Jump to content

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


1911 replies to this topic

#801 Heffay

    Rum Runner

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

Posted 23 May 2014 - 07:58 AM

View PostDimento Graven, on 23 May 2014 - 06:43 AM, said:

This was question was asked, and for whatever reason it resulted in ire and apparently was removed:

However, I believe this is actually very REASONABLE question given that, from what I understood the MM is going to go through an almost total revamp, AGAIN, to address the issues of implementing a workable "3/3/3/3" (which in my experience while it was working did ZERO to address the issue of pug stomping).

If even if it's not a 'total revamp', but just a SIGNIFICANT amount work that's going to occur, I too believe that "larger pre-mades" should should be accommodated.

We have the situation now where we can select multiple "game modes", the solo dropper should also have the option to select "Solo Only" and/or "Group" queues as well.

And please don't throw a fit about an old response already having been given. Since there's going to be, at least, a significant amount of changes to the MM, those reasons aren't necessarily going to apply any more, so YES, it SHOULD be reevaluated and reconsidered, as originally in closed beta, many of us thought this a GREAT feature of the game.


He asked Karl Berg to overrule the 3/3/3/3 rule that he's been told to implement. We don't even know if he disagrees with that rule. What if he says 3/3/3/3 and 4 man group limits is a great idea? Will Roadbeer stop his crusade to change it?

View PostDimento Graven, on 23 May 2014 - 07:58 AM, said:

A policy of "forgive and forget" with a business is VERY bad policy for the customers.


Are you kidding??

Edited by Heffay, 23 May 2014 - 08:04 AM.


#802 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 23 May 2014 - 08:04 AM

View PostHeffay, on 23 May 2014 - 07:58 AM, said:

He asked Karl Berg to overrule the 3/3/3/3 rule that he's been told to implement. We don't even know if he disagrees with that rule. What if he says 3/3/3/3 and 4 man group limits is a great idea? Will Roadbeer stop his crusade to change it?
Again, my experience with 3/3/3/3 when it was that it did NOT have the intended results.

All the matches I participated in that were 3/3/3/3 balanced were 12:0 stomps.

I have no access to the data, but it "felt" like the MM pulled in a LOT of "noobs", or at least people who had SIGNIFICANTLY LOW ELO scores for the chassis they were in at the time.

THAT, plus the sometimes 15+ minute wait time for a match seemed to me to indicate the idea of 3/3/3/3 might need some rethinking.

THAT being said, it seems to me that since the MM is going to be revamped (if not totally, then at least significantly) we have an opportunity to revisit the group size restrictions for public matches.

The reasons previously stated for not being able to allow 3-11 size groups were tied to issues with the MM and how it worked. Since that is almost definitely going to change significantly, WHY NOT look into it?

#803 Heffay

    Rum Runner

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

Posted 23 May 2014 - 08:07 AM

View PostDimento Graven, on 23 May 2014 - 08:04 AM, said:

Again, my experience with 3/3/3/3 when it was that it did NOT have the intended results.

All the matches I participated in that were 3/3/3/3 balanced were 12:0 stomps.



All the matches I participated in were fantastic battles. I can't wait for it to be implemented again.

Dimento Graven said:

THAT, plus the sometimes 15+ minute wait time for a match seemed to me to indicate the idea of 3/3/3/3 might need some rethinking.


The queues were a result of the algorithm not working properly properly. Karl Berg mentioned when run against historical production data, the average queue time was 40 seconds.

Edited by Heffay, 23 May 2014 - 08:08 AM.


#804 DragonsFire

    Member

  • PipPipPipPipPipPipPip
  • Wrath
  • Wrath
  • 655 posts

Posted 23 May 2014 - 08:10 AM

Gents, this has been a fantastic thread thus far with lots of good info from Mr. Berg. Let's continue that trend and if you want to have a discussion on the matchmaker or business practices, kindly start another thread or take it to PM's.

#805 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 23 May 2014 - 09:10 AM

View PostHeffay, on 23 May 2014 - 08:07 AM, said:

All the matches I participated in were fantastic battles. I can't wait for it to be implemented again.
I guess when you're on the 'giving' end of a 12:0 win, it can seem fantastic...

Quote

The queues were a result of the algorithm not working properly properly. Karl Berg mentioned when run against historical production data, the average queue time was 40 seconds.
I don't understand the disparity though. What was different between "historical data" and post patch?

What did we users do differently that would have resulted in such wildly different queue times?

#806 Heffay

    Rum Runner

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

Posted 23 May 2014 - 09:33 AM

View PostDimento Graven, on 23 May 2014 - 09:10 AM, said:

I guess when you're on the 'giving' end of a 12:0 win, it can seem fantastic...

I don't understand the disparity though. What was different between "historical data" and post patch?

