Jump to content

Hit Detection: Broken Again.


37 replies to this topic

#21 PanzerMagier

    Member

  • PipPipPipPipPipPipPipPip
  • Big Daddy
  • Big Daddy
  • 1,369 posts
  • Twitter: Link
  • Twitch: Link
  • LocationSome nameless backwater planet

Posted 18 September 2013 - 10:48 PM

I am a player with average 250 ping. HSR worked PERFECTLY right up until they implemented it for missiles. That broke everything. And it's just as bad as it was back then. I'll even go as far as to say that I had better hit registration before Hsr was implemented at all!

#22 GRiPSViGiL

    Member

  • PipPipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 1,904 posts
  • Twitch: Link
  • LocationHillsboro, OR

Posted 18 September 2013 - 11:17 PM

I agree, the hit detection has not felt solid at all. I see shots register hits I know I missed with and when I am center mass nothing.

I especially like when I can see a targetable portion of a mechs side or arm next to terrain or some building only to magically have something block the exposed part of the mech. This occurs at range when my weapons would be able to converge to the spot and seemingly not have anything obstructing projectile path.

#23 Nauht

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 2,143 posts

Posted 18 September 2013 - 11:22 PM

What a joke this is still an issue at launch.

When the reviews come in from the bigger gaming sites I'm gonna make sure they know about this fundamental problem.

A shooter where your shots dont register. Ha!

#24 Sgt Helmet

    Member

  • PipPipPipPipPip
  • Knight Errant
  • 101 posts

Posted 18 September 2013 - 11:39 PM

Check you ping at first drop after you login and then in the second drop.

For me the first drop is around 200ms (prepatch around 150ms) and in second drop and onwards 600ms+.

HSR seems to go offline at that point. Hits do not register on moving target at all and on stationary sometimes. Movement, target acquisition and zooming has a huge delay. Basically second drop is unplayable, did restart of the client few times and same thing happened every time.

First drop has no issues at all. Hits register ok and everything else is fine.

There's a feedback thread for this: http://mwomercs.com/...feedback-12242/
So if you observe this, fill the poll there and make a post. Or have any other issues. I trust that devs are checking those out as they direct us to give feedback there.

Sadly this patch rendered the game unplayable for me, so I'm hoping for a quick resolution on this issue.

Edited by Sgt Helmet, 18 September 2013 - 11:41 PM.


#25 Elfman

    Member

  • PipPipPipPipPipPip
  • 202 posts
  • LocationUK

Posted 19 September 2013 - 12:19 AM

SRM's for me seem to have the biggest problem unfortunatly most of the mechs I drive use them. Seems the only way i can get reliable damage (300+) these days is by using slighly cheesy dual AC/20 or CLBX/10 builds.

Also have noticed that Artemis seems to not be having hardly any affect on SRM's and isnt worth its weight or slots.

#26 Alistair Winter

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Storm
  • Storm
  • 10,823 posts
  • LocationBergen, Norway, FRR

Posted 19 September 2013 - 01:27 AM

Unfortunately, I'm not even confident that separate servers for Europeans and Australians / Asians would solve this problem.

So yeah...

<unequips SRMs, equips lasers>

#27 Cola

    Member

  • PipPipPipPipPip
  • 171 posts
  • LocationSheridan

Posted 19 September 2013 - 01:31 AM

I too feel that Hit detection has once again failed with a patch. I shot a shut down jenner 3 times, in what I thought and looked like his CT, 2LL/AC10 was getting hit markers. and He walked away from 80 points of poentinal DMG with only light yellow all around.

#28 Zeee

    Member

  • Pip
  • 14 posts

Posted 19 September 2013 - 02:25 AM

i concur with u guys.....was just in a match, i released 7 bloody salvos of 5xSRM6 together with three 5xMLs and when i got blasted down it only showed as 55 BLOODY F%$&^%$^G DAMAGE!!!

WTF!!!

Edited by Zeee, 19 September 2013 - 02:25 AM.


#29 Nightfire

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 226 posts
  • LocationAustralia

Posted 19 September 2013 - 02:43 AM

What do you mean broken AGAIN!?
It's still broken, only a little more than is used to be.

Apparently 3 volleys of 6x Medium lasers on a stationary Boom Jager is 33 points.
1 volley from him of 2x AC20 is enough to instantly destroy my barely touched torso and engine.

I still have to lead fast mechs by as much as 2 mech lengths.

No, Hit Detection has never really worked properly. It just varies in how bad it is.

#30 Billygoat

    Member

  • PipPipPipPipPipPip
  • Survivor
  • Survivor
  • 298 posts

Posted 19 September 2013 - 03:22 AM

The most frustrating thing about it for me is that I can't even compensate by lag shooting at fast mechs.

