338 replies to this topic

[Ozric]

Wait, so I am as good in a Spider as I think I am?

[Hythos]

yashmack, on 17 October 2013 - 11:12 PM, said:

This happens not just with Spiders but every mech at some point, its not a hitbox issue its an issue with hit registration and with host state rewind...

The Spider was standing perfectly still. Absolute clear LOS. One weapon: ERLarge LASER. One full cool-down cycle. Two shots. Dead center, zoomed in @ 600m. No damage. That's not HSR.

[Hythos]

** Edit - additionally, I understand / know that it's not *just* with Spiders. However, the conditions *I* experience are different: Hitting other units and seeing a red reticle flash - but see no damage (HSR? Hit Detection?)
While (so far) only seeing what should be a hit but no reticle flash and no damage (hit-box geometry) as what I would expect was the issue with me shooting at the Spider last night.

[Gallowglas]

What I find confusing is that PGI publicly acknowledged the spider hitbox problem a few months ago and we've heard nothing else about it. Now, suddenly, we're seeing Garth say something else in an informal conversation. I think I'm probably not alone in feeling a bit bewildered. I was under the assumption that since this had been publicly acknowledged as a problem that it was slated to be actively addressed. Why can't we get clear communication on these issues, particularly when there's a complete about-face on stance?

Honestly, I've seen video evidence of people registering big damage on Spiders with their aiming reticle well ahead of the mech, while only getting a paper doll flash while putting reticle on target. I've personally experienced dropping three AC/20 rounds dead on target of a shut-down Spider. Does it happen all the time? No. However, can you seriously tell me that you believe that Spider pilots just happen to be better than other pilots to the extent that there's a logical reason why they can run around a group of assaults with impunity and consistently be the last mechs standing in a match while Jenner and Raven pilots can't do the same?

Is the problem crippling? No. Is it consistent? No. But to deny there's any problem is bewildering if you've played the game for any length. I'm never, ever one to jump in on public hype over something, but I strongly believe there's some amount of fire behind this smoke.

Edited by Gallowglas, 18 October 2013 - 10:13 AM.

[Mehlan]

Quote

What I find confusing is that PGI publicly acknowledged the spider hitbox problem a few months ago and we've heard nothing else about it.

Really? LINK it...

Quote

The Spider was standing perfectly still. Absolute clear LOS. One weapon: ERLarge LASER. One full cool-down cycle. Two shots. Dead center, zoomed in @ 600m. No damage. That's not HSR.

How so? Did you program the system? Do you think it turns it'self on and off based on the mech movement state?

[Deathlike]

Gallowglas, on 18 October 2013 - 09:19 AM, said:

What I find confusing is that PGI publicly acknowledged the spider hitbox problem a few months ago and we've heard nothing else about it. Now, suddenly, we're seeing Garth say something else in an informal conversation. I think I'm probably not alone in feeling a bit bewildered. I was under the assumption that since this had been publicly acknowledged as a problem that it was slated to be actively addressed. Why can't we get clear communication on these issues, particularly when there's a complete about-face on stance?

The thing is that communication between PGI members generally happens to be... less than stellar.

Frankly I don't think PGI devs play their own game enough (technically Paul should be #1 on this, because he "balances" the game) to know what's working and what's not.

Although, considering that I've seen certain PGI devs play with Spiders, I'm starting to think they don't realize whatever special "Spider sauce" they they may have on for their own mechs is translating to everyone else driving a Spider.

Edited by Deathlike, 18 October 2013 - 09:46 AM.

[Hythos]

Mehlan, on 18 October 2013 - 09:34 AM, said:

Really? LINK it...

How so? Did you program the system? Do you think it turns it'self on and off based on the mech movement state?

Wow. Not sure if you're just trolling.... Well, I can respond with the definition of "Host State Rewind". If the unit's location is 'rewound' to a previous location in 4 dimensions (X, Y, Z, and TIME) to when the shot was fired... it is considered as having been there all along. Now, if the target doesn't move from before the time I saw it, until the time that the first weapon-cycle occurs, followed by cool-down, and then a second weapon-cycle... Host State Rewind will have accounted for the unit standing still.
Because this time-frame exceeds HSR's active time-cycle, there should have been some appreciable damage...... unless, HSR is actually what, 5+? seconds? IS that what you're suggesting? Otherwise, comment is troll.

** Edit - my deduction would suggest that the conditions we're experiencing with the Spider is NOT related to HSR, but rather hit-box, hit-detection, or 'applying damage' which is apparently independent of the other two: (Ref:

Quote

Bumping damage is going to help deal a small amount more damage to small Mechs, but the larger Mechs are going to be destroyed VERY quickly."

)

Edited by Hythos, 18 October 2013 - 10:12 AM.