What did we users do differently that would have resulted in such wildly different queue times?


None of the matches were 12:0. They were see-saw battles that changed depending on a ton of factors.

I don't understand the disparity either. What we do know is that people are self-balancing now even with no actual change in wait times, so 3/3/3/3 will be just fine as people try to optimize their play time vs those who play only 1 class. The problem is that the algorithm that was doing the MM is flawed even though the concept has been proven both that people will self balance AND the matches are better.

#807 Cimarb

    Member

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

Posted 23 May 2014 - 11:26 AM

View PostDimento Graven, on 23 May 2014 - 08:04 AM, said:

Again, my experience with 3/3/3/3 when it was that it did NOT have the intended results.

All the matches I participated in that were 3/3/3/3 balanced were 12:0 stomps.

I have no access to the data, but it "felt" like the MM pulled in a LOT of "noobs", or at least people who had SIGNIFICANTLY LOW ELO scores for the chassis they were in at the time.

THAT, plus the sometimes 15+ minute wait time for a match seemed to me to indicate the idea of 3/3/3/3 might need some rethinking.

THAT being said, it seems to me that since the MM is going to be revamped (if not totally, then at least significantly) we have an opportunity to revisit the group size restrictions for public matches.

The reasons previously stated for not being able to allow 3-11 size groups were tied to issues with the MM and how it worked. Since that is almost definitely going to change significantly, WHY NOT look into it?

I posted in the Launch Module Test Feedback thread about this, with (albeit limited) examples. 3/3/3/3 did NOT have a noticeable effect on match outcome. See my examples

As long as they leave the 1-group rule out, I can live with 3/3/3/3, but only because I already think the vast majority of matches are well-balanced as far as weight. I don't think it fixes what they think it fixes, and stomps will continue to happen with or without it.

#808 Cimarb

    Member

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

Posted 23 May 2014 - 11:31 AM

View PostHeffay, on 23 May 2014 - 09:33 AM, said:


None of the matches were 12:0. They were see-saw battles that changed depending on a ton of factors.

I don't understand the disparity either. What we do know is that people are self-balancing now even with no actual change in wait times, so 3/3/3/3 will be just fine as people try to optimize their play time vs those who play only 1 class. The problem is that the algorithm that was doing the MM is flawed even though the concept has been proven both that people will self balance AND the matches are better.

The matches being "better" is totally subjective, though. I think they are exactly the same, as I just pointed out, Dimento thinks they are worse and you think they are better. I'm pretty sure it has much more to do with the time you play, your Elo (if that even exists) and your luck.

Regardless, the weight queue drop down is IMMENSELY more useful and efficient of a balancer than 3/3/3/3 is, even though it was changing so often back and forth the last few nights that you wound up just dropping in whatever because you couldn't switch mechs quick enough to make use of the information. It was usually quicker to just watch the numbers and wait a minute or two to drop when your weight class percent dropped.

#809 Wintersdark

    Member

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

Posted 23 May 2014 - 07:20 PM

Stomps will continue to happen no matter what. Simply because of the way the game plays, if one side gets a lucky kill or two at the start of the match, it's highly likely to snowball into a very one-sided match, even if the teams were very well matched in terms of mechs and skill.

In short, a stomp is not in any way indicative of a matchmaker failure. It CAN be a result of a badly made match, but it's really much more likely that that's just how the match played out. There are countless reasons why a match can end up a stomp that have no bearing on player skill or mechs used.

Edited by Wintersdark, 23 May 2014 - 07:21 PM.


#810 Modo44

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • 3,559 posts

Posted 23 May 2014 - 07:28 PM

Most of the above is redundant. A lot of matchmaker considerations were covered earlier in the thread. Check the first post before spamming.

#811 Randalf Yorgen

    Member

  • PipPipPipPipPipPipPipPip
  • 1,026 posts
  • Locationwith in 3m of the exposed Arcons rear ct

Posted 24 May 2014 - 04:36 AM

View PostHeffay, on 23 May 2014 - 07:58 AM, said:


He asked Karl Berg to overrule the 3/3/3/3 rule that he's been told to implement. We don't even know if he disagrees with that rule. What if he says 3/3/3/3 and 4 man group limits is a great idea? Will Roadbeer stop his crusade to change it?





Actually he didn't ask for the 3/3/3/3 to be altered, he asked if it was possible for the group size to be looked into and if possible for it to be implemented. you know groups from 5-11. Someone took exception and then a mod stepped in and hit a fly with a sledgehammer. I sent a report about this mod to support, they told me to report it to the mods, never the less it is in the files of support.

Imagine a hen house being looked after by a bunch of foxes and the hens go to the farmer and say that the foxes are not behaving correctly and the farmer telling the hens to take it up with the foxes. Even the Military is smarter than this and has other chains of command you can go to when there is an issue within your chain of command.


