Jump to content

Matchmaking Phase 4 Im Anmarsch


45 replies to this topic

#1 RapierE01

    Member

  • PipPipPipPipPipPipPipPip
  • 1,576 posts
  • LocationEden

Posted 12 April 2013 - 02:38 PM

Quote

Posted Today, 08:13 PM
Posted Image
POPULARIt's been two months since our last major update to matchmaking, where we introduced our first pass at Elo and saw a tremendous improvement in the quality of matches being kicked off. With the next patch we're bringing in a new round of changes, and we've got two big ones lined up: Weight Class Matching and Elo Rating Adjustment.


WEIGHT CLASS MATCHING

We're adding the ability for us, the developers, to adjust our weight matching tolerances on the fly. This is giving us a controlled way to tighten up the weight imbalance between teams that players are experiencing.

Fun fact: About 74% of games kicked off are within what we consider tolerable weight matching limits (the difference between a heavy vs. assault Mech on the opposing team). Of the remaining 26% of matches, about 7% are what we would call "horrendously bad" (i.e., they carry the weight difference of one or two full assault Mechs between teams).

When we tighten up the Weight Matching, we're going to drop our weight tolerance to 0 and see what happens. Teams will be almost equally built in terms of weight class, though they will draw from a wider skill range of players to compensate. We'll be keeping an eye on the system while doing this and will continue to tune it until we're happy with the match quality and wait times.


ELO RATING ADJUSTMENT

We discovered a bug that was causing player Elo values to become inflated. Players that would win a game against a weaker opponent were being rewarded as though they were the underdog in that match. This caused Elo values to increase by a lot when they should have increased by only a little, skewing the player Elo distribution so that it became top-heavy.

What we've done (beyond correcting the bug) is a complete re-simulation of Elo ratings for every player in the game. Using match history we've been logging since February 5th, we determined what every players Elo rating would have been if we were using the corrected algorithm. The difference we noticed is huge, and the new player Elo distributions look healthy. We're going to push these ratings to the servers when we patch on April 16th.

So, how will this affect things? Well, without going into the details of how Elo works, this significantly improves our ability to match similarly skilled players. The old algorithm had an unintended (although interesting) bias towards people who had played more games, whereas the new rating measures only how you're playing. The bottom line is that we can kick off more close matches.


ROLLOUT PLAN

So when will we get to see this stuff? You won't have to wait long; this goes live with the April 16th patch. Other fixes, such as better detection of disconnected players while matching are also going in at this time as well. However, we'll be staging the release of the Weight Matching tweaks to be sure the system is working correctly (and to give us time to collect a bit of data on the new operating conditions).

Here's the plan:
On Tuesday, April 16th, we'll patch as normal and will put in the adjusted Elo values.
On Thursday, April 18th (tentative date), we'll drop the weight matching tolerance to zero. This won't require a patch or an interruption of service.

We're excited to bring the next round of changes live and look forward to matching up with you. See you on the Battlefield!


Mal schauen wie es wird. Im Thread dazu wird schon wieder rumgeheult noch bevor es online ist. Ich denke es wird die ganze Sache verbbessern denn so sind 4 DDC + Kleinkram gegen 4 Cataphract + Kleinkram net mehr möglich und es wird die Imbalance etwas aufheben

Edited by RapierE01, 12 April 2013 - 02:40 PM.


#2 Quardak

    Member

  • PipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 1,301 posts
  • LocationRaumsystem Kitzingen

Posted 13 April 2013 - 01:58 AM

Jo ersmal abwarten. Es klingt doch schonmal gut, auf alle besser als was wir haben.

Ob die Daumenschraube dann zu eng ist und die Matches zu lange dauern um gefunden zu werden... wir werden sehen.

Interessanter find ich den Hinweis auf die Fehlberechnung! Endlich darf ich wieder gegen die Noobs spielen die genauso schlecht sind wie ich :)

#3 Vetrecs

    Member

  • PipPipPipPipPipPipPip
  • The Hammer
  • The Hammer
  • 786 posts
  • LocationGei Fu

Posted 13 April 2013 - 02:04 AM

View PostQuardak, on 13 April 2013 - 01:58 AM, said:

Endlich darf ich wieder gegen die Noobs spielen die genauso schlecht sind wie ich :)

Als Crashdummy in der Test-Area? :ph34r:
Verdammt, dann bin ich nicht mehr alleine ;)

#4 EvangelionUnit

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 776 posts
  • LocationWarframe

Posted 13 April 2013 - 03:19 AM

mal sehen was dann für ein wurm im system ist ... wenn viel spieler wieder mit unmöglich hohen ratings abgestraft werde ?
vllt sehe ich dann wieder mehr leute in trail mechs ...

