Jump to content

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


1911 replies to this topic

#1101 evilC

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,298 posts
  • LocationLondon, UK

Posted 25 June 2014 - 09:52 AM

@Karl

Oh, and any news on when we will see zoom / crosshair fixes?
I realized of course that the z-depth of the adv zoom pip window (See "Rotataion causes translation" above) is the reason we are seeing the targetting boxes move around when using adv zoom (even without TrackIR), and will also affect the crosshair with adv zoom enabled and arm lock disabled.

Having considered this, it now all makes sense why I have trouble aiming PPCs with advanced zoom, and why if using it, having arm lock on is absolutely essential for accurate aiming.

I would say that this is more important than fixing TrackIR, as it affects everybody - just most people do not understand why.

#1102 East Indy

    Member

  • PipPipPipPipPipPipPipPip
  • The Hammer
  • The Hammer
  • 1,217 posts
  • LocationPacifica Training School, waiting for BakPhar shares to rise

Posted 25 June 2014 - 12:40 PM

Karl, what about a kind of blind betting on tonnage in which a team that has the lowest total weight receives a small bonus to C-Bills and XP? Think 2.5%. Basically, a reward to the lowest bidder, so if a player is confident in how he performs with a 20/40/60/80-tonner, he's giving his team a chance for better rewards, win or lose, if the matchmaker is unable to match his weight.

#1103 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 25 June 2014 - 06:26 PM

I have been wondering about how CASE works. Could we get a description of its function/behavior in MWO for both IS and Clan (such as how it handles explosions from ammo and Gauss, does it distribute all damage like in the original description or reduce damage down, like to one ton)?

I'd love to read about it in Breakdown in the Command Chair area, if this sorta thing shouldn't be answered here.


And again, thank you for taking time reading and responding to us.

Edited by Praetor Knight, 25 June 2014 - 06:31 PM.


#1104 Kageru Ikazuchi

    Member

  • PipPipPipPipPipPipPipPip
  • The Determined
  • The Determined
  • 1,190 posts

Posted 25 June 2014 - 07:15 PM

View PostPraetor Knight, on 25 June 2014 - 06:26 PM, said:

I have been wondering about how CASE works. ...


I'm not a Dev (and didn't even stay in a Holiday Inn express last night) but, what I understand is:
- CASE restricts damage from an ammo explosion to the protected component (i.e.: side torso), preventing it from transferring to another component (i.e.: center torso).
- CASE does not reduce damage to a component ... a full ton of ammo explodes, you're going to lose that component (and in the case of a side torso, the attached arm, as well) ... a partial ton of ammo explodes, it will damage (and might destroy) the component, it might score a critical hit on another item stored in that component (such as another ton or partial ton of ammo), and might destroy it (or cause it to explode)
- CASE is only effective (and only marginally effective, at that) in side torsos on Inner Sphere 'mechs
- Clan CASE weighs nothing, and costs zero crit slots, so is included on every component for Clan 'mechs, so it is marginally safe (and marginally efficient, depending on the 'mech and component) to equip ammo in the same component as the weapon

#1105 Wintersdark

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • 13,375 posts
  • Google+: Link
  • Twitter: Link
  • LocationCalgary, AB

Posted 25 June 2014 - 11:12 PM

View PostWillothius, on 25 June 2014 - 02:41 AM, said:

Well having 2 half ton slots also DOUBLES the chances of a crit in any ammo bin! But, I guess you'd hope that by that time you'd indeed have emptied out one.

While it increases the chance of a crit hitting an ammo bin, it also provideseach chance with 10hp. Not helpful if a PPC is critting you, but an LBX for example? Those one damage pellets will be distributed between the two ammo bins (and other internal items like gyros and stuff) so it'll take twice as long to destroy a bin.... And when it does, IF there's an explosion (10%) it'll be for at worst half the damage.

So, in theory you'd generally want to fill slots with half tons... Except! Many half ton ammo bins actually carry slightly less than half a full ton's shots.

#1106 Cimarb

    Member

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

Posted 26 June 2014 - 05:21 AM

View PostWintersdark, on 25 June 2014 - 11:12 PM, said:

So, in theory you'd generally want to fill slots with half tons... Except! Many half ton ammo bins actually carry slightly less than half a full ton's shots.

That has got to be a bug? I don't use half tons myself, but it definitely shouldn't be that way...

#1107 Shlkt

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 319 posts

Posted 26 June 2014 - 06:46 AM

View PostCimarb, on 26 June 2014 - 05:21 AM, said:

That has got to be a bug? I don't use half tons myself, but it definitely shouldn't be that way...


Not a bug, just the result of odd numbers. It mainly affects the AC/20. A full ton of ammo gives you 7 shots, so half a ton of ammo rounds down to just 3 shots. For the clan version they just multiplied the IS version by 5 due to burst fire, so you have 35 shots vs. 15 shots for the half ton.

Edited by Shlkt, 26 June 2014 - 06:47 AM.


