Jump to content

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


1911 replies to this topic

#421 Klappspaten

    Member

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

Posted 21 April 2014 - 10:54 AM

Quick info:
Karl Berg is on the latest NGNG.tv Podcast #109

#422 Tedarin

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 77 posts
  • LocationThe North

Posted 21 April 2014 - 12:29 PM

View PostKlappspaten, on 21 April 2014 - 10:54 AM, said:

Quick info:
Karl Berg is on the latest NGNG.tv Podcast #109

https://soundcloud.c...g-109-karl-berg

#423 shellashock

    Member

  • PipPipPipPipPipPip
  • Little Helper
  • 439 posts

Posted 21 April 2014 - 12:35 PM

Not sure if this has been asked before or if you are allowed to comment on this, but I would appreciate you replying to these:

- What is the current state of MASC, and by extension, the Flea?

- Have you had any breakthroughs in fixing the network issues connected to MASC's function or is the investigation still ongoing?

and

- If you had to give a rough estimate (if even possible), how long would it be from right now (21'st of April), until MASC could be implemented?

#424 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 April 2014 - 02:38 PM

Yeah, NGNG was fairly last second, but was quite fun to participate in. They asked me beforehand if there were things I wanted to discuss. It's been a fairly crazy last few weeks. That day in particular was very busy; so I ended up pulling a blank on topics, and kind of left them hanging. Still, they managed to drag a few answers out of me, which are hopefully of interest to some of you.

#425 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 April 2014 - 02:43 PM

View Postshellashock, on 21 April 2014 - 12:35 PM, said:

- What is the current state of MASC, and by extension, the Flea?

- Have you had any breakthroughs in fixing the network issues connected to MASC's function or is the investigation still ongoing?

and

- If you had to give a rough estimate (if even possible), how long would it be from right now (21'st of April), until MASC could be implemented?


This would require investigation still. Depending on the success of the recent HSR fixes, when they all make it to production at least, and on what the maximum object speeds we would be required to support, network simulation may or may not still be a concern.

Obviously, the date estimate relies on the answer to the investigation. If there are no network changes, then the date is determined by design giving us a document outlining expected behaviour, and gameplay adding the required functionality to the game. If networking changes are required, the date would likely be dominated by the extent of those changes.

#426 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 21 April 2014 - 02:47 PM

View PostKarl Berg, on 21 April 2014 - 02:43 PM, said:

...

Obviously, the date estimate relies on the answer to the investigation. If there are no network changes, then the date is determined by design giving us a document outlining expected behaviour, and gameplay adding the required functionality to the game. If networking changes are required, the date would likely be dominated by the extent of those changes.
So... How I interpret this statement is:

Once HSR is confirmed, 'working' AND once design actually ASKS for it (as you've just stated they haven't given you a requirements document yet), it will be some undetermined time after that.

So ultimately, MASC isn't being worked on, because Design hasn't even 'thought it out yet.'

The answer to when is: Some undetermined time after it's actually designed.

Well... At least we can stop holding our breaths on THAT feature!

Thanks Karl!

#427 Tekadept

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • 1,290 posts
  • LocationPerth, Australia

Posted 21 April 2014 - 04:41 PM

View PostKarl Berg, on 21 April 2014 - 02:43 PM, said:


This would require investigation still. Depending on the success of the recent HSR fixes, when they all make it to production at least, and on what the maximum object speeds we would be required to support, network simulation may or may not still be a concern.

Obviously, the date estimate relies on the answer to the investigation. If there are no network changes, then the date is determined by design giving us a document outlining expected behaviour, and gameplay adding the required functionality to the game. If networking changes are required, the date would likely be dominated by the extent of those changes.

IS Knockdowns/collisision in the same boat? awaiting HSR fixes, or does that have its own Network physics issues seperate to HSR issues?

Karl Berg is the only person that could ALMOST make me want to listen to NGNG, but sadly I still can't bring myself to do it :).

Edited by Tekadept, 21 April 2014 - 04:42 PM.


#428 Cimarb

    Member

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

Posted 21 April 2014 - 05:06 PM

View PostKarl Berg, on 16 April 2014 - 11:06 PM, said:


Not here; but Paul has agreed to do a segment with NGNG tomorrow later today.

View PostKarl Berg, on 21 April 2014 - 02:38 PM, said:

Yeah, NGNG was fairly last second, but was quite fun to participate in. They asked me beforehand if there were things I wanted to discuss. It's been a fairly crazy last few weeks. That day in particular was very busy; so I ended up pulling a blank on topics, and kind of left them hanging. Still, they managed to drag a few answers out of me, which are hopefully of interest to some of you.

Somehow, I'm guessing you got shoehorned into the last second thing because Paul "agreed" to do it and couldn't? I can't wait to hear it (going to do that right after I post this), but like I said earlier in this thread, it would be really nice if you didn't have to do all of the community outreach around here... we need some answers on other topics once in a while!

