Jump to content

Matchmaker Adjustment - 06/05/2014


292 replies to this topic

#61 Jak Darren

    Member

  • PipPip
  • The Bludgeon
  • The Bludgeon
  • 33 posts

Posted 06 May 2014 - 06:55 PM

View PostNikolai Lubkiewicz, on 06 May 2014 - 03:47 PM, said:


I appreciate this suggestion. However, it's important to note that every player's definition of content is different.
For a great many players, and indeed for ourselves as developers, new Mechs do count as content just as much any other feature in the game we have added to, whether it's whole new facets like Private Lobbies or smaller additions like cockpit items inspired by the Community. With that in mind, I ask if you be more specific about what particular kind of content you would personally like to see more of. Thanks!



Taken from Dev Blog 0:

"The initial ideas revolved around revitalizing or rebooting the series by focusing on two areas:[/color]
  • Increasing the overall graphical fidelity of the entire game, especially the BattlMechs.
  • Solve some of the long standing gameplay issues from previous MechWarrior games.
From these two key areas, early design pillars emerged in the form of:[/color]
  • Urban Combat to address circle strafing and long range sniping.
  • Dynamic Battlefield further expanded on the concepts of Urban Combat and mechs having roles.
  • Dynamic pacing, introduced an early version of one of the final pillars, Information Warfare."



Taken from Official Developer Update, June 15th, 2012:

"We expect to have each of the features we have announced and discussed on the website (with one exception), in the game by the Open Beta date (which cannot yet be announced other than Summer 2012).
The one exception is the Community Warfare pillar which is a complex system but extremely important. In not wanting to delay the game, logic dictates it be added post-launch. Once fans are completely familiar with creating their 'Mech and pilot trees, the depth of Community Warfare will be added, with the core of the community experience projected to be in-game within 90 days of open beta."



I want your original design pillars.

I want roles. ACTUAL roles. I want Information Warfare. Not ECM bubbles.

I want Community Warfare. Something that was PROMISED to us by the END of 2012.

I WANT WHAT YOU SAID YOU'D DELIVER.

WE TRUSTED YOU. YOU CANT EVEN PATCH IN A LOBBY SYSTEM CORRECTLY.

THIS IS NOT MECHWARRIOR, THIS IS COUNTER STRIKE IN MECHS. YOU HAVE NOT ADDED AN *ACTUAL* GAME MODE FOR TWO YEARS. "Walk forward and shoot each other and pretend the base matters in a slightly different way" IS NOT A GAME MODE.

I CANNOT BELIEVE INVESTORS STILL HAVE FAITH IN YOUR ABILITIES.

Edited by Jak Darren, 06 May 2014 - 06:56 PM.


#62 Pariahsis

    Banned - Cheating

  • 5 posts
  • LocationLong Beach, California

Posted 06 May 2014 - 06:55 PM

View PostNikolai Lubkiewicz, on 06 May 2014 - 03:47 PM, said:


I appreciate this suggestion. However, it's important to note that every player's definition of content is different.
For a great many players, and indeed for ourselves as developers, new Mechs do count as content just as much any other feature in the game we have added to, whether it's whole new facets like Private Lobbies or smaller additions like cockpit items inspired by the Community. With that in mind, I ask if you be more specific about what particular kind of content you would personally like to see more of. Thanks!


Here are some ideas for new content that I would like to see instead of new 'Mechs:


1. A campaign with a story, like what previous MechWarrior titles had, to immerse players in Battletech lore, and with the option to play the campaign cooperatively with a lance of your friends online. This should be player(s) vs AI. I would even settle for a remake of a campaign from a previous Battletech title but with MWO’s updated graphics.


2. More game modes beyond assault, skirmish, and conquest. Here are a few ideas:

a. Capture the Flag (only ‘Mechs with hand actuators can pick up the flag)

b. Escort VIP(s)

c. POW Rescue

d. Raid (steal supply crates from the enemy base and return them to your base. Only ‘Mechs with hand actuators can pick up supply crates. On time out, team with most supply crates wins)

e. Defend and Destroy (The original Starsiege Tribes had this game mode. It’s like assault, but instead of standing on a capture point, players had to destroy different components of the enemy base while protecting their own base’s components from destruction)

f. Something with a dropship as a combatant.