The hit detection is so spotty and inconsistent, it's usually not even worth shooting at fast moving lights at all unless they're running directly toward or away from me. There's simply no way to predict where the hitboxes are at any given moment, which is even worse in some ways than it was before HSR was introduced. At least back then, I knew roughly where to put my lasers to hit them by factoring in latency, but that hasn't worked for months. Nevermind that even when I'm lucky to actually score a "hit" (ie. the crosshair goes red) with a large weapon like an AC/20, only 50% of the time is the damage even applied at that point.

We've all got stories about pegging Spiders with a big gun, seeing the explosion, seeing the crosshair turn red, only to have it keep on running with yellow armour on whatever section you struck.

I'd just love to know exactly what PGI did to screw it up so badly since the very good initial Laser HSR patch and the not-as-good-but-still-serviceable ballistic HSR patch. Did applying it to missiles REALLY wreck all that previous good work? If that's the case, maybe it's time to roll missile HSR back and let us deal with firing delay on SRMs again. It's not like they're actually registering hits at the moment anyway.

#31 DaZur

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 7,511 posts
  • LocationWisconsin

Posted 19 September 2013 - 04:58 AM

While I'm not going to say HSR is not borked a little... I think some of you guys are dead-set to blame the farmer for the taste of the milk from the cow.

HSR is a work-around (attempt to mitigate) player-induced lag... Not a definitive solution. If you really think HSR can perfectly mitigate pings greater than 200ms (Not to mention 400+ pings) you've not played enough online gaming... ;)

Net speed is only one aspect of line quality... packet loss, latency and jitter is also of equal consideration, for gaming sometimes even more important that pure speed and ping.

For grins and giggles... try this broadband test site: http://www.megapath.com/speedtestplus/

I'm now curious how folks line quality measures out...

Edited by DaZur, 19 September 2013 - 05:27 AM.


#32 Billygoat

    Member

  • PipPipPipPipPipPip
  • Survivor
  • Survivor
  • 298 posts

Posted 19 September 2013 - 05:17 AM

View PostDaZur, on 19 September 2013 - 04:58 AM, said:

While I'm not going to say HSR is not borked a little... I think some of you guys are dead-set to blame the farmer for the taste of the milk from the cow.

HSR is a work-around (attempt to mitigate) player-induced lag... Not a definitive solution. If you really think HSR can perfectly mitigate pings greater than 200ms (Not to mention 400+ pings) you've not played enough online gaming... ;)


The problem isn't that everybody thinks it will be 100% perfect even with high ping like mine. I'm playing from the other side of the world, I know it's never going to be perfect.

That said, when PGI first implemented HSR for lasers and then for ballistics, the results were excellent (lasers) to fair (ballistics/PPCs). This coming from somebody who plays from Australia. I was singing PGI's network developers' praises because those initial implementations were some very good work in less than ideal circumstances (ie. project management's poor choice of engine).

The problem is (at least for me) that now, the quality of the hit detection and damage registration has regressed significantly since then. Missiles have never worked right, though I'm prepared to accept that they would be the most technically difficult to deal with. However, the fact that both laser and ballistic hit detection in general have both degraded tangibly since then leads me to believe that there is something else PGI has done in the mean time that has affected it. Something that can have a technical solution because it has been much better in the past than it is now.

Edited by Billygoat, 19 September 2013 - 05:19 AM.


#33 DaZur

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 7,511 posts
  • LocationWisconsin

Posted 19 September 2013 - 05:26 AM

View PostBillygoat, on 19 September 2013 - 05:17 AM, said:

The problem is (at least for me) that now, the quality of the hit detection and damage registration has regressed significantly since then. Missiles have never worked right, though I'm prepared to accept that they would be the most technically difficult to deal with. However, the fact that both laser and ballistic hit detection in general have both degraded tangibly since then leads me to believe that there is something else PGI has done in the mean time that has affected it. Something that can have a technical solution because it has been much better in the past than it is now.

While pure speculation my part... I'm coming to the conclusion that as they added more weapon categories to the HSR correction work-load, it became far more sensitive to line quality and as a result, dependent upon the number participants with poor line quality / speed... the accuracy of the HSR correction becomes more-and-more flaky.

That said... it boils down to tuning the HSR to handle a wider range of inequity. Will it be perfect?... Never. And any expectation of it being perfect is an irrational expectation IMHO... No offense to anyone with a less than stellar internet connection.

Edited by DaZur, 19 September 2013 - 05:26 AM.


