Jump to content

Matchmaker Actually Seems "fair"?!?


31 replies to this topic

#21 TexAce

    Member

  • PipPipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 2,861 posts
  • LocationGermany

Posted 03 October 2013 - 06:43 AM

on a side note here an interesting match from yesterday where our pug group won against what looks like a misplaces 12man or at least 2-3 4-man groups of steiners.

Posted Image

Uploaded with ImageShack.us

so it seems at least elo-wise the MM is working but you can't imagine the first shock we had when we saw that enemy list :D

Edited by TexAss, 03 October 2013 - 06:47 AM.


#22 PappySmurf

    Member

  • PipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • 842 posts

Posted 03 October 2013 - 06:44 AM

View PostMatthew Craig, on 02 October 2013 - 03:00 PM, said:

It's not that weight balancing isn't happening it is, actually over 50% of games on any given day are matched within 50 tonnes, but when your talking about 10s of thousands of games on any given day your also going to get a lot that look like what you've posted.

As mentioned earlier there are plans in the works to address this when lobbies are introduced with CW.


OMG I cant wait to just play private matches with my friends. Only question I have is why has it taken 2 years to implement lobbys when there were thousands of requests for them? :D

#23 HUBA

    Member

  • PipPipPipPipPipPip
  • 481 posts

Posted 03 October 2013 - 07:44 AM

It need more then only weight and elo to make a even match. The most important thing is the individual configuration of the mech. With enough xp and money you can make a mech twice as good as it's default variant. ELO helps here, because better player will mostly have better tuned mechs. Also you can match good player against good player which make it more interesting for both sides. A match new vs elite isn't fun, also for both sides.

IMO the best way would be a BV modified by elo. So the individual strength of the mech and player will be considered. AND a regimentation of the quantity for each class e.g. 1-3 light, 3-5 medium, 2-4 heavy and 1-3 assault, which could replace the weight regimentation.

#24 Electron Junkie

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 192 posts

Posted 03 October 2013 - 07:54 AM

View PostHUBA, on 03 October 2013 - 07:44 AM, said:

It need more then only weight and elo to make a even match. The most important thing is the individual configuration of the mech. With enough xp and money you can make a mech twice as good as it's default variant. ELO helps here, because better player will mostly have better tuned mechs. Also you can match good player against good player which make it more interesting for both sides. A match new vs elite isn't fun, also for both sides.

IMO the best way would be a BV modified by elo. So the individual strength of the mech and player will be considered. AND a regimentation of the quantity for each class e.g. 1-3 light, 3-5 medium, 2-4 heavy and 1-3 assault, which could replace the weight regimentation.


Yup this is pretty much what I meant... It's a pain in the rear to post to the forum when I have to go back and click the ready button all the time /joke... But yes There needs to be a "Standard" loadout for the company 2 Lights, 4 Mediums, 4 Heavies, 2 Assaults or the like.

#25 Aggressor666

    Member

  • PipPipPipPipPip
  • Survivor
  • 158 posts
  • LocationQuebec

Posted 03 October 2013 - 11:20 AM

View PostCurccu, on 02 October 2013 - 11:56 AM, said:

about every game is similar to this pic... numbers should be about right.

Posted Image

yup pretty much what 80% of matches look like altho sometimes far worse with 5 atlases facing 3 spiders 2 jenners and 2 ravens

Edited by Aggressor666, 03 October 2013 - 11:20 AM.


#26 Wintersdark

    Member

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

Posted 03 October 2013 - 12:09 PM

View PostPappySmurf, on 03 October 2013 - 06:44 AM, said:


OMG I cant wait to just play private matches with my friends. Only question I have is why has it taken 2 years to implement lobbys when there were thousands of requests for them? ;)


The answer is very easy, actually.

Lobbies have been high on the to-do list all along. However, as lobbies are entirely a UI construct, they'd have had to be developed on top of UI1.5. This would have been 100% "wasted" effort, as they'd need to be scrapped and re-built for UI2.0.

Further, the makeup of the lobby relies on some Community Warfare functionality, so delays in CW delayed the lobby as well

UI2.0 and Community Warfare was expected to be live some time ago, so back a while (keep in mind development of a major system like this takes months before you see it live) there seemed little value in delaying UI2.0 and CW (as they WOULD be the same teams working on a lobby) in order to build a construct they'd throw away and rebuild immediately afterwards.

In retrospect maybe they should have "wasted" that time, given how long UI2.0 and CW have taken to go live... But hindsight is 20/20, and even so... it would have delayed UI2.0 and CW even more.

#27 Advil

    Member

  • Pip
  • 12 posts

Posted 03 October 2013 - 02:30 PM

I'm pretty new here but already playing with friends in premade 4 man groups.

These are some interesting problems. I can see why people keep talking about the Lobbies.

Based on what we have all encountered so far in our first hundred games or so, getting the tonnages balanced should be #1 right now.

I can't see how this is going to get done without getting all 3 of these things going at once:
1) Lobbies to get lances organized
2) Mech Selection... during the ready phase so you can choose from at least your 4 ready mechs to get an appropriate tonnage/build for the upcoming map.
3) Tonnage Limits... so that the teams must conform to some kind of balance against each other.

It's hard to imagine doing any one of those without doing all of them at once or nothing gets accomplished.

#28 HumanDuracell

    Member

  • PipPipPip
  • Knight Errant
  • Knight Errant
  • 60 posts
  • LocationOn that hill over there, reporting your position to my teammates. They would SO love to meet you...

Posted 03 October 2013 - 04:26 PM

One possible resolution in factoring premades into PuG drops could be to limit the weight class of each lance similar to something that was in the 1st Edition of the Mercenaries handbook from TT Battletech.