[Gallowglas]

Mehlan, on 18 October 2013 - 09:34 AM, said:

Really? LINK it...

Though it doesn't specifically mention Spiders, here's the quote where they mention the problem. Granted, they don't say "hitbox", but they do say it impacts smaller mechs more than bigger mechs. Note that they say the size rather than the speed impacts the registration problem. It's possible it's related to something other than the Spider hitbox specifically, but that chassis definitely exhibits the problem. Maybe because it's one of the smallest mech chassis.

Quote

Many of you have cited hit detection errors. We’re seeing this as well. While it happens across almost all Mechs, it’s most noticeable with small Mech chassis. Bumping damage is going to help deal a small amount more damage to small Mechs, but the larger Mechs are going to be destroyed VERY quickly. We are investigating the root cause of these detection errors but it’s a deep problem. We need to find out if it’s in HSR (host state rewind) or is it in the simulation running on the server etc etc. Once found we will be pouring heavy resources on to the problem to fix it ASAP. However, I can tell you this, it’s going to take a while to find due to its complexity.

http://mwomercs.com/...general-update/

Edited by Gallowglas, 18 October 2013 - 10:06 AM.

[Kenyon Burguess]

Captain Midnight, on 17 October 2013 - 02:14 PM, said:

There is no universal hitreg problem, there are hitreg problems on SPECIFIC mechs, namely the spider. I can hit jenners, ravens, and locusts no problem. The exact same shots miss against spiders.

yesterday I took an ac/20 to the back area in my locust (p). the screen did the mega-shake and I ran into a building in the confusion (don't judge me dammit, im not used to the speed yet) when I refocused I was undamaged armor (99%). its not just spiders imo, its all things small and fast getting hit by things that use aoe explosions.

[Mehlan]

Quote

Though it doesn't specifically mention Spiders, here's the quote where they mention the problem. Granted, they don't say "hitbox",

That is, precisely why I stated 'link it'

Quote

What I find confusing is that PGI publicly acknowledged the spider hitbox problem

This claim that PGI acknowledged a HITBOX problem.

Quote

Wow. Not sure if you're just trolling.... Well, I can respond with the definition of "Host State Rewind". If the unit's location is 'rewound' to a previous location in 4 dimensions (X, Y, Z, and TIME) to when the shot was fired... it is considered as having been there all along. Now, if the target doesn't move from before the time I saw it, until the time that the first weapon-cycle occurs, followed by cool-down, and then a second weapon-cycle... Host State Rewind will have accounted for the unit standing still.
Because this time-frame exceeds HSR's active time-cycle, there should have been some appreciable damage...... unless, HSR is actually what, 5+? seconds? IS that what you're suggesting? Otherwise, comment is a troll

Was not, the comment that I address/referred to..stating it was not/could not be an HSR issue? The flaw in your logic above and the statement in saying it could not be an HSR issue, is assuming it's working like you say it should ^.

[Clit Beastwood]

Geist Null, on 18 October 2013 - 10:10 AM, said:

yesterday I took an ac/20 to the back area in my locust (p). the screen did the mega-shake and I ran into a building in the confusion (don't judge me dammit, im not used to the speed yet) when I refocused I was undamaged armor (99%). its not just spiders imo, its all things small and fast getting hit by things that use aoe explosions.

Yup - took my YLW 5 AC20 hits to kill a legged locust. As is my norm now, I hit TAB and wrote down the ping and the number of shots. There's definitely a connection between latency and hit registration. sub-100ms pings and locusts come apart easy. 300-400 ping, locusts are beastly to kill with autocannons. Lasers seem to be less affected by the pings.

[Gallowglas]

Mehlan, on 18 October 2013 - 10:18 AM, said:

That is, precisely why I stated 'link it' This claim that PGI acknowledged a HITBOX problem.

Nitpick word choice if you like, but to me and I am sure many others, when you say that SIZE and not SPEED are the problem, I call that a hitbox problem. Maybe that's a misnomer, but I don't have anything else to call it nor does anyone have any other technical details to further clarify. And maybe that's the source of the confusion between what Garth said and what the earlier post indicates. If that's the case, surely someone could acknowledge and clear up that the latter is still acknowledged as a problem. Don't just say "there's nothing to fix". That's a surefire way to cause confusion and misunderstanding.

In short, if there's a problem which exists in game that impacts our ability to hit Spiders, then there's something to fix, even if it's not in the purest technical sense specifically the Spider's hitbox. You should know the gist of what people are talking about and respond accordingly.

Edited by Gallowglas, 18 October 2013 - 10:39 AM.

