318 replies to this topic

[Carrioncrows]

Marmon Rzohr, on 20 September 2013 - 03:53 AM, said:

The proposed Dragon design would not make XLs any less viable. From the front it will still be hard to hit the STs from the front, but it would give you the option of twisting to absorb damage if you want to.

Yes, exactly.

[Karl Streiger]

Hows going on - will we have those hitboxes today with a hot fix - or do we have to wait 1 1/2 weaks until the next patch?

I do want - I can even wait 2 additional months for CW or UI 2.0 but this is great....

[DaZur]

@OP

Very nice synopsis and an admirable intent... Problem is it's not likely plausible and here's why.

There are a couple different ways in which games institute damage-collision (hit boxes):

• Geometry = Higher-level combat games actually use the models lowest-level LOD (level-of-detail) geometry to register hits / damage. This typically results in the most accurate damage model but comes as a cost of CPU cycles.
  
• Bounding boxes = Using global axis (X,Y,Z) points based off the vertex points of the given object, the game numerically defines a "bounding box" that is representative of that objects hit-box.

In both of examples the hit-box is attached to an object it's meant to represent. If MW: is using geometry-based hit-boxes, it's almost impossible "fudge" the hit-boxes outside the actual geometry unless you actually cut / break off bits / parts of other objects and associate them via the linked hierarchy... which is model-messy, time consuming and overtly complicated.

If it's volume based, in theory, as you illustrated, PGI "could" extend / shape the hit-boxes to balance their ratios. But this also creates problems. If say the arm hit-box extends into the side-torso and the arm is destroyed and detached, so to is the hit-box... This leaves a side-torso that now has a bad bounding damage volume in that there will be a portion of the side-torso geometry that does not have a matching hit-box volume... i.e. shots to that area will simply pass through and not register damage as there as the hit-box volume was detached with the arm.

Long and short... PGI can fudge ever so slightly but cannot cross geometry / objects / parts with their hit-boxes, lest when the lowest level object is removed from the mech (arm to a side torso or a side-torso to a center torso etc..) it will leave a gap in the geometry where the hit box profiles no longer match / register damage.

Not trying to be a Debbie-downer... just throwing the facts out there for reference.

[-Muta-]

Should be frustatign to have such a bic CT

[N a p e s]

Make it happen PGI!

[General Taskeen]

You know when they said the Orion was ready and yet they just wanted more "meta" from releasing the Victor and Quickdraw? How in the world could they have released the Orion, with the protruding CT, after apparently having it "ready" all that time, and not have seen how that hitbox design would somehow not be an issue based on what has been shown to them so far?

Hitbox philosophy needs to be worked on now, before introducing more Mechs that may or may not suffer the same fate. It is more work for them fix if they don't.

[Steve Pryde]

Awesome post, but you forget the catapult, same issue with the center torso and the problem with his head isn't fixed either. Wanted to buy the Protector but with this big center? No chance.

[Tweaks]

I'm probably one of the few, but I really don't like these suggestions. I think it makes all 'Mechs work the same way and almost nullifies their differences.

The point of having 'Mechs with different types of geometry and different types of designs is that they don't behave the same way and don't all have the same weaknesses either.

By making left and right torsos consistently equal in proportion with the CT on all 'Mechs, it means there's no more challenge in knowing the specifics of each 'Mech anymore and learning their weak points. You just have to aim at about the same spot on each 'Mech you see and you'll get the same results.

I think this would make the game very boring and too much of a simple shooter with robots.

I vote no, and although they "may" be adjusting the CT hitbox of some variants because they are just wrong, I don't think it's more than one or two that would be affected by this.

I mean, for example, your idea of a perfect Hunchback 4G is that the whole right side of its giant shoulder mounted ballistic hardpoint is part of the arm? That's just wrong. Shooting this has no effect on the arm, the actuators and shoulder joints are beneath that, not in it.

I think most of the hitboxes now are based on the internal skeleton, the limb actuators, the location of internals and and junctions between components, and also the shape and form of the armour plates that's on top of it. The devs have to take all that into consideration when defining them.

Edited by Tweaks, 20 September 2013 - 12:10 PM.

[John MatriX82]

Carrioncrows, on 18 September 2013 - 11:01 PM, said:

Feel free to post up your thoughts, but at the very least I hope it gets PGI thinking about hitboxes and the future.

Well posted, you already did good before, this is even better!

Rascula, on 18 September 2013 - 11:36 PM, said:

The Awesome has had a fan base crying out for this kind of work for a very long time now and nothing has happened and the Kintaro was flat out broken rather than just borked hence why it got attention so fast.

