Jump to content

Paging Karl Berg...karl Berg, Please Pick Up The White Courtesy Phone...


1911 replies to this topic

#501 Mawai

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 3,495 posts

Posted 24 April 2014 - 10:58 AM

Hi Karl,

Apparently some additional bugs have been found in the hit registration code with the potential to have a significant impact on SRM's (Paul NGNG podcast).

However, unless there is different hit registration code for every weapon (which seems a bit unlikely), is it possible that the bug will affect multiple weapons?

For example, as far as I know, laser fire is resolved in several 'pulses' over time. Could the hit registration of these pulses also be affected by these bug fixes for SRMs? (One reason I ask, is that there are times when laser fire appears to be less effective than I expect it should be ... i.e. laser is held on target (which isn't moving that much) but doesn't seem to do as much damage as expected ... of course that is all subjective but there could be laser hit registration effects also involved ... perhaps especially when multiple lasers are fired simultaneously).

Anyway, I was wondering if there are any details on the nature of the bug and the scope of possible impact.

#502 TheCaptainJZ

    Member

  • PipPipPipPipPipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 3,658 posts
  • LocationUnited States

Posted 24 April 2014 - 11:10 AM

View PostMawai, on 24 April 2014 - 10:58 AM, said:

Hi Karl,

Apparently some additional bugs have been found in the hit registration code with the potential to have a significant impact on SRM's (Paul NGNG podcast).

However, unless there is different hit registration code for every weapon (which seems a bit unlikely), is it possible that the bug will affect multiple weapons?

For example, as far as I know, laser fire is resolved in several 'pulses' over time. Could the hit registration of these pulses also be affected by these bug fixes for SRMs? (One reason I ask, is that there are times when laser fire appears to be less effective than I expect it should be ... i.e. laser is held on target (which isn't moving that much) but doesn't seem to do as much damage as expected ... of course that is all subjective but there could be laser hit registration effects also involved ... perhaps especially when multiple lasers are fired simultaneously).

Anyway, I was wondering if there are any details on the nature of the bug and the scope of possible impact.


iirc, one of the major bugs with srms that is getting fixed is an issue with how cryengine handles explosions. it wasn't simulating all the appropriate explosions within the same frame (some sort of cap on it) and some were being delayed til the next frame at which point the target would have moved.

#503 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 24 April 2014 - 02:18 PM

View PostKlappspaten, on 24 April 2014 - 07:02 AM, said:

I would say Karma ^^

But its still ture, I have expierienced that too.
Sometimes the missle turrets fire at me when I'm not even in the 450 meter radius and another Mech has triggered the turret.
I found a pretty good example of it, right here.

There was at least one, possibly two mechs that were in front of me, line of site but the missile turret targeted me.



Anyway, I'm guessing the game is prioritizing turret targets by whom ever is direct line of site, closest, and has the highest damage dealt that round.

Edited by Dimento Graven, 24 April 2014 - 02:20 PM.


#504 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 24 April 2014 - 02:57 PM

View PostDimento Graven, on 24 April 2014 - 02:18 PM, said:

Anyway, I'm guessing the game is prioritizing turret targets by whom ever is direct line of site, closest, and has the highest epeen that round.

Fixed that for you :lol:

#505 TheCaptainJZ

    Member

  • PipPipPipPipPipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 3,658 posts
  • LocationUnited States

Posted 24 April 2014 - 03:04 PM

It may not "see" the other friendlies ahead of you if it's targeting, say, the legs of mechs. Could just be a targeting issue from the turret's perspective.

#506 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 24 April 2014 - 03:05 PM

View PostCimarb, on 24 April 2014 - 02:57 PM, said:

Fixed that for you :lol:
Well for some, that's quite a sizable target, no doubt. :lol:

#507 aniviron

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,752 posts

Posted 24 April 2014 - 03:48 PM

View PostDimento Graven, on 24 April 2014 - 02:18 PM, said:

I found a pretty good example of it, right here.

There was at least one, possibly two mechs that were in front of me, line of site but the missile turret targeted me.



Anyway, I'm guessing the game is prioritizing turret targets by whom ever is direct line of site, closest, and has the highest damage dealt that round.


Another thought: given the logic on how turrets target components with lasers, the LRM turrets might just choose the target with the lowest health/health percentage within their entire max firing distance of 1km and hit that.

#508 Deathlike

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 29,240 posts
  • Location#NOToTaterBalance #BadBalanceOverlordIsBad

Posted 24 April 2014 - 04:08 PM

View PostDimento Graven, on 24 April 2014 - 02:18 PM, said:

I found a pretty good example of it, right here.

There was at least one, possibly two mechs that were in front of me, line of site but the missile turret targeted me.



