 May 2013 - 10:58 AM.

[MustrumRidcully]

OneEyed Jack, on 28 May 2013 - 10:54 AM, said:

Possible? Probably. Balanced? No

How can it be a method that balances hit boxes is possible but it wouldn't balance hit boxes?

Quote

The mechs that have missile bay doors give up something to use them. Applying the same thing to other mechs with no downside would be exactly the opposite of balance.

What if the resistance was not actually static, but instead dependent on proximity to the center of the surface? Say, because an Awesome side torso is 20 % larger than for the "ideal" 80 ton mech, it gets a damage reduction going from 0 to 40 % from the center to the most outlying point. (It's probably a bit more complicated, something about integrals I figure).

Maybe it has a 20 % larger side torso, but also a 20 % smaller arm, so instead of a damage reduction, the damage is partially transferred, from 0 in the center of the side torso to 40 % at the most outlying end towards the arm.

Something like that.

[scJazz]

MustrumRidcully, on 28 May 2013 - 10:48 AM, said:

Basically, a damage reduction value. Yes, perhaps. But how could you arrive at a balanced one? Is there a way to do it with a mathematical model, or is it all gut feeling.

Do we even need to start with complicated? Bearing in mind that my idea is to implement an existing mechanic to help out the gimped mechs. Therefore not requiring a hell of a lot of effort from PGI.

Just start with sticking the existing door mechanic on the gimped designs. Period. 10% damage reduction as long as your not stupid enough to actually open the fake door. Although it does cover up the missile hardpoint on a Dragon. So he has a choice. After that thought can be put into 5%, 10%, 15%, etc. After that they can change the "door" code to include a "door" that doesn't open, close, get destroyed, or add the 0.5sec delay.

This still leaves Dragon's in a sensitive spot when it comes to very large CT causing large splash damage from missile strikes but gives the design a quirk. Dragon's are allergic to missiles! Fit an AMS! Etc... similar stuff for the Awesome.

[MaddMaxx]

I can't see having the "gimped" mechs having damage reduction going anywhere as what it really equates to is more armor in those areas. No one will give a **** about your supposed "gimped" mech but they will be very dissatisfied that their "non-gimped" Mech does not have that same amount of armor.

What with all the talk about Cone of Fire and Spreading of Damage, why not use the Hit Boxes to do that spread.

Example:

I hit the CT 2 times, it takes that damage, I, someone, hit that same CT a 3rd consecutive time, well 50% of that damage value simply gets shifted over to the Left or Right Torso. Same for the other sections. All damage is always applied, so no one gets cheated, and everyone has the same armor levels.