So whilst I agree with most of your analysis of the hit boxes, and agree they need some tweeking I think as far as the developers are concerned they are working as planned and I doubt anything will change.

Yep and you know why? Because Kintaro had an Hero mech that was so flawed that the backlash was much steeper than the one for the Pretty baby; that, together to the normal cbills unsold variants (and PGI's wonderful telemetry) told them that something had to be done, especially for the GB.

Evidently the PB sold more, so they are still thinking Awesome's hitboxes (and scale) are.. awesome.


Also Dragons as pointed out in this thread are basically dead. I'm going to get rid of my beloved 1N soon, killed by
-the absurd tube-like SRM pattern,
-huge CT,
-wrong scale,
-abysimal new Gauss rifle mechanic,
-an entire extra lance of enemy mechs that can shoot it with even more ease,
-the new dull movement mechanics that paired up with the lack of JJ makes its movement hard and prevents often good flanking on several maps such as canyons,
-Elo putting that chassis against fairly more efficient death machines,
-lack of collisions (I know it was bugged, but yanking lights in Dragons was.. yummy),

that truly make it a moving wreck of an old and now forgotten era. At least for me of course. I hope they'll rework the hiboxes in the future, but they did it once for a newly relased mech that isn't part of those of the old bulk (like the Awesome is) that clearly have already "been paid" by the founders and early buyers.

The kintaro is fresh, evidently the new eager 3PV players and those gathered by the explosively outstanding launch, weren't responding well and they had to cash in asap reworking the hitboxes. Meh.

Edited by John MatriX82, 20 September 2013 - 11:38 AM.

[General Taskeen]

Carrioncrows, on 19 September 2013 - 02:55 PM, said:

A accomplished marskman would still be able to get around this "Flaw of the big hunch design" but this should more than help alleviate loosing the cannon out the door.

For this, I would like to point something out. The big hunch on a Hunchback was not actually really a "flaw" in TT, as far as I know, it was not easier to hit when rolling, its just all random, no matter what size - but I know what you are saying.

The flaw was really just very little ammo. And when translating from TT to real-time the problem is inherent to Mech Warrior trying to shoe-in hitbox area's and trying to keep them 'accurate' to the original art, and then combine that with pin-point aim, is why we have arrived at this problem and the Devs that need to be more careful in their hitbox design philosophy. I think hitbox design philosophy was something other Dev groups got a little better in previous Mech titles.

If I wasn't at work, I would draw on your re-adjusted Hunchback hitboxes with "damage reduction" area written all over that right torso, but only on the sides of it, instead of the front, to account for PGI wanting to keep the original hunch art into real-time.

[Sable Dove]

I like your thinking. This would go a long way to improving survivability for many mechs.

However, if you don't mind, I'd like an explanation as to why you think the Catapult is fine as it is, when proportionally it has one of the largest CT hitboxes in the game (not to mention the giant head hitbox). As bad as the Dragon, but slower, a larger mech in general, and doesn't have arms capable of protecting its torso at all.

The Catapult may be okay now, but if you change all the other bad-hitbox mechs, the Catapult will be useless, because every other heavy will have 2-3 times as much effective armour as it.


And as a side note, you would be doing less damage overall, because only 50% damage is transferred from a destroyed section. Not that it's a bad thing. More survivability is a good thing.

[Carrioncrows]

DaZur, on 20 September 2013 - 06:22 AM, said:

@OP

Very nice synopsis and an admirable intent... Problem is it's not likely plausible and here's why.

There are a couple different ways in which games institute damage-collision (hit boxes):

• Geometry = Higher-level combat games actually use the models lowest-level LOD (level-of-detail) geometry to register hits / damage. This typically results in the most accurate damage model but comes as a cost of CPU cycles.
  
• Bounding boxes = Using global axis (X,Y,Z) points based off the vertex points of the given object, the game numerically defines a "bounding box" that is representative of that objects hit-box.

In both of examples the hit-box is attached to an object it's meant to represent. If MW: is using geometry-based hit-boxes, it's almost impossible "fudge" the hit-boxes outside the actual geometry unless you actually cut / break off bits / parts of other objects and associate them via the linked hierarchy... which is model-messy, time consuming and overtly complicated.




If it's volume based, in theory, as you illustrated, PGI "could" extend / shape the hit-boxes to balance their ratios. But this also creates problems. If say the arm hit-box extends into the side-torso and the arm is destroyed and detached, so to is the hit-box... This leaves a side-torso that now has a bad bounding damage volume in that there will be a portion of the side-torso geometry that does not have a matching hit-box volume... i.e. shots to that area will simply pass through and not register damage as there as the hit-box volume was detached with the arm.