Anyway, I'm guessing the game is prioritizing turret targets by whom ever is direct line of site, closest, and has the highest damage dealt that round.


Well, I still see them trying to "leg me" w/o having actual proper LOS to my legs. It happens a lot while running a light mech. So, it occasionally does "nothing" vs me.

Edited by Deathlike, 24 April 2014 - 04:08 PM.


#509 TheCaptainJZ

    Member

  • PipPipPipPipPipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 3,658 posts
  • LocationUnited States

Posted 24 April 2014 - 04:10 PM

Turrets seem to shoot "down" moreso than they ever shoot "up" at a target. I suppose it's related to the terrain providing cover to mechs on hills.

#510 Klappspaten

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,211 posts
  • LocationGermany

Posted 24 April 2014 - 05:09 PM

View PostDimento Graven, on 24 April 2014 - 02:18 PM, said:

I found a pretty good example of it, right here.

There was at least one, possibly two mechs that were in front of me, line of site but the missile turret targeted me.



Anyway, I'm guessing the game is prioritizing turret targets by whom ever is direct line of site, closest, and has the highest damage dealt that round.


What you experienced in this short clip is pretty easy to explain.
If there are multiple targets available to the turrets they shouldn´t all shoot at the same target. So if there are 5 turrets you only need to be the fifths closest do the turret that shoots at you.

#511 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 24 April 2014 - 09:18 PM

View PostKlappspaten, on 24 April 2014 - 05:09 PM, said:

What you experienced in this short clip is pretty easy to explain.
If there are multiple targets available to the turrets they shouldn't all shoot at the same target. So if there are 5 turrets you only need to be the fifths closest do the turret that shoots at you.
But that's not what happened here. There were as many as 4 'mechs that were in front of me, in direct line of sight of the turret that fired at me, and at the point it finally fired at me, there were STILL 2 Assault 'mechs directly in front of me, neither of which appeared to ever receive any missile or laser fire from turrets. Heck the ONLY turret that would have had direct line of sight of us 3 was the one that FINALLY started firing at me, and again, my understanding is it would have fired at the first 'in range' 'mech in its direct line of sight.

And trust me, the laser turrets WILL focus fire on you you're the closest 'mech, I've had THAT happen to me too. The laser turrets only start choosing different targets when you are no longer the closes 'mech to all of them. As soon as a team mate gets closer to one of the laser turrets than you, there's a chance it will start firing at that teammate.

#512 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 25 April 2014 - 12:30 AM

View PostDimento Graven, on 24 April 2014 - 06:03 AM, said:

Karl,

Another question that maybe you can get some answers on.

TURRETS - How does the turret system designate target priority?

I >>KNOW<< it's not whatever enemy is in range, line of sight, first/closest.
<snip..>


This I'm not too sure about. I'll try asking the gameplay team about intended turret behaviour tomorrow.

View Postshellashock, on 24 April 2014 - 07:31 AM, said:

I am all but positive that this hasn't been asked before here.

How often is "The Plan" updated, and is a fixed timeline for updating.

Ie, "The Plan" is updated every month and is only updated on the first day of the second week of said month, etc.


And this is handled by Fox :) I'll see if he's interested in popping in and answering a few questions.

#513 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 25 April 2014 - 12:37 AM

View PostModo44, on 24 April 2014 - 08:03 AM, said:

There was a lot of talk about the Elo delta between teams. Can you tell us how the Elo delta within each team is? Will the launch module improve that?


This is a really interesting question. We're starting to gather information on this using PTS of course; but even a full day of public test data is worth at most a few minutes of production. There's just that much difference in population. 3/3/3/3 obviously puts a lot of constraints on which users are valid match candidates, so based on this alone I'd expect skill matching to worsen by some amount. The introduction of the user-explicit cancel action makes this a much harder call in practice however. I'd like to wait until we've acquired a day or two of actual production data with 3/3/3/3 before trying to answer this.

#514 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 25 April 2014 - 12:56 AM

View PostBuso Senshi Zelazny, on 24 April 2014 - 10:08 AM, said:

Which brings me to my question for Karl and the team. I'm not really sure if background in any of this material is required for game design, but it is obviously relevant in many cases. Karl clearly has some knowledge in these areas based on his previous posts, but I'm curious how the rest of the team stands up. What are the backgrounds of some of the other members of the team in the Gameplay and Design departments? Are there any with engineering or physics backgrounds, or did they come from a coding background and had to learn the physics stuff after the fact?

Also, could you elaborate a little bit more on this from Paul's latest interview with NGNG? What aspects are you working on more closely with the design team? Does this kind of interaction between departments help or hinder the production and release of features? Does it allow the team to develop features faster, or simply to make them "better" once they are released?