There would be four classes of lance; Light, Medium, Heavy and Assault.
Each lance type would have a maximum and Minimum weight:

Light: 80 - 120 tons
Medium: 125 - 200 tons
Heavy: 205 - 280 tons
Assault: 285 - 380 tons

The weights are set in such a way that you cannot simply max out the tonnages in each weight class of 'mech.

Eg; a Light Lance cannot have four 35 ton 'mechs, a Medium lance cannot have four 55 ton mechs, etc.

This would mean that you would not be able to have 4 Atlases in a lance, though you could however have 3 Atlases and either an Awesome or a Victor (or a lower tonnage mech if you should so choose.)The players would be required to think about the lance composition.

When Matchmaker determines which players / premade teams will join a particular game, the matchmaker would use the lance type as the basis of tonnage matching. Each team could perhaps be composed of 1 Light or Medium lance, 1 Heavy lance and 1 Assault lance.
This would encourage players in premade teams to think about how to make the best use of the tonnage requirements of a particular lance type as exceeding the weight restriction would move the lance type up. (IE: You could have four Jenners in your lance, but due to the lance weight restrictions it would count as a Medium lance for matchmaking purposes, meaning you would be at a 60 ton disadvantage to your team.) The only exception to this would be assault lances, in which case you would not be able to start searching for matches while exceeding the weight restrictions.
Individual players would be arranged into the relevant Lance type by the Matchmaker.
This should result in a reasonably balanced match tonnage-wise when matchmaking. The only time where there would be a vast difference is if a premade team has gone over the threshold for the lance, knocking the lance type up one, (Light to Medium, Medium to Heavy, etc,) in which case the premade team have only got themselves to blame.

Example lance compositions.
Light Lance: 4x Spider. Total Tonnage = 120 tons
Medium Lance: 1x Victor, 1x Raven, 1x Jenner, 1x Hunchback. Total Tonnage = 200 tons
Heavy Lance: 1x Atlas, 1x Centurion, 1x Dragon, 1x Cataphract. Total Tonnage = 280 tons
Assault Lance: 2x Atlas, 1x Highlander, 1x Orion. Total tonnage = 365 Tons (15 ton disadvantage)

The beauty of this lance system is that the pre-made lances would be self regulated, (the players would not want to give a huge tonnage advantage to the opposing team,) allowing matchmaking to concentrate on the random player mech weights for balancing purposes. In addition it could be easily implemented in Community Warfare when it arrives:

Planetary assault? No problem! 2x Heavy and 1x Assault Lance. or alternatively; 1x Medium, 1x Heavy and 1x Assault Lance.
Recon or Quick strike? 2x Light and 1x Medium Lance or 1x Light and 2x Medium Lances.
Standard Engagement? 1x Light or Medium, 1x Heavy and 1x Assault Lance.

The Matchmaker would not need to struggle with factoring in a premade lance as the lance type has already been determined by the players, therefore it can use a preset figure:

Example
A pre-made Medium lance drops in with random players. The matchmaker assigns 200 tons for the pre-made team and proceeds to balance out the remaining tonnage with randomly dropping players using the Standard Engagement example above as a template.

Edited by Edanomel, 04 October 2013 - 12:59 AM.


#29 Rebas Kradd

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,969 posts

Posted 05 October 2013 - 07:25 PM

Anyone who doesn't know how to beat a premade of assaults (i.e. go around and basecap) is someone the devs shouldn't be wasting their time listening to.

Boosting cap rewards would be a terrific way to do passive weight-balancing. Teams would actually have incentive to basecap, and opponents would therefore have to maintain some balance. The rewards for engaging head-on are already there (i.e. the fun of killing big stompy robots); right now, basecap does not offer much incentive.

#30 Electron Junkie

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 192 posts

Posted 06 October 2013 - 11:28 AM

I think the problem with base capping can be solved very easily....

Change it from base capping to base DESTROYING!!! Just give the base enough hit points that a squad of lights will take half the match to do so... Problem solved.

Edited by Electron Junkie, 06 October 2013 - 11:29 AM.


#31 HUBA

    Member

  • PipPipPipPipPipPip
  • 481 posts

Posted 06 October 2013 - 01:52 PM

View PostEdanomel, on 03 October 2013 - 04:26 PM, said:

One possible resolution in factoring premades into PuG drops could be to limit the weight class of each lance similar to something that was in the 1st Edition of the Mercenaries handbook from TT Battletech.

There would be four classes of lance; Light, Medium, Heavy and Assault.
Each lance type would have a maximum and Minimum weight:

Light: 80 - 120 tons
Medium: 125 - 200 tons
Heavy: 205 - 280 tons
Assault: 285 - 380 tons

The weights are set in such a way that you cannot simply max out the tonnages in each weight class of 'mech.

Eg; a Light Lance cannot have four 35 ton 'mechs, a Medium lance cannot have four 55 ton mechs, etc.


The idea is good but there is a problem. Light mechs seems to be stronger as they tonnage wise should. Also is there a huge difference between a standard mech and a tire 2 mech with the same tonnage. The only way I see is a BV like this combined with ELO.

#32 Fabian Wrede

    Member

  • PipPipPipPipPipPip
  • 203 posts

Posted 06 October 2013 - 03:13 PM

so far all drops been totaly unbalanced, ELO MM is a big fail like any for of skill based mm is. The fact that new player gain to much ELO to fast make for very unbalance games . Also on conquest with the larger maps side with most lights alwasy wins.


MM must take Wight clsses into account better than it dose now.. MM was actually better when all that matterd was weight class





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users