Long and short... PGI can fudge ever so slightly but cannot cross geometry / objects / parts with their hit-boxes, lest when the lowest level object is removed from the mech (arm to a side torso or a side-torso to a center torso etc..) it will leave a gap in the geometry where the hit box profiles no longer match / register damage.

Not trying to be a Debbie-downer... just throwing the facts out there for reference.

Yup I knew this.
<-- working towards degree in game and art design.

Some of these are indeed easy fixes and I figured the LOD method wouldn't be preferred for MWO just because I doubt they would need that level of precision at the cost cycles for what is ultimately 7 hit boxes.

I was betting on the Bounding Box theory.



And though redrawing the bounding boxes would become time consuming in some cases they are already going to do a "Mech pass" going back to those older mechs to make them "Modular" so when you switch out weapons it changes what the mech actually looks like.

But if you also remember they fixes the Jager head hitbox in a mater of days. (EDIT: which now that I think about it may not of been a fix due to LRM-apocalypse splash) The KTO took longer because they actually pulled the mech apart and redid where the KTO torso twisted from which is a much bigger fix outside of just bounding box.

*Fingers crossed*

=)

-Crow

Tweaks, on 20 September 2013 - 11:08 AM, said:

I'm probably one of the few, but I really don't like these suggestions. I think it makes all 'Mechs work the same way and almost nullifies their differences.

The point of having 'Mechs with different types of geometry and different types of designs is that they don't behave the same way and don't all have the same weaknesses either.

By making left and right torsos consistently equal in proportion with the CT on all 'Mechs, it means there's no more challenge in knowing the specifics of each 'Mech anymore and learning their weak points. You just have to aim at about the same spot on each 'Mech you see and you'll get the same results.

I think this would make the game very boring and too much of a simple shooter with robots.

Look at the mechs I posted that needed a redesign. Even the Hunchback which I admitted was boarder line if not blatantly over the top "Janky" solution.

The individual weakness of the mechs did not disappear, they have just been mitigated. Why? Because those weakness would massively penalize the player for taking that mech instead of another with the same tonnage that doesn't have the inherent weakness of an easy to hit center torso.

As I also said, not every mech needs this. Most mechs are fine, but the ones I grabbed are some of the more extreme examples that absolutely need dealt with.

Sable Dove, on 20 September 2013 - 02:12 PM, said:

I like your thinking. This would go a long way to improving survivability for many mechs.

However, if you don't mind, I'd like an explanation as to why you think the Catapult is fine as it is, when proportionally it has one of the largest CT hitboxes in the game (not to mention the giant head hitbox). As bad as the Dragon, but slower, a larger mech in general, and doesn't have arms capable of protecting its torso at all.

The Catapult may be okay now, but if you change all the other bad-hitbox mechs, the Catapult will be useless, because every other heavy will have 2-3 times as much effective armour as it.


And as a side note, you would be doing less damage overall, because only 50% damage is transferred from a destroyed section. Not that it's a bad thing. More survivability is a good thing.

I still see many many K2's and Kat's on the battlefield that acquit themselves well in both sniper and brawler roles. I see the side torso peeled off just as much as I see the CT hammered, some times they have a STD engine sometimes an XL.

I do agree with you on the head hitbox that needs to be downsized a lot, and the pelvic / pelvis area should be split between the legs. I felt once that happens the Catapult should be about perfect.

I am not trying to completely rewrite the "Quirks" of a particular Battlemech, I am just trying to mitigating the more absurd hitboxes into something that makes a bit more sense.

The Catapult has a big nose and the Dragon has a big nose.

The difference is when the Catapult torso twists he starts taking damage on his left and right torso's where as the dragon doesn't. Is it perfect? Probably not but i didn't as someone put it "make them all the same"

It may just come down to a personal preference thing.

But hey if the Catapult needs it, lets do it.

It sets precedence for the the Madcat.

Edited by Carrioncrows, 20 September 2013 - 03:21 PM.

[Sable Dove]

I can't really agree with that. I can count on one hand the number of times I've seen a Catapult side-corred, and even fewer times when it's lost a side torso and lived (except for the obvious dual AC20 build). When a Catapult torso-twists, you might get maybe a quarter of the damage going to your side torso if you're lucky. Against skilled players, all torso-twist does is make it harder to keep track of your enemy while they pile damage on your CT.