#34 Billygoat

    Member

  • PipPipPipPipPipPip
  • Survivor
  • Survivor
  • 298 posts

Posted 19 September 2013 - 05:34 AM

View PostDaZur, on 19 September 2013 - 05:26 AM, said:

While pure speculation my part... I'm coming to the conclusion that as they added more weapon categories to the HSR correction work-load, it became far more sensitive to line quality and as a result, dependent upon the number participants with poor line quality / speed... the accuracy of the HSR correction becomes more-and-more flaky.

That said... it boils down to tuning the HSR to handle a wider range of inequity. Will it be perfect?... Never. And any expectation of it being perfect is an irrational expectation IMHO... No offense to anyone with a less than stellar internet connection.


Yeah, you might be onto something there. Either the system as they designed it just buckles under the pressure of having to deal with the extra categories or it could even be that the hardware IGP is providing for them is underpowered.

The other possibility is that the quality of the hit detection has declined because of factors outside IGP and PGI's control. That is, the wider network around the data centre where they have their servers. I know the quality of my connection to the MWO game servers improved markedly (along with improvements to gameplay) when I started using a VPN to route around a couple of Sprint/Tinet nodes on my previous route that were lagging hard and dropping packets like crazy.

#35 MaddMaxx

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 5,911 posts
  • LocationNova Scotia, Canada

Posted 19 September 2013 - 06:08 AM

View PostBillygoat, on 19 September 2013 - 05:34 AM, said:


Yeah, you might be onto something there. Either the system as they designed it just buckles under the pressure of having to deal with the extra categories or it could even be that the hardware IGP is providing for them is underpowered.

The other possibility is that the quality of the hit detection has declined because of factors outside IGP and PGI's control. That is, the wider network around the data centre where they have their servers. I know the quality of my connection to the MWO game servers improved markedly (along with improvements to gameplay) when I started using a VPN to route around a couple of Sprint/Tinet nodes on my previous route that were lagging hard and dropping packets like crazy.


And also add in the fact that we are 24 on the field now, not 16 and required calculations and data transferred increases, likely, exponentially. ;)

#36 Bromineberry

    Member

  • PipPipPipPipPipPip
  • 436 posts

Posted 19 September 2013 - 06:10 AM

View PostImperius, on 18 September 2013 - 03:32 PM, said:

I'm all for segregation via ping. I'd never vs mr 10k ping from Brazil again. That was a super bad match... Anyway it should go like this.

0~50
50~100
100~150
150~200
200+ (enjoy your game)

This is how ELO should work my ping is 40~60 avg so I really don't think it's fair I get paired up with anything above my tier or close. If you lag you should play with people that lag like you. That simple.



This is a horrible idea. First of all, I don't think that having a higher ping in the 100-200 range (as long as it isn't 1000+) has ever been very advantageous or disadvantageous. At least I never noticed anything like that.
Second, with a "ping splitting" like you suggest, you're simply screwed if you play at an unusual time in your timezone, because the ping difference is mostly due to geographic location. Someone from central Europe will always have a ping around 150ms. So, if it is night and harldy anyone plays in Europe. So, the player won't find a game, because there are hardly people in his ping range playing. This would be annoying ,especially because the player knows, that in the US there a re lots of players he could play with....but because his ping is 100ms slower, he is excluded from the game. I'm sure this player will happily pay for anything in MWO...


But, I have to agree: Hit detection is quite bad at the moment. Yesterday I blasted a Spider which first did a bee line towards and then away from me in the face/back with 4 LL. Nothing happend. In another game,one was standing still and I fired at the little ******...no damage registerd.

Edit:
bugg_r is considered a curseword and therefore is censored?

Edited by Bromineberry, 19 September 2013 - 06:14 AM.


#37 -Muta-

    Member

  • PipPipPipPipPipPipPip
  • The 1 Percent
  • 749 posts
  • Locationstill remains a mistery.

Posted 19 September 2013 - 06:12 AM

View PostTisun, on 18 September 2013 - 12:20 PM, said:

I have a theory that when they put in the new movement system it screwed up HSR. More specifically, HSR might not have been designed to predict the new movement code. I don't have any concrete evidence besides for them implementing the movement code after HSR, and that shooting at things with JJs usually leads to a miss.


I have to disagree on this one because I find way easier to hit flying targets and ofc causing the respective damage.

#38 Mehlan

    Member

  • PipPipPipPipPipPip
  • 481 posts
  • LocationTx

Posted 19 September 2013 - 06:17 AM

Quote

I have to disagree on this one because I find way easier to hit flying targets and ofc causing the respective damage.
Probably has to do more with how JJ's are...coded. Take any JJ mech, when they activate the jets, they cannot/deviate from the path until they land... it's a straight line only.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users