3. More maps:

a. Mech Factory

b. Palace

c. Underwater

d. Orbital space station



#63 Smokeyjedi

    Member

  • PipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 1,040 posts
  • LocationCanada

Posted 06 May 2014 - 06:59 PM

Long search to find match, and than 2 minutes in everyone disconnects.......i was 3 or 4 to go........

#64 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 06 May 2014 - 07:07 PM

I want more new MAPS, (One map for every system in the Inner Sphere, he he) and then let Community Warfare begin "Content"!

Edited by OZHomerOZ, 06 May 2014 - 07:08 PM.


#65 Dan Nashe

    Member

  • PipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 606 posts

Posted 06 May 2014 - 07:11 PM

I think the problem was probably that the *more* people you have in queue, the *worse* it will get for people in anything but the least popular build. Add that to some MM inefficiencies and you're in trouble.

E.g.,it feels like the MM puts 12 people on one team, and then tries to find 12 to match them, instead of putting people in in pairs.

e.g., it will put an 80 and a 85 on the last two spots in a team when there's a 55 and 50 floating in queue, and not smart enough to swap the 80 for the 55 to balance the teams. But I don't know because MM logic is secret.

I actually doubt ELO is necessarily the problem. Maybe a contributor.

But anyone who plays wow knows that:
10,000 players run dungeons.
3 roles: tank, dps, healer
Need 1 tank, 1 healer, and 3 dps
Average wait time for a tank: 30 seconds
Average wait time for a healer: 4 min.
Average wait time for a DPS: 30 minutes.

I still don't understand how 3/3/3/3 avoids that problem.
Don't get me wrong, 3/3/3/3 sounds like an awesome match. It's one of the reasons I do leagues.
But if anything less than exactly 25% of your online population is playing each class, I don't know how you don't get everyone hung up on the least popular class, with the wait time getting longer as you add more people?

Granted, the cycle time is 10 min in MWO. So that is a fundamental difference from WoW, where new dps join the queue as tanks and healers leave, whereas in MWO, those tanks and healers all requeue in 10 min.

/shrug

I could be wrong though :-),

But if I'm right:
Limit premades to 3 of any weight class.
But allow the matchmaker to build teams of up to 4 of any weight class.
Just make sure both teams have the same.
That should release a lot of pressure if you can have 3/2/3/4 or whatever.
Additionally, that means if premades favor 3 assaults, there's still room in the queue for pugs in assaults.

The other possible problem someone else pointed out is:
84 people queue solo (84% of launches)
16 people queue doubles (8% of drops)
12 people queue triples (4% of drops)
16 people queue quads (4% of drops)

which means 84/128, or only 2/3 of the population is queueing solo. If the usual group only has 2-3 people, 75-80% of the population needs to be solo to fill up a 12 man with solos.

On the public test, I bet you had more solo's compared to 12s.
If 95% of the population drops solo, you can just let pressure off with 12 PuGs v 12 PuGs.

This is tougher to address, other than to make sure that if one team has a two groups of 3, the other team gets two groups of 3 or a group of 4 and a group of 2. /shrug.

But, this is all me making stuff up. I don't know how it actually works.

#66 Trauglodyte

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,373 posts

Posted 06 May 2014 - 07:16 PM

Quote

I appreciate this suggestion. However, it's important to note that every player's definition of content is different.
For a great many players, and indeed for ourselves as developers, new Mechs do count as content just as much any other feature in the game we have added to, whether it's whole new facets like Private Lobbies or smaller additions like cockpit items inspired by the Community. With that in mind, I ask if you be more specific about what particular kind of content you would personally like to see more of. Thanks!


Niko, content is different for everyone just as you said. What people want is to have fun - THAT is the end goal content. Everyone is growing tired of dropping into a match, most of the time which is greatly imbalanced in fire power and skill, only to have the game with no real reward except for space bucks. I want to know that taking Forrest Colony from the other side meant something. I want to know that winning a Conquest game for resources was important to something.
  • Fix the MM so people are playing together within actual acceptable parameters (weight and a stricter ELO)
  • Give us something for which to fight
If you want to know what frustrates people the most, it is another patch with nothing more to do than to play a Mech version of Duck Hunt. Fix that and you'll have people spending money in waves.

#67 Jun Watarase

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 3,504 posts

Posted 06 May 2014 - 07:19 PM

Quote

Our monitoring of the matchmaker immediately following the patch has revealed that wait times had more than doubled as a result of this fix being added without further changes to the matchmaker already in the pipeline.


