78 replies to this topic

[focuspark]

is it an issue with SRM hit detection (known bug) or Spider hit detection?

[aniviron]

focuspark, on 04 October 2013 - 02:28 PM, said:

is it an issue with SRM hit detection (known bug) or Spider hit detection?

Replicable with ballistics. It's easier to do with SRMs.

[focuspark]

aniviron, on 04 October 2013 - 03:23 PM, said:

Replicable with ballistics. It's easier to do with SRMs.

well then, it's the chassis. does Raid (bug spray) work?

[Gralzeim]

Here's hoping a 'fix' doesn't entail the sort of band-aid 'fix' that they did with the Raven's legs (which was supposed to 'fix' issues with people having a hard time hitting their legs, prior to HSR, except they never removed it after HSR), but rather an actual fix.

Now, I pilot a spider sometimes (5V, 12 jump jets!), but I die easily whenever I make a mistake, or get caught in front of stuff, or get engaged by a Jenner, etc. I've never experienced this "Spider takes 30 AC/20 rounds to the chest and walks away unharmed" thing people say happens. Haven't experienced it when using AC/20s on Spiders, either.

Not saying it doesn't happen, but even if there is a hitbox issue, it's obviously not the only thing causing people to have a hard time killing them (and other fast mechs). There's also latency to factor in (HSR isn't perfect, and PGI did say they're still working on that, but there's only so much magic that code can do with a 280ms connection), and other potential variables.

Not to mention as others have shown, it's clearly not just something exclusive to the Spider (but of course they're going to get singled out anyway because what's more annoying than a tiny fast little pest tickling you?).

Going to be annoyed if any patches to address issues only affect the Spider, and not the other mechs also being affected by hit box issues (to their benefit or detriment). But given how they handled the Raven, I wouldn't be surprised if they didn't just fix the issue, but also inflated the hitboxes beyond the model "because it's hard to hit anyway".

Maybe I'm just being pessimistic here.

[Throe]

Here's an idea: get a buddy to pilot a Spider, and test it out on your team mate. Definitely not a great situation for anyone, but at least there's no lasting negative impact here(aside from your team mates getting a little hot over it). And it makes it easy to verify if your findings with the pilot receiving the damage(of course, you won't have any other way to verify the damage, since you can't target team mates.

Yes, I'm a Spider pilot. I'd volunteer myself, but I'm probably not going to be making any appearances in-game any time soon.

[shintakie]

Xanilos, on 04 October 2013 - 01:53 AM, said:

im pretty sure the cicada issue is one that isnt favorable to the pilot ive had an X-5 for months and ive noticed and enormously disproportionate amount of damage goes to my CT

The problem with the Cicada is the same problem every mech with a protrudin front torso has. PGI decided it'd be super duper smart to make the entire section that juts out count as the CT regardless of how incredibly screwed that made pilots of the mech. The only mech that doesn't (or didnt, dunno if they "fixed" it) was the Raven back in CBT.

The Cicada, Dragon, and Catapult are the three that suffer the worst from this problem. I will bet one of my beloved macadamia nut cookies that the Locust has the exact same issue when it gets released.

Really though, all hitboxes are screwed in MWO. I have no idea why they've never been fixed, or even put on a priority list other than the "we'd like to get to them soon, but wishy washy response that really means nothin" we've gotten before.

Its really really hard to play mechs I actually want to play when PGI insists on makin the most ridiculous hit boxes ever.

[Ensaine]

And yet..... we have a sale on these same broken toys.

[Mechthild B Steigen]

There where situations when I was hitting a spider with lasers over minutes and the armour was only orange, on ravens too (would have been enough to core an Atlas). This hit detection issue on lights is quite unpredictable, even though most of the time it works OK. Light mechs should be difficult to hit, but sometimes something is bugged.

[Accursed Richards]

I know I regularly see games where the last survivor of one team is a single Spider who'll tank the entire other team for a minute or more.

[Masterzinja]

I had something similar happen when shooting what I presumed was an afk raven with a pair of ERlarge las a few days ago in a live match. The raven was sitting at theta on tourmaline desert map and took about 16 shots dead center to it's nose from about 400m With no visable effect. I figured he must have crashed and somehow left his mech image sitting there, but then the raven came to life a few seconds after I stopped shooting him and ran off with no damage at all. No lag; other mechs shooting at me and receiving damage just fine.

[Mehlan]

Masterzinja, on 11 October 2013 - 07:24 AM, said:

I had something similar happen when shooting what I presumed was an afk raven with a pair of ERlarge las a few days ago in a live match. The raven was sitting at theta on tourmaline desert map and took about 16 shots dead center to it's nose from about 400m With no visable effect. I figured he must have crashed and somehow left his mech image sitting there, but then the raven came to life a few seconds after I stopped shooting him and ran off with no damage at all. No lag; other mechs shooting at me and receiving damage just fine.

You could have been fine, he could have been 'lagged'...

[Lugh]

White Bear 84, on 01 October 2013 - 03:28 PM, said:

We need to keep raising this in the ATD and put some pressure on. This really needs to be fixed ASAP.

It's been broken for far longer than 3 months... it'll be fixed when collision is back in the game... and you have single drop hardcore mode.

