Jump to content

- - - - -

Matchmaking Phase 4 Follow-Up - Feedback


277 replies to this topic

#121 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 23 April 2013 - 09:58 PM

View PostJess Hazen, on 23 April 2013 - 03:40 PM, said:

remove elo and we're good

Nahh, it is doing it's job.. in PUG games.

The only disappointment with the MM is, that they plan on 12v12 and do not have any tools for proper group play up.
The syncdropping teams, that will just have no other choice will perfectly ruin the PUG match experience and further hurt the new players experience.
Because they just add the communication advance to ELO and therefore srew the system for the other players.

But hey, it is a PGI invented problem, as they fail to provide a lobby for private matches of any size.

Sad is though, that this is nearly as certain as the "amen" in church is. It is so predictable to happen and yet, they once again will release it and then wonder why their official forum is burning down....

While it could be so damn stupid simple....

1.PUG matches, max 1 team of 4 per side + elo
2.private matches with no rewards but also without any elo or weightmatching or whatever... a group challenges group thingy.
3. the 12 v12 qeue with what ever PGI sees fit for official group competitions with rewards

done, everyone happy, elo working mostly, fair matches to everyone's flavor.

Edited by grayson marik, 23 April 2013 - 10:00 PM.


#122 Black Templar

    Com Guard

  • PipPipPipPipPipPip
  • Veteran Founder
  • Veteran Founder
  • 300 posts
  • LocationGermany

Posted 24 April 2013 - 01:35 AM

working as intended most of the time :(

#123 Krzysztof z Bagien

    Member

  • PipPipPipPipPipPipPip
  • Giant Helper
  • 710 posts
  • LocationUć, Poland

Posted 24 April 2013 - 04:12 AM

This tug of war example is not the best here - in that game there's only one factor to consider: collective strenght of the team being a sum of strenght of individual players (which can actually be easily measured and compared, so we can easily say that one player is better than the other), well, maybe also teamwork to some degree.
In a game like MWO there are far more things that can influence win or loss: players' skill (and that one is really hard to measure, it's even hard to tell what that skill actually is), overall teamwork (here it is much more important), whether team has voice comms or not (as it influences teamwork greatly), number of disconnects/AFKs and bugs that can make it harder or even impossible to play, team composition (some mechs are simply better than the others, and their loadout also matters), even the fact that one team has greater number of ECM units (as it was proven somewhere, I can't it find now, more ECM = higher chance to win). Probability to win a game does not equal skill, and in fact one has very little to do with the other. Player skill will become even less important factor in the equation with introduction of 12v12.

Edit: one more thing. Fore every player that says his experience improved thanks to Elo matchmaking there is one that says otherwise, so I belive it doesn't actually do anything and, as I mentioned earlier, it is merely a placebo effect.

Edited by Krzysztof z Bagien, 24 April 2013 - 04:30 AM.


#124 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 24 April 2013 - 06:41 AM

View Postgrayson marik, on 23 April 2013 - 09:58 PM, said:

1.PUG matches, max 1 team of 4 per side + elo
2.private matches with no rewards but also without any elo or weightmatching or whatever... a group challenges group thingy.
3. the 12 v12 qeue with what ever PGI sees fit for official group competitions with rewards

done, everyone happy, elo working mostly, fair matches to everyone's flavor.


They already said that teams stay at 4 for 12 man.

....and I can't get an 8 man anytime I play anyway....so yeah.

#125 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 24 April 2013 - 06:58 AM

View PostYokaiko, on 24 April 2013 - 06:41 AM, said:


They already said that teams stay at 4 for 12 man.

....and I can't get an 8 man anytime I play anyway....so yeah.

That is the reason for:

1.PUG matches, max 1 team of 4 per side + elo
2.private matches with no rewards but also without any elo or weightmatching or whatever... a group challenges group thingy. ---> ok and variable group sizes I should have added.
3. the 12 v12 qeue with what ever PGI sees fit for official group competitions with rewards

#126 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 24 April 2013 - 06:59 AM

View Postgrayson marik, on 24 April 2013 - 06:58 AM, said:

That is the reason for:

1.PUG matches, max 1 team of 4 per side + elo
2.private matches with no rewards but also without any elo or weightmatching or whatever... a group challenges group thingy. ---> ok and variable group sizes I should have added.
3. the 12 v12 qeue with what ever PGI sees fit for official group competitions with rewards


yeah great.