#429 Rebas Kradd

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,969 posts

Posted 21 April 2014 - 05:19 PM

They have a special session with Paul coming. I hear there will be beer. It oughta be fun.

#430 p4r4g0n

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • 1,511 posts
  • LocationMalaysia

Posted 21 April 2014 - 05:49 PM

Karl, been reading a number of posts / threads recently where players seem to be facing massive / long losing streaks. (e.g. here and here). It might just be perception bias, trolling or just players in the outlier low Elo band but given the number of people who have expressed frustration with the new player experience recently, I'm wondering if there is a problem somewhere with the matchmaking system.

It seems very odd to see reports of >26 straight losses or consistently losing over a 1 month period assuming the numbers are correct.

Just wondered if you had any thoughts about this.

#431 Cimarb

    Member

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

Posted 21 April 2014 - 07:20 PM

View PostRebas Kradd, on 21 April 2014 - 05:19 PM, said:

They have a special session with Paul coming. I hear there will be beer. It oughta be fun.

I love being a White Knight, and want nothing more than for this to happen, but I won't be holding my breath either. Please, Paul, prove me wrong, I DARE YOU!

#432 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 April 2014 - 08:03 PM

View PostDimento Graven, on 21 April 2014 - 02:47 PM, said:

So... How I interpret this statement is:

Once HSR is confirmed, 'working' AND once design actually ASKS for it (as you've just stated they haven't given you a requirements document yet), it will be some undetermined time after that.

So ultimately, MASC isn't being worked on, because Design hasn't even 'thought it out yet.'

The answer to when is: Some undetermined time after it's actually designed.

Well... At least we can stop holding our breaths on THAT feature!

Thanks Karl!


Yes, and no.

First, a bit of background on our development pipeline. Handing me a design document for technical breakdown and implementation is a very formal step in the development pipeline. This happens when developer resources become available, and the product owners have decided the feature in question is the highest priority use of those resources. At this point design provides myself and a few others their completed document. We sit down in a technical review and point out edge cases, undefined or contradictory behaviours, technically risky or needlessly complicated complicated components, etc.. Design revises the document to address our concerns, and we iterate until all concerns have been addressed. We'll then do high-level task breakdowns to identify which specific resources are required for which subtasks, and estimate the number of sprints required to complete the work. Production takes this data away and does their thing. Sprints start, stuff gets implemented, etc...

In that light:

- Design almost certainly has a document relating to MASC; just because I don't have it in hand, doesn't mean it doesn't exist. I'm sure design has dozens, if not hundreds of documents in various stages of completion.

- If they handed us a MASC document right now, the technical risk would be called out during the tech review. The feature would be blocked on investigation, and potentially additional network related work required to mitigate those risks.

- The developers required to work on MASC are mostly overlapping with the developers required for LaunchModule, Clans, and CW. Scheduling MASC now would directly compete with those features for development resources, and therefore push the delivery date for those features back.

Hopefully that clears things up a bit.

#433 Sybreed

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,199 posts
  • LocationQuebec

Posted 21 April 2014 - 08:10 PM

Wow, I just realized this thread existed. What types of questions can we ask Karl? I'd mostly talk about hardpoints and the heat system... though I'm not sure I would get much answers since these are touchy subjects :)

#434 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 April 2014 - 08:26 PM

View Postp4r4g0n, on 21 April 2014 - 05:49 PM, said:

Karl, been reading a number of posts / threads recently where players seem to be facing massive / long losing streaks. (e.g. here and here). It might just be perception bias, trolling or just players in the outlier low Elo band but given the number of people who have expressed frustration with the new player experience recently, I'm wondering if there is a problem somewhere with the matchmaking system.

It seems very odd to see reports of >26 straight losses or consistently losing over a 1 month period assuming the numbers are correct.

Just wondered if you had any thoughts about this.


I have the data from the ten thousand foot perspective. In this case, I would have to go back and look at the specific match data for grimunk to see whether his/her particular experience was due to bad luck, a consistent run of failures on the matchmaking side, or whatever else may have caused this. Assuming a 50% chance of win/loss, and assuming the outcome of each game is independent, a 26 run losing streak has astronomically low odds; something like one millionth of one percent. We might have had enough games played to make this a reasonable occurrence by this point, but I would expect this was caused by other factors. There *is* correlation between matches due to player behaviour for example, breaking the independence of each outcome. It's possible to skew Elo via grouping behaviours for another, breaking that 50 percentile base. Matches frequently have non-zero Elo delta's between teams for a third, yet again making that 50% base in the probability calculation incorrect.

#435 ICEFANG13

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,718 posts

Posted 21 April 2014 - 08:27 PM

Karl I have no questions but I sincerely thank you for answering other questions. Taking time to speak to us on your forums and in such a thoughtful, insightful, and thorough way. Its quite appreciated.