Can you please clarify as to what the tolerances are for this? I would rather have 1 premade per team and closer games than a pointless 12-2 curb stomp. It usually takes me less than 30 seconds to get into a match, so doubling that to a minute or so isnt very bad compared to 5 minutes of curb stomping.

#68 T0ADKILLERD0G

    Member

  • PipPip
  • Overlord
  • Overlord
  • 37 posts
  • LocationMinneapolis, MN

Posted 06 May 2014 - 07:28 PM

View PostInRev, on 06 May 2014 - 04:01 PM, said:


Yes, because the best way to grow a healthy fanbase is to allow newbies or casuals to be farmed by metaslayers. :)

If "pros" were really so good, they wouldn't need to farm bads to enjoy the game. But we all know that a notch in the W column is all that really matters, ignore all the rest.

I'm no fan of the way Elo is implemented in this game (because it's horrible and doesn't really work anyway), but blatant complaining is far worse than even a failed attempt to make things fair. This is a game, not real warfare, and definitely not a job. Leave the rest at the door, please.


Completely agree

#69 T0ADKILLERD0G

    Member

  • PipPip
  • Overlord
  • Overlord
  • 37 posts
  • LocationMinneapolis, MN

Posted 06 May 2014 - 07:32 PM

View PostTrauglodyte, on 06 May 2014 - 07:16 PM, said:


Niko, content is different for everyone just as you said. What people want is to have fun - THAT is the end goal content. Everyone is growing tired of dropping into a match, most of the time which is greatly imbalanced in fire power and skill, only to have the game with no real reward except for space bucks. I want to know that taking Forrest Colony from the other side meant something. I want to know that winning a Conquest game for resources was important to something.
  • Fix the MM so people are playing together within actual acceptable parameters (weight and a stricter ELO)
  • Give us something for which to fight
If you want to know what frustrates people the most, it is another patch with nothing more to do than to play a Mech version of Duck Hunt. Fix that and you'll have people spending money in waves.




As a person who loves the books this is what i want...

#70 Appogee

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 10,966 posts
  • LocationOn planet Tukayyid, celebrating victory

Posted 06 May 2014 - 07:40 PM

View PostNikolai Lubkiewicz, on 06 May 2014 - 03:47 PM, said:

I ask if you be more specific about what particular kind of content you would personally like to see more of.

1. Maps.
2. Game Modes.
3. Community warfare.

We have way more Mechs and cockpit items than needed, but nowhere near enough maps, and still no community warfare. Some additional game modes would be good too.

I realise maps, game modes and CW are being worked on. However, the fact that new Mechs and cockpit items arrive with amazing rapidity, and yet it's been 5 months since the last map, and 18 months since CW was first supposed to be delivered, shows that PGI isn't putting enough of its limited resources where they are most urgently needed.

Edited by Appogee, 06 May 2014 - 07:49 PM.


#71 Mister Blastman

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 8,444 posts
  • LocationIn my Mech (Atlanta, GA)

Posted 06 May 2014 - 07:52 PM

This isn't rocket science, guys. What is going on? With 84%+ of players... pugging... what's the deal? ...........

#72 Cimarb

    Member

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

Posted 06 May 2014 - 08:06 PM

MATCHMAKER:

Dump 3/3/3/3 and 1 group per side, as they obviously aren't working, and instead just balance each person/group with an equivalent on the other team.

Two 4-mans of equivalent Elos? One on each team.
Two 6-mans? One on each team.
Fill in the blanks with solos, separated equally by Elo on each team.

Launch. Rinse. Repeat.

CONTENT:

Communication is all we need.

For instance, WHAT is going on with CW? We haven't heard diddly squat for months. No command chair updates. No tidbits to whet our appetites. Nothing.

What happened with the two maps that were rumored to be coming soon? What is the team working on, what are the hold ups, etc....

I appreciate all the work you guys are putting into the game, but without communication at least as good as Karl has been doing, you aren't going to get adoring fans any time soon.

Just have people take turns spending an hour a day (so 5-7 total man hours amongst your whole crew each week) answering questions on the forums.

Ask Karl how his thread has been going - lots of complements, to this day, and almost zero vitriol for its entire existence. Now look at the hundreds of other threads like it where no staff have communicated on it - the exact opposite. It really isn't rocket science.