So I have to deal with pugging or find 12 players......which is just pugging with coms.

Lovely

#127 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 24 April 2013 - 07:05 AM

View PostYokaiko, on 24 April 2013 - 06:59 AM, said:


yeah great.

So I have to deal with pugging or find 12 players......which is just pugging with coms.

Lovely

NOPE EXACTLY NOT!
You can play with groups of 5 or 6 or 9 or 11 against other groups.

Ok, as for the grind... you are in PGI's hands.

but: I expect them to never allow such a thing like an old school lobby, where anyone can play as he likes with rewards on.

So my suggested sollution is far far better than the one that is currently coming donw the road.


but yes,

Quote

So I have to deal with pugging or find 12 players......which is just pugging with coms.

Iam on your side. a REAL group mode where you can match up against any team YOU like with any ruleset YOU like on a map YOU like.... but hey, we are speaking about IGP and PGI, who seem to share some of the scariest paranoia nightmares about control over the multiplayer experience....

Edited by grayson marik, 24 April 2013 - 07:05 AM.


#128 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 24 April 2013 - 07:20 AM

View Postgrayson marik, on 24 April 2013 - 07:05 AM, said:

Iam on your side. a REAL group mode where you can match up against any team YOU like with any ruleset YOU like on a map YOU like.... but hey, we are speaking about IGP and PGI, who seem to share some of the scariest paranoia nightmares about control over the multiplayer experience....



Yeah fantastic way that is working out.

They just ensured that there is a full jihad anytime they post anything....anywhere.

That is also one of the issues with 8 mans now. The money is crap because you generally spend a longer time running around and positioning the XP COSTS YOU MONEY to use because you need to convert to GXP to use it.

SO why go through the energy except for general e-peen.

#129 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 24 April 2013 - 07:22 AM

View PostYokaiko, on 24 April 2013 - 07:17 AM, said:



Yeah fantastic way that is working out.

They just ensured that there is a full jihad anytime they post anything....anywhere.

Well, I sayed so earlier and I do so again:

The "PUGSTOMP- random boringness- empty 8v8 queue- no competitive group play- dying community - Problems" are engineered and invented by PGI/IGP it is NOT a player problem!

If I only provide a random queue and an other queue where it is awesomly difficult to do team vs team games in, every gamer ends up on the random queue, and all those who would rather play differently thereby ruin the experience for the random lovers and newbies.

But it is not the players fault!

#130 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 24 April 2013 - 07:25 AM

I agree completely.

I did back when they came up with the "phase 1" brilliance.

#131 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 24 April 2013 - 07:34 AM

yepp!

Well, I understand the need for some automated matchmaking for CW, to be able to incoorperate all kind of events and stuff.

But this game should have had a lobby from the very start. Next to the matchmaker, a lobby would have provided a complete different dataset on weapons and almost any other aspect of the game. It would have opened doors to teams testing stuff like missized hitboxes and so on because you would hava had an environment to do so with a decent lobby.

After all, a lobby would have risen quality of test results and therefore sped up project development and success by a large portion! --> It would even have given room for the devs joining the public server to test something against real players and gather feedback instantly!

And last but not least, a Lobby would have enabled PGI/IGP to conduct some REAL tournaments with the community and commentators and streams and all kind of PR stuff to boost public interest into the game.

And while a decent Lobby is a WIN WIN for every style of gamer, for every noob joining in, for every Battletech nerd trying to set his dreams of conquest into a stalking robot, and for PGI/IGP when it comes to testing and community events.....
they still missed the point completely .. this is so far beyond my comprehension... I still can't understand it after months and months of discussions about matchmaking and how to make it fun and fair for all of us, nerdies and casuals.

Edited by grayson marik, 24 April 2013 - 07:38 AM.


#132 Ryvucz

    Zunrith

  • PipPipPipPipPipPipPipPipPip
  • 2,839 posts
  • LocationColorado Springs, Colorado

Posted 24 April 2013 - 12:36 PM

I'm definitely not a fan of ELO.

The old match making system may have pitted me against an 8 man, but by far, it was more likable.