Although, honestly, it seems that PGI does not much reference the forums, should someone at PGI read this, as a customer etc, I would suggest that Karl would receive a raise.

#436 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 21 April 2014 - 08:35 PM

View PostSybreed, on 21 April 2014 - 08:10 PM, said:

Wow, I just realized this thread existed. What types of questions can we ask Karl? I'd mostly talk about hardpoints and the heat system... though I'm not sure I would get much answers since these are touchy subjects :P


If it's about design, I probably can't answer, due to not being an authority on these subjects. If it's about future design, I definitely can't answer! That's covered under NDA.

If it's about software architecture or algorithms, fire away and I'll answer if I can. :( If I don't know the specifics, like with that little upside-down dude, I'll try to follow up and find an answer.

#437 p4r4g0n

    Member

  • PipPipPipPipPipPipPipPip
  • Knight Errant
  • 1,511 posts
  • LocationMalaysia

Posted 21 April 2014 - 08:54 PM

View PostKarl Berg, on 21 April 2014 - 08:26 PM, said:


I have the data from the ten thousand foot perspective. In this case, I would have to go back and look at the specific match data for grimunk to see whether his/her particular experience was due to bad luck, a consistent run of failures on the matchmaking side, or whatever else may have caused this. Assuming a 50% chance of win/loss, and assuming the outcome of each game is independent, a 26 run losing streak has astronomically low odds; something like one millionth of one percent. We might have had enough games played to make this a reasonable occurrence by this point, but I would expect this was caused by other factors. There *is* correlation between matches due to player behaviour for example, breaking the independence of each outcome. It's possible to skew Elo via grouping behaviours for another, breaking that 50 percentile base. Matches frequently have non-zero Elo delta's between teams for a third, yet again making that 50% base in the probability calculation incorrect.


Thanks for confirming the feeling I had that such long losing streaks were low probability events. Given that, I am surprised at the apparent number of people who seem to be experiencing this recently as evidenced by the number of threads / posts that I've come across where this is either mentioned or brought up specifically.

Incidentally, I listened to the NGNG #109 podcast and gotta say I'm surprised I was a little surprised that your session was so short. Possibly due to your not giving them topics to talk about? :(

#438 Kageru Ikazuchi

    Member

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

Posted 21 April 2014 - 10:03 PM

View PostKarl Berg, on 21 April 2014 - 08:03 PM, said:

... our development pipeline. Handing me a design document ... Hopefully that clears things up a bit.

Wow -- I've pretty much accepted "these things take time" and "U.I. 2.0 = Bottleneck" for why we hear about things "coming soon", and then nothing for months.

What are some of the things we might not be aware of (that you can tell us about) that are actually being worked on right now?

#439 ThatBum42

    Member

  • PipPipPipPipPipPip
  • Philanthropist
  • 220 posts

Posted 22 April 2014 - 03:30 AM

Just wanted to say I really enjoy the game and genuinely appreciate the effort you all are putting in. I understand the 'when it's done' approach to development, as long as I can have my stompy robot fun with my friends and think up crazy loadouts :(. There's been an amazing amount of progress given your available manpower, based on vids I saw from the early days. Please don't let the haters get to you!

I was wondering what the bandwidth usage is like. Perhaps per day of operation, or per match even? Also, what's the hardware specs on the servers? This kinda stuff interests me, being a budding network engineer and all.

Edited by ThatBum42, 22 April 2014 - 03:30 AM.


#440 Klappspaten

    Member

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

Posted 22 April 2014 - 04:45 AM

View PostKarl Berg, on 21 April 2014 - 02:43 PM, said:


This would require investigation still. Depending on the success of the recent HSR fixes, when they all make it to production at least, and on what the maximum object speeds we would be required to support, network simulation may or may not still be a concern.

Obviously, the date estimate relies on the answer to the investigation. If there are no network changes, then the date is determined by design giving us a document outlining expected behaviour, and gameplay adding the required functionality to the game. If networking changes are required, the date would likely be dominated by the extent of those changes.


I know this is probably a design question. But I think it could be an issue with MASC too.
So maybe you could just forward the question to the design guys.

Do you guys know that the Locust sometimes damages its own legs just by running at top speeds around 170kp/h?
I even died of that once when I had both legs cherry red. I was running along without an enemy anywhere near me and BAM I died from falling damage without actually falling off of anything.
Its not a lot of damage, but it is damage. And when the speeds go even higher than that due to MASC it could become a serious issue.

If neccessary I would try to get some footage of it, but I can´t do it today since I´m not at home.

Edited by Klappspaten, 22 April 2014 - 04:50 AM.






18 user(s) are reading this topic

0 members, 18 guests, 0 anonymous users