#73 VXJaeger

    Member

  • PipPipPipPipPipPipPipPip
  • The Wrath
  • The Wrath
  • 1,582 posts
  • LocationFinland

Posted 06 May 2014 - 08:08 PM

A fine example how Excel-leadership works.... IT DOESN'T.

#74 NinetyProof

    Member

  • PipPipPipPipPipPipPip
  • 547 posts
  • LocationSan Diego, CA

Posted 06 May 2014 - 10:05 PM

3 + 3 + 3 + 3 .... plus "1 group minimum" ... plus ELO banding = Dead On Arrival ... if you want any sort of decent queue times. It's not rocket science.

Not to mention considering a "Awesome" as equal to a "Banshee / Atlas" is the definition of "imbalance".

3 + 3 + 3 + 3 is noble, but naive. Try just using "weight banding" plus tonnage instead. You can still have artificial 3 + 3 + 3 + 3 ... but maybe the ranges are sometimes only 10-15 tons apart and not some arbitrary edict.

4 Locust + 4 Cicada + 4 Dragons + 4 Awesomes is no match for 4 Jenners + 4 Shadow Hawks + 4 CTF + 4 Atlas is in now way "balanced" even though it's 3 + 3 + 3 + 3 on each side. Yes, that is an edge case, but the point stands.

Yes, the idea of the 4x3 balance is ideal, but unfortunately, the chassis in the weight classes are not ideal ... really should be more like 6 weight classes when you consider the effectiveness of all the mechs. All your doing PGI is sticking your head in the sand on this one. Do it well, or don't do it at all.

Weight banding is the way to go ... incredibly fast queues ... people get to play pretty much what they want, and you get at least some semblance of balance. Yes, that means there might be matches without lights, or maybe matches without assaults ... or some variation, but as long as you have bands of weights, you will get some form of balance, which is what is needed.

I don't know why I am wasting my breath ... just going to fall on deaf ears.

#75 DeathlyEyes

    Member

  • PipPipPipPipPipPipPip
  • The Messenger
  • 940 posts
  • Google+: Link
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationMetaphorical Island somewhere in the Pacific

Posted 06 May 2014 - 10:56 PM

Let me just point out one last thing, the whole point of ELO match making is to try to mitigate 12-0 stomps. 12-0 stomps occur almost always it seems. I seriously feel the issue has to do deal with game mechanics and the matchmaker. It seems once one team gets a slight advantage numerically the other team implodes. Respawn/dropship mode would alleviate this. In addition since the 12-0 stomps seem to happen regardless perhaps mitigating the affects of ELO or eliminating it entirely while you get your new matchmaking parameters to work would have minimal affect on the games being played anyway.

You gotta remember the people who don't group up drop way more frequently. When they die they leave the match almost immediately because they have no reason to sit around and wait. Meanwhile premade groups have to wait on their fellow members to finish match. Solo players launch far more often whereas premade groups have to wait for all members to be readied up. Most serious players play in groups The easiest way to raise your to raise your win loss ratio and consequently your ELO is playing in groups. This means most people in premade groups tend to be higher ELO. So you have an over-represented minority dropping individually against an underrepresented minority or apparent minority and you are stacking that over-represented lower ELO group and hoping your matchmakers constraints can find teams. The fact is there are likely a lot less solo players in queue than you think and fewer so with ELOs that meet constraints to matched with the premades.

Edited by SLDF DeathlyEyes, 06 May 2014 - 11:07 PM.


#76 Karl Streiger

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • Wrath
  • Wrath
  • 20,369 posts
  • LocationBlack Dot in a Sea of Blue

Posted 06 May 2014 - 11:07 PM

12:0 stomps have nothing to do with ELO.
Hell we have tested the private module
9 on 9, 6 on 6
and usually it was 9:1 or 6:2 or vice versa - and of course it did change. 1:6; 6:1 ....

I have yesterday played with a premade group of 4 - we always found a Match instantly. We where always the Alpha Lance.
I believe it has to do something with ELO - i for example have a really really worse ELO - it would be hard for you to find somebody with a worser - I'm pretty sure the other guys in my team had a W/L of 1 or above...not 0.6.
But the skill and abillity of the other players was purely random - i have seen a ProTeam on the other side.
Have seen really newbies in Champion Mechs - and of course some "full mastered" Mechs.