Anyone even remotely familiar with the mechs shoots for - and hits - the CT. From the side, the CT hitbox is easily twice the size of the ST. I really don't think that having a grossly over-sized CT hitbox really qualifies as a quirk when the mech has no benefits in return other than unintentionally-good hardpoints (being able to use dual Gauss or dual AC20s, for instance, which are getting nerfed regularly)

Also, your hitboxing of the Dragon disagrees with the in-game testing here: http://mwomercs.com/...x-localisation/
I'm guessing you're going by the mechbay hitbox indication, which is very wrong. The mechbay makes the Catapult's ST look significantly larger than they are, and the Dragon's look much smaller than they are. In terms of hitboxes, the Catapult's are definitely worse than the Dragon's.

The Dragon hitbox, while large, is not nearly as large as the Catapult's (proportionally or absolutely), and as I said, it's faster, and has arms that at least somewhat block the torso. If you change the Dragon, it becomes a faster, tougher, and smaller target. Your explanation of the Dragon fix effectively doubles its armour. So now the 60-ton Dragon is sprinting around with an effective armour of 176, while my slower, much larger, 65-ton Catapult is stuck with its 84 CT armour because it's hard not to hit the CT.


In terms of bad hitboxes, from worst to best (more or less):
Jenner (somewhat mitigated by size and speed), Orion (as far as I know; I'm not sure what the actual hitboxes are in-game), Catapult, Awesome (probably; I don't see them around enough to judge), Dragon, Cicada.

You've proposed fixes for all of these mechs except the Catapult, which has some of the worst hitboxes in the game. Humongous, impossible-to-miss CT, and a big head that's also easier to hit than most mechs (and if you do miss, you still hit CT, so it's not wasted damage).

[Tweaks]

Carrioncrows, on 20 September 2013 - 02:39 PM, said:

The individual weakness of the mechs did not disappear, they have just been mitigated. Why? Because those weakness would massively penalize the player for taking that mech instead of another with the same tonnage that doesn't have the inherent weakness of an easy to hit center torso.

That's the thing I disagree the most with. It's perfectly fine that some 'Mechs do penalize the player in a certain way compared to another "possibly" better 'Mech of the same tonnage, simply because it has other advantages. For example, a Hunchback 4G pilot has to learn to protect his right torso or he's pretty much dead. However, if he can, he can use his left torso as a shield safely (for a while at least) since it's got no weapons in it. That's part of the whole fun. The Orion has to protect his center torso and can't just stay facing the enemy while shooting or he will die. He has to twist often between shots.

The point is that each 'Mech variant has a learning curve, and it's not just about feeling its size and knowing how it handles turns and such, but knowing its weaknesses and knowing how to fight efficiently despite them and to get the most out of it. When someone says "I'm a Hunchback pilot", it has to mean something, and not just that this guy has a Hunchback in his 'Mechbay.

I compare this to driving sports cars. They may all look similar, but a Ferrari handles very differently than a Porshe, and it takes time to get used to each even though they may be in the same "class" in terms of speed and power.

Personally that's what I like about this game. The different 'Mech chassis are not simply aesthetic covers over identical metal skeletons. They actually are anatomically different, feel different, handle differently and have different key weaknesses and strengths. It takes skill and practice to master multiple chassis of the same class and it's just fine like this.

Like I said though, for balance purposes, they could adjust the model and hitboxes of certain 'Mechs which are sort of unfair right now. The Orion and the Spider come to mind, but it's about the only ones I can think of, but I don't agree with the overall changes you're suggesting.

Edited by Tweaks, 20 September 2013 - 04:29 PM.

[Zakie Chan]

Good post, I fully support all proposed changes.

However the new leg design would be best if accompanied with a modified legging system. I could see a huge increase in leg deaths for mediums and lights.

[Fate 6]

I logged in after a month of not playing just to upvote this

[Carrioncrows]

Tweaks, on 20 September 2013 - 04:22 PM, said:

That's the thing I disagree the most with. It's perfectly fine that some 'Mechs do penalize the player in a certain way compared to another "possibly" better 'Mech of the same tonnage, simply because it has other advantages.

I am all for quirks of different mechs, but there is a big difference between a perk and a massive penalty that makes a mech all but usable.

Are all my proposed changes the best idea and should be used with no deviation? Hell no.

But it's a start in the right direction for a lot of these mechs.

I'm not trying to kill the personality of the mech, but I "DO" want to use it with out getting my *** handed to me every time I step out in a match, not because i'm a bad pilot or they are great shots but because the hit boxes are atrocious.

Sable Dove, on 20 September 2013 - 04:04 PM, said:

***

You convinced me. Working up designs.