#5 Quardak

    Member

  • PipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 1,301 posts
  • LocationRaumsystem Kitzingen

Posted 13 April 2013 - 06:25 AM

View PostVetrecs, on 13 April 2013 - 02:04 AM, said:

Als Crashdummy in der Test-Area? :P
Verdammt, dann bin ich nicht mehr alleine :ph34r:


Man wäre das geil wenn man da in einer der "Bots" reinspringen könnte und den Tester erschrecken!

Was würd ich doof gucken wenn sich plötzlich der Atlas drehen würde und mir nen Alpha auf die Nase haut :( :D

#6 EvangelionUnit

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 776 posts
  • LocationWarframe

Posted 13 April 2013 - 07:24 AM

eine voll speed im kreis laufen routine wäre schon toll für die mechs im training ground, so kann man sehen wie die waffen treffen "würden" ... schließlich bleibt kaum jemand länger als ne notabschaltung stehen (ausnahme sniper)

#7 w0rm

    Member

  • PipPipPipPipPipPipPipPipPip
  • Giant Helper
  • 2,162 posts

Posted 13 April 2013 - 09:39 AM

Die Änderungen an der ELO-Berechnung sorgen hoffentlich dafür das im Top-Tier die Matches besser werden.

#8 EvangelionUnit

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 776 posts
  • LocationWarframe

Posted 13 April 2013 - 10:53 AM

View Postw0rm, on 13 April 2013 - 09:39 AM, said:

Die Änderungen an der ELO-Berechnung sorgen hoffentlich dafür das im Top-Tier die Matches besser werden.

klar, weil ich dann nicht mehr dabei bin!
(ich spiel so schlecht >_<)

Edited by EvangelionUnit, 13 April 2013 - 10:53 AM.


#9 Garagano

    Member

  • PipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 862 posts
  • LocationGermany

Posted 13 April 2013 - 02:44 PM

View Postw0rm, on 13 April 2013 - 09:39 AM, said:

Die Änderungen an der ELO-Berechnung sorgen hoffentlich dafür das im Top-Tier die Matches besser werden.

dann bleiben die ecm-light-cap-premades endlich unter sich *duckundwegrenn*

#10 Vincent Lynch

    Member

  • PipPipPipPipPipPipPipPip
  • Shredder
  • Shredder
  • 1,652 posts
  • LocationVienna

Posted 13 April 2013 - 03:59 PM

Ich hab jetzt keine Lust, das im Original-Forum zu lesen.
Aber: wer zur Hölle wird da whinen?
Wenn sie es genau so hinkriegen wie hier beschrieben, ist das doch das Beste für alle Beteiligten, davon hat nun wirklich niemand einen Nachteil!

#11 EvangelionUnit

    Member

  • PipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 776 posts
  • LocationWarframe

Posted 13 April 2013 - 04:18 PM

View PostVincent Lynch, on 13 April 2013 - 03:59 PM, said:

Ich hab jetzt keine Lust, das im Original-Forum zu lesen.
Aber: wer zur Hölle wird da whinen?
Wenn sie es genau so hinkriegen wie hier beschrieben, ist das doch das Beste für alle Beteiligten, davon hat nun wirklich niemand einen Nachteil!

na ja außer sie matchen 4 player premade mit public only playern ...

#12 Garagano

    Member

  • PipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 862 posts
  • LocationGermany

Posted 14 April 2013 - 01:53 AM

Das gewhine bezieht sich u.a. auch darauf, dass wieder nur Gewichtsklassen zugeordnet werden, also so wie schon vor ELO.
Das war aber schon damals nur suboptimal.
Besser ist eine (mehr oder weniger) exakte Tonnage Zuordnung beider Teams!
Aber PGI scheint das nicht so richtig einführen zu wollen... hoffentlich ist das nur eine vorübergehende Lösung und die Tonnage Angleichung kommt noch vor dem offiziellen Release.

View PostOmid Kiarostami, on 12 April 2013 - 11:44 AM, said:


To clarify, it'll be similar to the forced Weight Class balancing that we used to have. We're not doing any tonnage based matching at this time.


#13 Iqfish

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,488 posts
  • Google+: Link
  • LocationGermany, CGN

Posted 14 April 2013 - 02:08 AM

Irgendwie ist der Post doch widersprüchlich oder?
Erst sagt er:

"When we tighten up the Weight Matching, we're going to drop our weight tolerance to 0 and see what happens. Teams will be almost equally built in terms of weight class, though they will draw from a wider skill range of players to compensate."

Worunter Ich soviel verstehe wie:

"Wir werden die Teams mit einer Gewichtstoleranz von 0 zusammenstellen, heisst 500 vs 500 Tonnen 370 vs 370 Tonnen"

Und dann sagt er:

"To clarify, it'll be similar to the forced Weight Class balancing that we used to have. We're not doing any tonnage based matching at this time."

Worunter Ich das genaue Gegenteil von Obrigen verstehe...