Edit, I just remembered my tag line, it's kind of the exact issue that is going on here.

Edited by Randalf Yorgen, 24 May 2014 - 04:37 AM.


#812 Mawai

    Member

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

Posted 24 May 2014 - 10:39 AM

View PostCimarb, on 23 May 2014 - 06:07 AM, said:

Please don't take this as personal, but can we (the community) get over this "90 days" thing?

Whoever said that originally obviously had no idea what they were saying and probably upset a LOT of people that actually DID know what was going on. It happens all the time in advertising, and I'm sure development is the same way. Let's get over it and worry about what is currently going on. It's like complaining about 3PV still...



LOL - the statement about CW was made by the president of PGI, Russ Bullock, on these forums, shortly before the the transition to open beta. The closed beta community was asking that the game be somewhat more feature complete before going to open beta and Russ (the president of a small company of 50 people whose only project was MWO and who one might expect to be knowledgeable about planning, schedule and roadmap - stated 90 days after open beta). This turned out to obviously and patently false showing exactly what you mentioned above about knowing what they were talking about. I suspect that if the comments had been made by anyone else then they would have been forgotten by now ... but the facts are that they were made by the person who has the responsibility to actually know the state of their product when speaking to the community ... and the fact that they were off by YEARS in their statement makes it hard for folks who have been around for a while to actually trust anything that Russ says about the game.

Please also note that communications issues misstatements, misunderstandings or even just ignorance of the game they are building have happened repeatedly over the last 2 years ... ti is why PGI as a whole has such a bad reputation for communications and why Karl's comments in this thread are so massively appreciated like a breath of fresh air in a room full of choking smoke :)

I'm quite willing to give PGI the benefit of the doubt but honestly their track record this year has not been stellar either ...
- UI2.0 released without addressing any of the comments from public tests as early as November. Lots of poor work flows and no support for many standard UI features like double clicking among others. They got it out the door on time though.
- The launch module was scheduled for late April. They introduced private matches but the matchmaker is MIA and Karl has to re-write it. This means that it is running 4 to 6 weeks late and even then I am not confident that the community will self-distribute into 3/3/3/3 without substantial queues for some classes. No one minds dropping in a light on release day ... but when they have to grind they might prefer to use their heavy weight class hero mechs ... as soon as that happens there will be a substantial heavy mech queue ... folks who have to wait 10 or 15 minutes to play a match with the mech they want have three choices ... wait, be unhappy and play a different class, quit ... any situation in which one of the solutions to a problem is "quit" is bad for MWO. Time will tell whether this works or not but the matchmaker design needs to be flexible.

Anyway, the recent track record does not bode well for a timely delivery of CW ... though I will keep my fingers crossed.

#813 Cimarb

    Member

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

Posted 24 May 2014 - 01:33 PM

View PostMawai, on 24 May 2014 - 10:39 AM, said:

LOL - the statement about CW was made by the president of PGI, Russ Bullock, on these forums... but the facts are that they were made by the person who has the responsibility to actually know the state of their product when speaking to the community ... and the fact that they were off by YEARS in their statement makes it hard for folks who have been around for a while to actually trust anything that Russ says about the game.

That is my point. Take it for what you will, but despite his best efforts to keep us informed, Russ has made many similar blunders in what is actually happening at PGI. It may be that he only knows what his managers know, who are telling him highly over-on fidget timelines to keep him happy, or it may be that he just isn't processing the information correctly before restating it to us, but regardless, I would take the word of someone actually working on it far more than someone at the top.

#814 Roadbeer

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • 8,160 posts
  • LocationWazan, Zion Cluster

Posted 24 May 2014 - 03:23 PM

View PostHeffay, on 23 May 2014 - 07:58 AM, said:


He asked Karl Berg to overrule the 3/3/3/3 rule that he's been told to implement.


Well, if that isn't a blatant lie...

#815 Deathlike

    Member

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

Posted 24 May 2014 - 10:35 PM

View PostRoadbeer, on 24 May 2014 - 03:23 PM, said:

Well, if that isn't a blatant lie...


It's in Heffay's nature. Let it go. You know these people when you see them.

#816 Mawai

    Member

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

Posted 25 May 2014 - 11:36 AM

View PostDeathlike, on 24 May 2014 - 10:35 PM, said:


It's in Heffay's nature. Let it go. You know these people when you see them.


Here ... I will toss in the reference to what was said and folks can be the judge ...

Russ said that 3/3/3/3 can't work correctly with the current matchmaker implementation and it needs re-writing. Karl got the job.

http://mwomercs.com/...on-matchmaking/

#817 Deathlike

    Member

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

Posted 25 May 2014 - 01:40 PM

View PostMawai, on 25 May 2014 - 11:36 AM, said:


Here ... I will toss in the reference to what was said and folks can be the judge ...