[Xeno Phalcon]

Hitbox system is desperately in need of some hardcore tweaking, currently one of its biggest victims are the hunchback, dragon, awesome and the like. But what happens when we get mechs with massive shoulder pods like the timber wolf? the current system would make it a terrible mech for XL engines thats for damn sure.

[Tweaks]

Carrioncrows, on 20 September 2013 - 07:20 PM, said:

I am all for quirks of different mechs, but there is a big difference between a perk and a massive penalty that makes a mech all but usable.

Are all my proposed changes the best idea and should be used with no deviation? Hell no.

But it's a start in the right direction for a lot of these mechs.

I'm not trying to kill the personality of the mech, but I "DO" want to use it with out getting my *** handed to me every time I step out in a match, not because i'm a bad pilot or they are great shots but because the hit boxes are atrocious

Well, sorry to say this, but you'll just have to learn to play the 'Mech you want to play like everybody else and just try harder until you get it. And if you keep getting your {behind} handed to you every time you step out in a match, well tough luck, but you just need to practice more. There are tons of players doing well with the same 'Mechs you're being {behind}-handed by. It's not an easy game. Other than two 'Mechs I can think about, the rest is perfectly balanced to me and I see plenty of people do great with them, Don't blame the hitboxes for being the cause of your lack of skills...

Xeno Phalcon, on 20 September 2013 - 08:19 PM, said:

Hitbox system is desperately in need of some hardcore tweaking, currently one of its biggest victims are the hunchback, dragon, awesome and the like. But what happens when we get mechs with massive shoulder pods like the timber wolf? the current system would make it a terrible mech for XL engines thats for damn sure.

All of the 'Mechs you mentioned are just fine like that. You just need to learn to pilot them correctly to avoid being disabled. Stop trying to just rush into a brawl head-first like a {mentally challenged person} and you may have more luck. I've seen plenty of Dragons and Hunchbacks do good in matches and survive.

Edited by Tweaks, 20 September 2013 - 08:26 PM.

[Faithsfall]

Tweaks, on 20 September 2013 - 04:22 PM, said:

That's the thing I disagree the most with. It's perfectly fine that some 'Mechs do penalize the player in a certain way compared to another "possibly" better 'Mech of the same tonnage, simply because it has other advantages. For example, a Hunchback 4G pilot has to learn to protect his right torso or he's pretty much dead. However, if he can, he can use his left torso as a shield safely (for a while at least) since it's got no weapons in it. That's part of the whole fun. The Orion has to protect his center torso and can't just stay facing the enemy while shooting or he will die. He has to twist often between shots.

The point is that each 'Mech variant has a learning curve, and it's not just about feeling its size and knowing how it handles turns and such, but knowing its weaknesses and knowing how to fight efficiently despite them and to get the most out of it. When someone says "I'm a Hunchback pilot", it has to mean something, and not just that this guy has a Hunchback in his 'Mechbay.

I compare this to driving sports cars. They may all look similar, but a Ferrari handles very differently than a Porshe, and it takes time to get used to each even though they may be in the same "class" in terms of speed and power.

Personally that's what I like about this game. The different 'Mech chassis are not simply aesthetic covers over identical metal skeletons. They actually are anatomically different, feel different, handle differently and have different key weaknesses and strengths. It takes skill and practice to master multiple chassis of the same class and it's just fine like this.

Like I said though, for balance purposes, they could adjust the model and hitboxes of certain 'Mechs which are sort of unfair right now. The Orion and the Spider come to mind, but it's about the only ones I can think of, but I don't agree with the overall changes you're suggesting.

While you have mentioned in the last paragraph that the hitboxes on the Orion are unfair you did mention at the start that the orion pilot needs to torso twist more to avoid the incoming fire, well I am one of those orion pilots and torso twist on the orion still allows a large amount of incoming fire to still hit CT.

But this still isn't just a matter of torso twist, for instance (it happens to the best of us) you come around a corner and you are suddenly facing another mech, you both fire a snap shot/alpha, in certain mechs Orion/awesome for instance most of that snap shot hit's CT while firing at something else it might hit CT/ST.

Straight away 1 of those mechs has taken more dmg to 1 area than the other, You both start to torso twist and again certain mechs like the Orion still allow some if not all of the dmg through to CT, while the other mech absorbs some now to ST/arm.

At each point of weapons fire certain mechs are having 1 area dmg'd unless the shot is really wild while the other is able to distribute the dmg more evenly over their mech.

The way the hitboxes have been reworked by Carrioncrow will allow for more survivability of alot of mechs, this again can be seen as a good thing as some mechs go down far to easily to stray shot's, now this still won't help you if you constantly walk straight towards the enemy, but would help those pilot's that are at least attempting to protect themselves.