#14 Zerberus

    Member

  • PipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 3,488 posts
  • LocationUnder the floorboards looking for the Owner`s Manual

Posted 14 April 2013 - 02:13 AM

Sprich ab Dienstag werden die ganzen scheinheiligen "Ich bin einfach gut weil ich weiß daß ich einfach gut bin" möchtegern-Pro-Gamer wieder verstärkt gegen Lanzen aus Trial mechs droppen, und ab Donnerstag haben sie die Tonnage auch nicht mehr auif ihrer Seite?? This could get interesting. :P:D

#15 RapierE01

    Member

  • PipPipPipPipPipPipPipPip
  • 1,576 posts
  • LocationEden

Posted 14 April 2013 - 02:18 AM

@ IQ ich vermute das soll heissen das nicht wenn du drei Atlas hast die auch 3 Atlas haben. Sondern das nur die Gesammttonnage stimmig ist und die Mechklassen zusammenpassen. (z.b: Awesome + Cataphract + Hunchback vs: Stalker + Jägermech + Centurion)

Viel besser finde ich das net mehr 4 Atlas + Kleinkram drops vs Medium only gematched werden können

Edited by RapierE01, 14 April 2013 - 02:22 AM.


#16 Iqfish

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,488 posts
  • Google+: Link
  • LocationGermany, CGN

Posted 14 April 2013 - 02:24 AM

View PostRapierE01, on 14 April 2013 - 02:18 AM, said:

@ IQ ich vermute das [...]


Das ist ja das Problem an dieser Firma, selbst wenn sie eine Ankündigung machen, ist diese so unklar formuliert, dass weitere Fragen aufgeworfen werden.
Mir persöhnlich ist ELO eigentlich egal, Ich hatte erst ein, zwei mal ein unfaires Match (aus über 1500)
Ich hoffe eher, dass das Spiel mit 8-Mann mal wieder spielbar wird, ohne das zwei Leute von Anfang an wegen Bugs, Disco's oder Abstürzen ausfallen...

#17 w0rm

    Member

  • PipPipPipPipPipPipPipPipPip
  • Giant Helper
  • 2,162 posts

Posted 14 April 2013 - 03:01 AM

View PostGaragano, on 13 April 2013 - 02:44 PM, said:

dann bleiben die ecm-light-cap-premades endlich unter sich *duckundwegrenn*


chapeau!

#18 Quardak

    Member

  • PipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 1,301 posts
  • LocationRaumsystem Kitzingen

Posted 14 April 2013 - 09:08 AM

Ligths sind ja auch pöse! Hatte aus Testzwecken einen 4er Drop mit 4x Jenner....

Gegner 3x ECM-Raven 3L und 1 Jenner... Ergebnis 4:0 für das Premade-Team.

Fakt 1: Wenn was leichtes gespielt wird ist es meist Raven 3-L oder Jenner (Commando+Spider sind schon selten)

Fakt 2: TS ist Random-ECM überlegen. Gegen TS+ECM wären wir sicher gnadenlos untergegangen.

Und bevor sich wer aufregt <_< ... wir haben es nur eine Runde gemacht weil es uns zu "lame" war :D

Edited by Quardak, 14 April 2013 - 09:09 AM.


#19 Garagano

    Member

  • PipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 862 posts
  • LocationGermany

Posted 14 April 2013 - 10:29 AM

Ich hatte heute schon 3 verschieden gaycapecm premades als Gegner. Widerliches Pack!
Das eigene team hatte natürlich jedes mal 0 ECMs.

Das traurigste an der ganzen MatchMaker Geschichte ist ja, dass PGI es ernsthaft nicht für notwendig hält, eine ECM balance herzustellen. o0
Dabei ist das ja IMHO noch wichtiger als eine tonnage gleichheit.

#20 Oberhupe

    Member

  • PipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 140 posts
  • LocationBerlin

Posted 14 April 2013 - 08:55 PM

View PostGaragano, on 14 April 2013 - 10:29 AM, said:

Ich hatte heute schon 3 verschieden gaycapecm premades als Gegner. Widerliches Pack!
Das eigene team hatte natürlich jedes mal 0 ECMs.

Das traurigste an der ganzen MatchMaker Geschichte ist ja, dass PGI es ernsthaft nicht für notwendig hält, eine ECM balance herzustellen. o0
Dabei ist das ja IMHO noch wichtiger als eine tonnage gleichheit.


Sorry aber das kann ich so überhaupt nicht überschreiben.
ECM ist ein taktischer Vorteil aber nicht mehr. 50T mehr Kampfgewicht ist ein größeres Problem weil man die nicht abgeschält bekommt.

Ich sehe auch den Riesenvorteil von ECM nicht mehr. Anschleichen kann durch Scouts verhindert werden. LRMs feuern mit Tag einwandfrei. Streaks sind nur ein paar Schadenspunkte und wenn alle die DInger haben, dannn muss man halt die ECMler ausknipsen.

An dieser Stelle geht PGI mMn den richtigen Weg.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users