#1108 Cimarb

    Member

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

Posted 26 June 2014 - 12:54 PM

View PostShlkt, on 26 June 2014 - 06:46 AM, said:


Not a bug, just the result of odd numbers. It mainly affects the AC/20. A full ton of ammo gives you 7 shots, so half a ton of ammo rounds down to just 3 shots. For the clan version they just multiplied the IS version by 5 due to burst fire, so you have 35 shots vs. 15 shots for the half ton.

Ahhhh, gotcha. That does make sense, though you would think they would round UP since you are sacrificing an additional crit slot...

#1109 TheCaptainJZ

    Member

  • PipPipPipPipPipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 3,659 posts
  • LocationUnited States

Posted 26 June 2014 - 01:36 PM

View PostCimarb, on 26 June 2014 - 12:54 PM, said:

Ahhhh, gotcha. That does make sense, though you would think they would round UP since you are sacrificing an additional crit slot...

No, because it's not supposed to be an advantage, just a way to fill that last .5 tons

#1110 Heffay

    Rum Runner

  • PipPipPipPipPipPipPipPipPipPip
  • The Referee
  • The Referee
  • 6,458 posts
  • LocationPHX

Posted 26 June 2014 - 02:01 PM

Dear Mr Berg,

A while back you mentioned that the hit registration was in the 70-80% range. With the implementation of the Buckton fix (as well as the other ones that have been recently implemented), has that gone up significantly? Can you share with us the new numbers?

Love,
Heffay

#1111 DarkonFullPower

    Member

  • PipPipPipPipPip
  • Big Brother
  • 191 posts

Posted 27 June 2014 - 03:37 PM

View PostWintersdark, on 25 June 2014 - 11:12 PM, said:

While it increases the chance of a crit hitting an ammo bin, it also provideseach chance with 10hp. Not helpful if a PPC is critting you, but an LBX for example? Those one damage pellets will be distributed between the two ammo bins (and other internal items like gyros and stuff) so it'll take twice as long to destroy a bin.... And when it does, IF there's an explosion (10%) it'll be for at worst half the damage.

So, in theory you'd generally want to fill slots with half tons... Except! Many half ton ammo bins actually carry slightly less than half a full ton's shots.


2 damage pellets actually. LBX Crit damage is double of the pellet damage, last I checked. But the rest stands.

Edited by DarkonFullPower, 27 June 2014 - 03:37 PM.


#1112 TheCaptainJZ

    Member

  • PipPipPipPipPipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 3,659 posts
  • LocationUnited States

Posted 27 June 2014 - 10:14 PM

View PostKarl Berg, on 09 June 2014 - 11:10 AM, said:

Interesting. I'll escalate this as well. It sounds like it's just slipped through the cracks.

edit: Indeed an oversight. QA is reopening and reassigning the bug now.

Karl, can you find out if this has been corrected now or not (ties in matches giving no xp)? We are trying to keep this guide up-to-date: http://mwomercs.com/...01#entry3514201

Edited by TheCaptainJZ, 27 June 2014 - 10:15 PM.


#1113 Deathlike

    Member

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

Posted 29 June 2014 - 01:27 PM

Karl,

Could you inquire to whoever's responsible for the Radar Deprivation module... about its interaction vs Target Decay (both Advanced and regular versions)?

It would be greatly appreciated.

Edited by Deathlike, 29 June 2014 - 01:27 PM.


#1114 SgtKinCaiD

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,095 posts
  • LocationBordeaux

Posted 30 June 2014 - 12:53 AM

@Karl

Dear Karl, since you are in charge of the new MM, can you confirm that with the reintroduction of the 4*3, we will see the ELO buckets back too ?

Thank you.

#1115 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 30 June 2014 - 01:50 PM

View PostHobo Dan, on 20 June 2014 - 11:37 AM, said:

Hello Karl. Bug Question

For quite some time now I’ve been plagued by the CryEngin FRead Error. What happens is that the game will crash to desktop (this can happen while logging in, while loading into a match, and less often during a match). When the game crashes I get the FRead Error message that states # out of ### files were miss read.

After some investigating, I found that every time this happened, it was because one of the game .pak files had been corrupted. Before Repair tool existed this involved either downloading the whole game again or getting ahold of fresh .paks. Now repair tool can handle it, but it still happens often to me. At least once a week. On patch weeks several times a night for a few days. Objects.pak is always a favorite, but many others have been corrupted as well. Locking the .paks doesn’t seem to work.

When I contact support, they suggest reinstalling the game, etc., etc. None of that has worked. I think I’ve had this issue in varying degrees of severity and frequency since August 2012.

I don’t see too many others out there reporting the issue. So either I’m an edge case, or a glutton for using the repair tool.

Any solutions?


Hey Hobo Dan, sorry for the delay on this response. Unfortunately this bug is apparently still open and under investigation. The bug is apparently very difficult to reproduce.

View PostSgtKinCaiD, on 30 June 2014 - 12:53 AM, said:

@Karl

Dear Karl, since you are in charge of the new MM, can you confirm that with the reintroduction of the 4*3, we will see the ELO buckets back too ?

Thank you.