[retakrew7]

Hit detection issues are included in the known issues section since October, finally.

Seems to be that 1%.

Edited by retakrew7, 13 October 2013 - 01:54 PM.

[Oznog]

ShadowSpirit, on 01 October 2013 - 09:22 PM, said:

The issue with the treb .... damage spreads to the CT or the CT hitbox overlaps. Shoot a treb with a few LRMs and you'll see the CT blown away vastly faster than every other part.

I believe this is the case. I keep sidestepping past cover JUST far enough to get the right arm past the collision detection on the obstacle, and as I'm ducking back and can't even see, I keep getting severe CT damage. That's just crazy.

[Kmieciu]

Ballistic weapons seem more reliable than lasers right now. Some shots fail to do damage, but a few AC20 blasts will leg that Spider or outright destroy.
Lasers are so bad, that multiple blasts from 6xML don't seem to do much against a mech moving at 152 kph. But as soon as that mech stops or looses a leg, a single salvo will core him.

Edited by Kmieciu, 14 October 2013 - 01:09 AM.

[Kazly]

If its any matter, I submitted this bug to the devs, with a youtube video of the problem, back in August. Video watched, bug confirmed, and a thanks for the submission - this was on the live servers. I just didn't post the video on the forums.

The area is small, very small. People complaining of fighting a spider who's taking no damage - well, lets just say if you hit that tiny area that many times that he's just not takign damage, you're a much better shot than me. More than likely, it has to do with your/the enemy ping, or just missing.

Kaz

Edited by Kazly, 14 October 2013 - 08:44 AM.

[aniviron]

Kazly, on 14 October 2013 - 08:42 AM, said:

If its any matter, I submitted this bug to the devs, with a youtube video of the problem, back in August. Video watched, bug confirmed, and a thanks for the submission - this was on the live servers. I just didn't post the video on the forums.

The area is small, very small. People complaining of fighting a spider who's taking no damage - well, lets just say if you hit that tiny area that many times that he's just not takign damage, you're a much better shot than me. More than likely, it has to do with your/the enemy ping, or just missing.

Kaz

I didn't "just post the video on the forums;" I emailed support, of course.

There's clearly a pretty big problem with this. There are quite a few similar threads to mine where you can watch people getting 3-4 dead-on hits with AC20 or massive SRM clouds that are all fire and explosions but leave the spider completely unscathed.

[Kazly]

aniviron, on 14 October 2013 - 04:34 PM, said:

I didn't "just post the video on the forums;" I emailed support, of course.

There's clearly a pretty big problem with this. There are quite a few similar threads to mine where you can watch people getting 3-4 dead-on hits with AC20 or massive SRM clouds that are all fire and explosions but leave the spider completely unscathed.

I agree, a mech not taking damage is a serious problem, and even though I primarily play the spider/jenner, I still submitted the ticket.

And don't take offense to what i said please - there is a connotative difference between "I didn't just post" and "I just didn't post" - one, I would read as an accusation, the other more a simple statement of my actions.

Feel free to re-read which one I posted I wasn't accusing you of anything

Kaz

[Fragger56]

Im putting this Spider Hitreg Analysis post here for better viewage, since the other thread tanked on views once I interjected logic into the OMG SPIDERS OP convo...


"I've just done some math and figured out how large the Spider is supposed to be and from this, it seems pretty obvious why spiders "take little damage"

A spider at full speed can move at up to 150+ KPH, at 150 KPH the mech is moving at a rate of 41.6 meters per second.

Now the rate of travel in M/S is important, as once you factor in ping, you should realise that the spider, on the server end can be a significant distance away from the client POV.

In my case, I typically have a ping of 50ms.
Within 50ms, a spider at 150KPH will have traveled 2.08 meters, now some of you might think that 2 meters isn't much, but a spider is only supposed to be around 10m tall.
2m on a spider is the difference between hitting the CT and the RT, or hitting the RT and missing to the side.


Any game with any amount of latency and server confirmed or hybrid client/server hitreg will have offset issues like this, its the nature of the internet and latency. In most games the targets and hitboxes are large enough that the offset caused by latency should be negligible, but in MWO we have the spider, which can cover a distance larger than one of its hitboxes in the time covered by a typical players ping.

As your ping goes up, this situation becomes worse, and spiders can become nearly impossible to hit, as at 150ms, there would be nearly a full spiders width worth of offset between where your client thinks it is and where the server says it is.

This issue can also be mitigated by using weapons with FAST projectile speeds (Gauss) and a small amount of lead.

I've gotten to the point that I can usually hit the exact component I want to on a fast spider as i aim for the next component over to compensate for the offset.

Example:
Spider is moving from right to left at max speed 300m in front of you, with gauss and a 50ms ping, if i aim at his RT (left from my POV) I'll get hitreg on his CT, if i aim CT, it will hit his LT (RT from my POV) and if I aim at his LT, the shots will likely miss his arm completely cause its so thin.

BTW, I think the Locust will be even worse than the spider when it comes to hitreg once it is released, due to the conclusions drawn from that math up above.
Locust will be smaller than the spider and move faster, leading to even more offset which will probably lead to more 'ZOMG LIGHTMECHS' threads..."