There are several of us with software engineering backgrounds. I think I'm the only engineer with professional physics experience, as I've done ground-up implementations of 3D rigid body simulators before. It really depends on what you're intending to work on, as different roles have wildly different requirements.

As to the iteration processes we're taking part with design, it's only been positive for us so far. The process is pretty much as I described earlier in this thread. Design puts together a formal document on some specific feature, detailing how they would like the feature to work. They submit the document to us, and we point out areas where further clarification is required, inconsistent or contradictory behaviours, portions that might have high technical risk, etc. On occasion we point out ways that we might reduce implementation costs by making small tweaks to the design. The design team writes down all our feedback and incorporates it into a second revision. They then submit that second revision for review. This process repeats until we all agree that the feature is ready for implementation.

#515 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 25 April 2014 - 12:59 AM

View PostMawai, on 24 April 2014 - 10:58 AM, said:

Hi Karl,

Apparently some additional bugs have been found in the hit registration code with the potential to have a significant impact on SRM's (Paul NGNG podcast).

However, unless there is different hit registration code for every weapon (which seems a bit unlikely), is it possible that the bug will affect multiple weapons?


View PostTheCaptainJZ, on 24 April 2014 - 11:10 AM, said:

iirc, one of the major bugs with srms that is getting fixed is an issue with how cryengine handles explosions. it wasn't simulating all the appropriate explosions within the same frame (some sort of cap on it) and some were being delayed til the next frame at which point the target would have moved.


Yes indeed, the explosion queue fix will affect multiple weapons. Lasers and other trace-fires will not be affected though.

#516 Iqfish

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,488 posts
  • Google+: Link
  • LocationGermany, CGN

Posted 25 April 2014 - 01:36 AM

View Postp4r4g0n, on 24 April 2014 - 07:36 AM, said:


Not in this thread but I believe it has been stated somewhere before. It is updated monthly but not sure about specific date. Depends on when time is available I think as I am not aware of a specific update day.


I asked Bryan on Twitter.
He actually updates it and he does it after every first patch of the month.

#517 Modo44

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • 3,559 posts

Posted 25 April 2014 - 02:00 AM

View PostKarl Berg, on 25 April 2014 - 12:37 AM, said:

3/3/3/3 obviously puts a lot of constraints on which users are valid match candidates, so based on this alone I'd expect skill matching to worsen by some amount. The introduction of the user-explicit cancel action makes this a much harder call in practice however.

On the other hand, Elo buckets should block some of the current edge cases. I wonder to what extent that will help, and if this can be improved further (other than the obvious -- making more, smaller Elo buckets, which has its own issues).

#518 Kmieciu

    Member

  • PipPipPipPipPipPipPipPipPip
  • Urban Commando
  • Urban Commando
  • 3,437 posts
  • LocationPoland

Posted 25 April 2014 - 03:46 AM

Karl, please take a look at this post - it's about the matchmaker:

http://mwomercs.com/...ost__p__3228282

#519 DarkonFullPower

    Member

  • PipPipPipPipPip
  • Big Brother
  • 191 posts

Posted 25 April 2014 - 04:01 AM

If Karl can answer this question for me, it will put to rest the only major worry I've had for this game since closed beta.

Where do Hardcore Lone Wolf players fit in Community Warfare?

The last time I found any answer to this question, it was that Lone Wolf = casual player. I cannot disagree with this enough.

Not everyone will want to choose a side, and may also not wish to be a part of a corp. But that does not automatically mean that they don't want to be apart of Community Warfare. It just means that they wish to remain independent and fight for whoever they feel like at the time, by their own choice.

Even something as simple as a checkbox for what factions to never random into would be a excellent starting point. The final goal would be that the Lone Wolf has full control over who he stays with, allowing him to remain with a group if he filled the gap they were missing adequately, without the player having to temporarily join their corp/faction. I'm positive that if such a system is not in place, you will have people changing their corp/faction constantly in order to play with who they want at the time. This will cause confusion among the players, as no one can tell who is temporary, and who is a devoted member. It would be better to clearly show that I am a temp Lone Wolf, and that I may leave whenever I want.

This single worry is the entire reason I have yet to spent a dime on this game, as I desire to play as a Hardcore Lone Wolf. If I cannot do so, then there is no point for me to play this game. I hope you can give me a clear answer so that I can finally deside if I should start spending money, or leave and never look back.

#520 EGG

    Member

  • PipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 322 posts

Posted 25 April 2014 - 04:49 AM

Hi Karl

I'd assume there's been manual data-mining of match stats to find optimal mechs / builds / weapons for the balance fixes we see come through. But I was wondering if there had been any movement towards hooking these stats back into other systems?

It would be interesting if the matchmaker were able to take into account the "match efficiency" of the Awesome(or specifically the AWS-8Q, or even a specific build), rather than just it's weight and class.

Cheers





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users