Russ said that 3/3/3/3 can't work correctly with the current matchmaker implementation and it needs re-writing. Karl got the job.

http://mwomercs.com/...on-matchmaking/


It's a lot more complicated than that.

For instance, the MM queue indicator (only visible when you're playing solo) gives you a look at the entire population. While this "may appear to be useful", this would actually be relatively useless when you consider the Elo buckets that 3/3/3/3 would have required. If the %s were based on your Elo for those weight classes (which means, they won't add up to 100%), then it would be a more useful indicator of how you can go about relieving the queues.

In any case, it's never as "simple" as it looks.

#818 SgtKinCaiD

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,096 posts
  • LocationBordeaux

Posted 26 May 2014 - 12:51 AM

View PostWintersdark, on 23 May 2014 - 07:20 PM, said:

Stomps will continue to happen no matter what. Simply because of the way the game plays, if one side gets a lucky kill or two at the start of the match, it's highly likely to snowball into a very one-sided match, even if the teams were very well matched in terms of mechs and skill.

In short, a stomp is not in any way indicative of a matchmaker failure. It CAN be a result of a badly made match, but it's really much more likely that that's just how the match played out. There are countless reasons why a match can end up a stomp that have no bearing on player skill or mechs used.


A stomp can be a pretty good indicator of matchmaker failure :
- if the MM were working correctly, we would still have 12-4 / 12-3 / etc. match, yes, but they would be slow landslide : one team slowly taking the advantage until the other can't fight back.
- what do we have now ? Lame roflstomp ending in less than 6 six minutes because the MM mix high and low ELO players. Basicaly it's not a 12vs12 match but rather 12v8 even less : the low ELO players are crushed within the first minutes of the match leaving their team with a missing lance if not more.

#819 TheCaptainJZ

    Member

  • PipPipPipPipPipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 3,685 posts
  • LocationUnited States

Posted 26 May 2014 - 01:22 AM

It's quite possible for a 12-0 win with all players of similar skill. One team might have set up some nice tactics, caught the enemies off-guard, and eliminated them without taking any losses due to, in short, playing smart or getting lucky.

Oh, they could be hurting pretty badly with missing arms and legs and such, but if the enemy team didn't coordinate fire, they might not have actually killed anyone.

There are simply too many variables and too many ways the match can go down.

#820 Karl Berg

    Technical Director

  • 497 posts
  • LocationVancouver

Posted 26 May 2014 - 01:07 PM

View PostFeatherwood, on 14 May 2014 - 10:14 AM, said:

Dear Karl,
I hope it's good time to ask questions, if so - here are few from me:
  • Can you please help to find (ask someone responsible) and publish a variable to permanently set cockpit light level (it's bind to '.' key by default)? It obviously has 0-3 values and by default set to 3, but it's not original CryEngine 3 variable, since I've tried every resemblant variable from SDK. It's quite annoying to turn this lighting off every match - reflections on the monitors irritate and distract me, so I'd rather prefer to set it once and for all in user.cfg. I'm sure many players will be as grateful for that as I would.
  • Can you please escalate to 3D modelers issue with missing back and top sides of cockpit in many Mechs? The problem is, that when you have lightning source behind you Mech - you get all the reflections and lighting effects inside the cockpit as if your 'Mech was transparent. This had been ruining my immersion since beta, but on all my reports I got replies that it's not considered a problem - freaking ridiculous! Why it was so hard to add few polygons on the cockpit model to make the it really closed space?
  • I am not big fan of 3333 idea (well, I find it stupid, to be honest), but I admit that it could bring some improvement comparing to old or current situation with balance. As many others I have my view on MM algorithm, but I wouldn't bother you with it, instead I'd like to ask one thing: in 3333 MM algorithm do you still calculate ELO rating for a group (1 per team)? If yes - what is the purpose of that?
I know that my questions are not exactly from your field, but still hope for your help. Thanks.


- I've looked into whether or not there is a user controllable cvar for default cockpit light levels. Currently there is not, but it doesn't appear to be too difficult to add one. I'll look into quickly adding this, but it takes a while for changes to trickle onto production.

- Missing cockpit geo. This is on the art side of course, so the best I can do is bring it up with the artist lead. They're pretty busy with Clan work right now, so I don't know if anyone will have time to go back and look at this currently.

- Yes, an Elo is calculated for groups. We use this calculated value (currently the average Elo of all team members, but I want to run analytics to determine whether or not this is actually a good metric to use) to match skill levels between groups, the same as we match skill levels between solo players.





4 user(s) are reading this topic

0 members, 4 guests, 0 anonymous users