[Bagheera]

Roadbeer, on 18 October 2013 - 04:39 AM, said:

So, in your world, the chat conversation should have gone like this,

Garth: "Ok, I was here to play the game, but allow me to take my time to explain to these petulant children in chat why they are mistaken. The other 21 of you don't mind that I take the whole 15 minutes..."
Garth: .."to do this do you? Because I can tell by their responses that they are going to be very receptive to anything I have to say."

A simple "______ is the problem, spider is a symptom." would have been just fine, short, to the point, and an actual answer. Saying "there is nothing to fix" regarding any topic of this game is just, well, incorrect.

Roadbeer, on 18 October 2013 - 04:39 AM, said:

As you say Whatever, There is a time and a place for those questions, and in the middle of a match isn't one of them. Bad form for asking it there, and REALLY bad form at the reaction, to not getting the answer he wanted. I picture a 5 year old with tourets who was just told that he can't have a piece of candy.

Actually, employees of a company that deals with the public should be ready and able to at least field polite stock responses anytime they interact with said public. It's not rocket science. Sure, the reaction was terrible, but the people reacting aren't paid employees of a company in communication with their customers. That doesn't excuse them, but it really doesn't excuse our man from dodging the question either. Like I said, the saloon in that there post-modern western is full of bad guys.

Edited by Bagheera, 18 October 2013 - 10:50 AM.

[Slab Squathrust]

Fierostetz, on 18 October 2013 - 10:25 AM, said:

Yup - took my YLW 5 AC20 hits to kill a legged locust. As is my norm now, I hit TAB and wrote down the ping and the number of shots. There's definitely a connection between latency and hit registration. sub-100ms pings and locusts come apart easy. 300-400 ping, locusts are beastly to kill with autocannons. Lasers seem to be less affected by the pings.

I don't use too much in the way of ballistics so I can't necessarily compare them to lasers. However just moments ago I was figthing a spider in a 6ml jenner. I was getting hit confirmations and the paperdoll would flicker but you saw very little change. Most of my hits did nothing even when getting solid connection for ~75% of the beam duration. However one finally registered and the leg went from yellow armor to orange internals. A second laser burn takes it off. Mind you, this spider was running around a group of heavies and assaults from my team and taking fire from some of them. Finally after the leg loss, I alphaed him in the back twice with 0 effect. He lasted about 8 seconds surrounded by enemies. Atlases fall faster than that spider.

Interestingly the ping was approximately 150. Not terrible. It was just frustrating how wildly inconsistent those shot fell.

[Clit Beastwood]

Slab Squathrust, on 18 October 2013 - 11:06 AM, said:

I don't use too much in the way of ballistics so I can't necessarily compare them to lasers. However just moments ago I was figthing a spider in a 6ml jenner. I was getting hit confirmations and the paperdoll would flicker but you saw very little change. Most of my hits did nothing even when getting solid connection for ~75% of the beam duration. However one finally registered and the leg went from yellow armor to orange internals. A second laser burn takes it off. Mind you, this spider was running around a group of heavies and assaults from my team and taking fire from some of them. Finally after the leg loss, I alphaed him in the back twice with 0 effect. He lasted about 8 seconds surrounded by enemies. Atlases fall faster than that spider.

Interestingly the ping was approximately 150. Not terrible. It was just frustrating how wildly inconsistent those shot fell.

When I rambo through a group in my spider I'm usually whipping my torso around all crazy to spread damage out, can't do it with legs.

[Asakara]

Gallowglas, on 18 October 2013 - 10:02 AM, said:

Though it doesn't specifically mention Spiders, here's the quote where they mention the problem. Granted, they don't say "hitbox", but they do say it impacts smaller mechs more than bigger mechs. Note that they say the size rather than the speed impacts the registration problem. It's possible it's related to something other than the Spider hitbox specifically, but that chassis definitely exhibits the problem. Maybe because it's one of the smallest mech chassis.



http://mwomercs.com/...general-update/

That paragraph you pulled out of context was regarding SRMs. Nowhere in those 3 paragraphs, that you pulled the middle one out of, does it mention other weapons or say it was not exclusive to SRMs.

Pretending otherwise is disingenuous in my book.

Full Context:

Quote

What about SRMs?!?! Buff them to 2.5!!!

Nuh uh. 2.5 SRM damage causes the same effect as the previous LRM-aggedon. While funny to test, the 6-SRM6 Catapult will decimate any Assault class Mech in 3 volleys. The third volley doesn’t even have to be a full volley, 2 volleys following up with a medium laser will probably kill most builds in the game. So what am I going to do about this? Well let me explain an issue…

Many of you have cited hit detection errors. We’re seeing this as well. While it happens across almost all Mechs, it’s most noticeable with small Mech chassis. Bumping damage is going to help deal a small amount more damage to small Mechs, but the larger Mechs are going to be destroyed VERY quickly. We are investigating the root cause of these detection errors but it’s a deep problem. We need to find out if it’s in HSR (host state rewind) or is it in the simulation running on the server etc etc. Once found we will be pouring heavy resources on to the problem to fix it ASAP. However, I can tell you this, it’s going to take a while to find due to its complexity.