My experience has been nothing but negative with ELO.

It's been more lopsided now than previously, only this time I'm not in the team that does the 8-0.

#133 Kageru Ikazuchi

    Member

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

Posted 24 April 2013 - 04:22 PM

First ... I agree that there should be a lobby system that would enable 16 (now, 24 soon) or less people to drop together in whatever map they want with whatever conditions they want ... but there should be no XP/C-Bills gained

This would allow people to play / train / test together without the randomness of PUG dropping, and with or without the cutthroat competitiveness of current 8-man drops.

I believe this is coming, but will be a small part MW:O, outside of the bigger community warfare battlefield.

#134 Hauser

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 976 posts

Posted 24 April 2013 - 05:27 PM

View PostKrzysztof z Bagien, on 24 April 2013 - 04:12 AM, said:

This tug of war example is not the best here - in that game there's only one factor to consider: collective strenght of the team being a sum of strenght of individual players (which can actually be easily measured and compared, so we can easily say that one player is better than the other), well, maybe also teamwork to some degree.
In a game like MWO there are far more things that can influence win or loss: players' skill (and that one is really hard to measure, it's even hard to tell what that skill actually is), overall teamwork (here it is much more important), whether team has voice comms or not (as it influences teamwork greatly), number of disconnects/AFKs and bugs that can make it harder or even impossible to play, team composition (some mechs are simply better than the others, and their loadout also matters), even the fact that one team has greater number of ECM units (as it was proven somewhere, I can't it find now, more ECM = higher chance to win).


I tried to keep the example really simple. I'm not trying simulate MWO. I'm trying to show that a good player will by virtue of being a good player will encounter more matches in which he has the better team. Even if he'll have the occasional game that is side against him.

The number of disconnects, afks and otherwise ineffective players doesn't really change this. They are just as likely to happen to either team. ECM is a bit trickier, if you bring a light withouth ecm you have a good chance to encounter one that does. However Hammers recent survey only showed a marginal advantage for the team with more ecm.

Now you're absolutely right about (voice)communication, teamwork, team composition, the mech and map combination and all those things. Al long with a players individual skill at piloting and aiming these really improve chances of winning. Now perhaps it's better to talk about rating effectiveness rather then player skill. Perhaps it is better to say that elo measures how good you are at winning games. And ultimately this is what the match maker tries to give people a 50% chance at.

View PostKrzysztof z Bagien, on 24 April 2013 - 04:12 AM, said:

Probability to win a game does not equal skill, and in fact one has very little to do with the other. Player skill will become even less important factor in the equation with introduction of 12v12.


Now you say this, but you don't explain why. Hell you already have trouble defining what player skill is. If you don't know what it is, how can you say it wont be a factor?

View PostKrzysztof z Bagien, on 24 April 2013 - 04:12 AM, said:

Edit: one more thing. Fore every player that says his experience improved thanks to Elo matchmaking there is one that says otherwise, so I belive it doesn't actually do anything and, as I mentioned earlier, it is merely a placebo effect.


I can look at my heavy stats and see that I'm not good at them. I can look at my medium stats and see I kick ***. I can't see my elo but I can make an educated guess after it. The difference between playing these two classes is like light day and night.

Edited by Hauser, 24 April 2013 - 05:30 PM.


#135 Kageru Ikazuchi

    Member

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

Posted 24 April 2013 - 06:29 PM

The good thing about all this is, the more we play, the better informed PGI is about the state of match making and the game overall ... while the opinions and observations of the vocal minority on the forums are being heard, the metrics and telemetry they're collecting are a reflection of our choices ... while the below quotes are not specifically germane to match making and Elo, they are important for understanding PGI's decision process ... from Bryan Ekman's reddit AMA:

http://mwomercs.com/...2013-730pm-est/
"We balance the wants, needs, and overall gameplay balance with what we see in the telemetry."
"Similar to how we use the k/d/movement heatmaps, we have telemetry that tells us how accurate weapons are, how often they are equipped, composition of each match in terms of mechs used etc.. This helps us determine what is popular, OPd etc.."