Anyhow non of this battles was a stomp: always 12:6; 12:8 but believe it or not - i would take most 12:1 or 12:3 every time over those games yesterday after the patch

#77 Javenri

    Member

  • PipPipPipPipPip
  • Bad Company
  • Bad Company
  • 171 posts
  • LocationAthens, Greece

Posted 06 May 2014 - 11:17 PM

My suggestion for public matches hasn’t changed and I will repeat it again:

Premade groups. Balance premade groups by the players in them. Allow more than four players in a premade group and more than one premade group per team. The total number of players in premade groups should not exceed 1 player between the two teams (1 player doesn’t make a huge difference in game play but allows more options to the matchmaker). For example if one team has 6 players in premade groups regardless of combination (4+2/3+3/6), it will be matched against a team that has 5/6/7 players in premade groups (again in any possible combination). The rest of the spots will be filled with solo players. Optionally, a limit to maximum 8 players in premade groups for each team can be applied. Although it is open for debate, I believe that a premade 8-man will not be vastly superior to a 4+3 premade groups (assuming all have voice communications on lance level).

Weight limits. I do like the 3/3/3/3 rule. I know some people disagree, but I like it. In the case though, that PGI decides that this model will not work well, I would like to suggest weight limits in combination with a restriction of up to 5 mechs per class (meaning no more than 5 heavy or 5 lights, etc). My proposed weight limit is the 700±50 tons per group. The number is derived from the sum of the average of each class multiplied by 3 (697.5 rounded to 700). For example lights range from 20 to 35 tons, their average is 27.5 tons. That will result to weight between 650-750 and no more than 5 mechs from a single class. This solution would solve the weight issues without being too restrictive or lax.

Edited by Javenri, 06 May 2014 - 11:20 PM.


#78 Sean von Steinike

    Member

  • PipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 2,880 posts
  • LocationUSA

Posted 06 May 2014 - 11:40 PM

View PostIqfish, on 06 May 2014 - 03:51 PM, said:

QFT.

People jump on the "I want more content" bandwagon really quick, but how do they define new content?

New Maps? New weapons? New Game Modes? All of those?

How about the stuff that we were led to believe we would have...dropship mode, community warfare, the reintroduction of knockdowns?

#79 PhoenixFire55

    Member

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

Posted 06 May 2014 - 11:46 PM

People told you wait times will be too long PGI, but you chose not to listen ...
Now you are gonna loosen up the boudaries for MM that'll lead to same old uber disbalance ...
3/3/3/3 is DOA ... thats Dead On Arrival in case you don't know.

So ... can we please get rid of ELO ... thats Egoistic Looser Obsessiveness in case you didn't know.
And while you are at it remove group limits that are hindering this game for 1.5 years.

#80 mrbash

    Member

  • PipPip
  • The Patron
  • The Patron
  • 27 posts

Posted 06 May 2014 - 11:55 PM

View PostNikolai Lubkiewicz, on 06 May 2014 - 03:47 PM, said:


I appreciate this suggestion. However, it's important to note that every player's definition of content is different.
For a great many players, and indeed for ourselves as developers, new Mechs do count as content just as much any other feature in the game we have added to, whether it's whole new facets like Private Lobbies or smaller additions like cockpit items inspired by the Community. With that in mind, I ask if you be more specific about what particular kind of content you would personally like to see more of. Thanks!


Niko, your post is astounding for its complete tone-deafness to the community. I am not sure if you are simply trying to antagonize the player base or if PGI is really this out of touch with what the MWO players are want. I am going to assume the later and present my expectations:
  • Maps. After 3 years all PGI has put out is 10 maps. If the company is unable to produce new maps why not allow for community submissions for maps? Or allow 'paying' clients to use custom maps?
  • Game Modes. 3 Game modes, of which 2 are nearly identical? Again,if the company is unable to create new game modes, allow the community to do so. Why not let people in private lobbies configure custom win conditions.
  • In-Game coms. Fixing this will solve many problems, and will both help and encourage new players.
Let me stress, that these are 'expectations' not 'nice to have'. To date the majority of the 'content' that has been produced has simply been 'aesthetic' or 'derivative'. Adjusting the LRM speed does change the meta-game but no one outside of a hard-core player would call that 'Content'. Similarly, having a new 10$ cockpit item is hardly considered 'Content'. Are these items the yardstick you want the gaming community to use in comparing MWO against other games?

Edited by mrbash, 07 May 2014 - 12:05 AM.






1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users