That being said, here’s an opportunity for you to decide your fate when dealing with SRMs. In this feedback thread (http://mwomercs.com/...-hit-detection/), there is a poll. You can vote to bump SRM damage up to 2.0, not this patch but next, or leave damage as is until hit detection is fixed. Keep in mind, this bump in damage is going to be a fairly big world of hurt for larger Mechs. NOTE: Cut off for voting will be Sunday 11:59PM PST.

Edited by Asakara, 18 October 2013 - 11:27 AM.

[Hythos]

Mehlan, on 18 October 2013 - 10:18 AM, said:

That is, precisely why I stated 'link it' This claim that PGI acknowledged a HITBOX problem.



Was not, the comment that I address/referred to..stating it was not/could not be an HSR issue? The flaw in your logic above and the statement in saying it could not be an HSR issue, is assuming it's working like you say it should ^.

Yes, that is true; it is my assumption that PGI is either honest, or correct, when they explain how it works... it is entirely possible that it does not work at all or is simply 'vaporware', considering certain track-records...

[Gallowglas]

Asakara, on 18 October 2013 - 11:26 AM, said:

That paragraph you pulled out of context was regarding SRMs. Nowhere in those 3 paragraphs, that you pulled the middle one out of, does it mention other weapons or say it was not exclusive to SRMs.

Pretending otherwise is disingenuous in my book.

I can see some room for debate about whether the hit registration issues referenced are isolated to SRM's. That really wasn't clarified and the point of the dev post about regressive impact of SRM damage on bigger mechs remains unchanged regardless of whether it was or it wasn't. That said, either you misunderstand the meaning of the word "disingenuous" or I'd have to be really stupid to source the quote.

Edited by Gallowglas, 18 October 2013 - 11:58 AM.

[Ngamok]

Corwin Vickers, on 17 October 2013 - 09:33 PM, said:

You need to aim ahead of the spider Homeless Bill, not behind them. You'll do much better.


All you older players should remember back when ballistics took like half a second to fire. That is because in a server authoratative game you send your desire to shoot to the server, it takes your request (for guys with 250 ping) 1/4th of a second to get there. The server does the calc and sends the result back to you. Half a second has passed before you see the results of your shot.

That pissed everyone off so they fixed it. Except they didn't fix it, they just gave you a simulation on your side that is sometimes right, and sometimes wrong. You still send your desire to shoot to the server which takes a 1/4 second. And then the server sends back the result to you, but you've already seen what happened, whether it was what really happened or not so the only thing that changes is the paper doll of the enemy mech. This is probably also why it takes a long time for paper dolls to update sometimes (also causing claims of invincible light mechs.)

As to exactly how HSR does it's prediction, that is a mystery but it seems to throw a monkey wrench into predicting where high ping enemies really are.


You guys should spend more time complaining to your ISP to get better ping than complaining here about spiders. One of the guys I'm playing with right now found out that his connection is going to Europe before coming back to the US. That's on Comcast.

Not really. It's like Call of Duty did to try to fix it. People with low ping were given a penalty to compensate. My ping here is between 36 to 42 usually. Maybe this HSR is penalizing me and can't predict things right. I don't know, but I hit spiders sometimes square in the nuts and they don't even flinch. Not always though. Why are commandos easier to hit and register damage than spiders? Could be that the commando has an older hit box configuration. Both models are almost identical in size.

[Asakara]

Gallowglas, on 18 October 2013 - 11:56 AM, said:

I can see some room for debate about whether the hit registration issues referenced are isolated to SRM's. That really wasn't clarified and the point of the dev post about regressive impact of SRM damage on bigger mechs remains unchanged regardless of whether it was or it wasn't. That said, either you misunderstand the meaning of the word "disingenuous" or I'd have to be really stupid to source the quote.

I do not see any room for debate on if that quote you pulled out of context was about SRMs or not. The title of the section even says "What about SRMs?!?! Buff them to 2.5!!!". In my mind all 3 paragraphs were about SRMs, not all weapons, not Spiders, etc..

As for understanding my use of "disingenuous", I was following its definition of "not sincere". Sincere meaning "free from deceit" (hopefully I do not need to define deceit too). I apologize for any confusion in my usage, and perhaps next time I will just say that "Pretending otherwise is deceitful in my book"... Though personally I find that less tactful.

Edited by Asakara, 18 October 2013 - 12:37 PM.