Some observations and opinions will differ from the average user experience ... let your voice be heard, but realize that just because it is what you saw or experienced, it may not be what everyone sees. I think I have played an above average number of games before Elo was recomputed (a week ago), and my experience since then has been different ... is it better or worse? I think it is probably too early to tell, but something changed ... I believe they're trying to make it better for everyone, and they're still trying to figure out exactly what that is.

#136 grayson marik

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • 1,436 posts
  • LocationGermany

Posted 24 April 2013 - 09:42 PM

Well, for one observation You do not need PGI's numbers:
Organized group play is extremely difficult with MWO but should have been addressed, since this is one of the communities that ran custom leagues for decades and is used to it.

But while it is not addressed, the teams that would play group vs group are syncdropping and if they cannot do 8 or 12 because not enough ppl online , they syncdrop smaller groups into the pug queue and therefore screw a good part of numbers and stomp a good part of pugs.

And this problem will get worse with 12v12! --> Not an assumption, not an observation --> a forecast!

Edited by grayson marik, 24 April 2013 - 09:46 PM.


#137 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 25 April 2013 - 12:34 AM

View Postgrayson marik, on 24 April 2013 - 09:42 PM, said:

Organized group play is extremely difficult with MWO but should have been addressed...


Really funny how they still claim its a team game, yet do nothing to improve game for groups ever since the whole "phase 1" started. Same time everything (including Elo matchmaker) is done to make game experience better for solo-puggers. Just shows what they really care about.

#138 Devil Fox

    Member

  • PipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 1,393 posts
  • LocationThe Fox Den

Posted 25 April 2013 - 01:51 AM

That's because they don't care about those that band together... but those that blow in and then out of this game spending a small bit of money...

They said something is coming but I don't hold high hopes for anything larger then 4man's, or any real incentive to not sync drop 8man teams as seems the current trend that PGI says is occurring... so all we have is their word, wonder how truthful that is...

#139 Krzysztof z Bagien

    Member

  • PipPipPipPipPipPipPip
  • Giant Helper
  • 710 posts
  • LocationUć, Poland

Posted 25 April 2013 - 04:26 PM

View PostHauser, on 24 April 2013 - 05:27 PM, said:

Now you say this, but you don't explain why. Hell you already have trouble defining what player skill is. If you don't know what it is, how can you say it wont be a factor?


I meant skill of individual player - the more people plays in single match the less relevant individual contribution is on average, I think that we can agree on that.

I really can't fully articulate my concerns with Elo system in English as it would require some mathematical terminology I only know in Polish and I'm not sure I can properly translate that without creating more confusion.



Here you can read about TrueSkill (Bayesian skill rating system which can be viewed as a generalisation of the Elo system used in Chess) developed by Micro$oft for their multiplayer games. Not like I'm great fan of them, but that work looks really solid from mathematical point of view.

It says:

Multiplayer online games provide the following challenges:
1. Game outcomes often refer to teams of players yet a skill rating for individual players is needed for future matchmaking. (and that's the main problem I see with Elo system being used in multiplayer games)
(...)
TrueSkil (...) addresses both these challenges in a principled Bayesian framework.


There's also a comparison to Elo system (they run tests in Halo 2 beta) - for 8vs8 games Elo system was inacurate in almost 40% of the games.
In fact Elo system was less acurate than Bayesian even in 1vs1 games!

As I understand it PGI implemented Elo system pretty straightforward, without any significant modifications. That's why I belive it can't do any significant change to matchmaking quality (even if it was implemented properly, and we can't be sure about that). Also, we have quite small playerbase (100k? maybe slightly more players total, and no more than 30k - 40k players at any given moment; and no, I don't have any data to prove it, it's just an educated guess) and in many cases there won't be enough players to be matched properly (skillwise) as tonnage would be (or is it already?) one of the factors matchmaker uses.

I hope I made myself clear :ph34r:

Edit: "I belive it can't", not "it can". One shouldn't think of solving complicated mathematical problems at 3AM.

Edit 2: also, funny fact - TrueSkill is patented. You can actually patent mathematical formula ;)

Edited by Krzysztof z Bagien, 26 April 2013 - 12:15 AM.


#140 Deathlike

    Member

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

Posted 25 April 2013 - 05:04 PM

You're saying that a player's contribution to teamwork is different from a player's contribution through skill? What an understatement!





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users