It will see much tighter Elo restrictions brought into place. How restrictive depends on how constrained we will be able to set these values to on production and still maintain our goals for worst case user wait times.

View PostDeathlike, on 29 June 2014 - 01:27 PM, said:

Karl, Could you inquire to whoever's responsible for the Radar Deprivation module... about its interaction vs Target Decay (both Advanced and regular versions)? It would be greatly appreciated.


Yup, will do. I'll see if I can get Alex to answer any questions you might have on this.

#1116 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 30 June 2014 - 02:10 PM

View PostTheCaptainJZ, on 27 June 2014 - 10:14 PM, said:

Karl, can you find out if this has been corrected now or not (ties in matches giving no xp)? We are trying to keep this guide up-to-date: http://mwomercs.com/...01#entry3514201


Not fixed yet, but thanks for reminding me! I just made sure this got bumped up the priority list again.

#1117 Cimarb

    Member

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

Posted 30 June 2014 - 04:46 PM

Karl, can you maybe elaborate on what I got Bryan started on? I'm curious about the "84% are solo drops" statement and how that stat may have changed since it was made. (sorry about the format...)

@cimarbs so are groups counted as a single drop, or is a 4-man considered 4 drops?

Posted ImageBryan Ekman ‏@bryanekman if you mean for Elo. 4 = 1 Elo aggregate score.

‏@cimarbs so elo is changed as a group, but is that considered 1 drop or 4? I.e. Does 1 group = 1 solo for drop stats? Or 4-man = 4 drops?

Posted ImageBryan Ekman ‏@bryanekman currently counts as 1 group elo, not 4. It's under review internally to see if we need to break it up.

@cimarbs thanks - does that mean the 84% solo drops stat may have been inaccurate? Or has that changed in the last few months?

Posted ImageBryan Ekman ‏@bryanekman it varies, I'd have to check again.

‏@cimarbs I would really appreciate, as would the forumites. Hot topic since "84% solo" was stated.

#1118 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 30 June 2014 - 05:33 PM

View PostHeffay, on 26 June 2014 - 02:01 PM, said:

Dear Mr Berg, A while back you mentioned that the hit registration was in the 70-80% range. With the implementation of the Buckton fix (as well as the other ones that have been recently implemented), has that gone up significantly? Can you share with us the new numbers? Love, Heffay


That's actually the tricky thing about the Buckton fix. Hit registration did not improve at all. What was happening was after host state rewind verified the hit, the gameplay code would do a damage check on the dedicated server with distance falloff using unrewound positions (incorrect positions), and as a result was determining incorrect damage values to be applied.

#1119 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 30 June 2014 - 05:57 PM

View PostCimarb, on 30 June 2014 - 04:46 PM, said:

Karl, can you maybe elaborate on what I got Bryan started on? I'm curious about the "84% are solo drops" statement and how that stat may have changed since it was made. (sorry about the format...)


Heh, obviously two different things being discussed there. ;)

Regarding those grouping stats that you were inquiring about:

I've been asked about those stats several times in this thread even, and I don't know what time period they were pulled from or how exactly they were gathered. They are certainly out of date right now; and they are about to be completely obsoleted this Wednesday. Your best bet is to ask Paul or Bryan for updated numbers once we've gathered a few weeks of data on production with the new grouping systems.

And as to what Bryan was talking about:

We're still in the middle of ongoing investigations as to the measurable impacts of grouping on win outcomes in our game. We literally have millions and millions of data points to examine; so we can produce some highly statistically significant results. Because of this, I have some really neat results regarding the impact of premades on solo games and premades of other sizes. This is all in addition to all of the various predictive tests we've been running on different algorithms for Elo convergence using production data. It's fascinating stuff, we've had several extremely late nights examining, hypothesizing about, running experiments on, and generally discussing this data as a group, and we have some really cool results that I'm hoping we can share at some point soon. So that's likely why Bryan was a bit confused about the topic of discussion.

Note that these investigations are still ongoing. We will likely be making some fundamental changes to the way we track and measure skill ratings in the game; but those changes will be made only with high confidence, after all our investigations are completed.

#1120 ShinVector

    Liao Mercenary

  • PipPipPipPipPipPipPipPipPip
  • The Hammer
  • The Hammer
  • 3,711 posts

Posted 30 June 2014 - 06:31 PM

View PostKarl Berg, on 30 June 2014 - 05:33 PM, said:


That's actually the tricky thing about the Buckton fix. Hit registration did not improve at all. What was happening was after host state rewind verified the hit, the gameplay code would do a damage check on the dedicated server with distance falloff using unrewound positions (incorrect positions), and as a result was determining incorrect damage values to be applied.


You know Karl... That almost sound like the same problem that we have with lasers HSR and moving targets. Even though we see the hit, we always get poor damage and inaccurate damage.

It feels like HSR is not working at all of lasers most of time. If the server is using 'unrewound positions' to calculate laser damage.... That would explain we feel the need to lag lead shoot to get better damage with lasers.

I hope Neema is able to investigate (or already investigating) this possible issue.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users