Jump to content


New Matchmaking rule. Thoughts?

Matchmaking New MM

  • Please log in to reply
42 replies to this topic

zeur0sis #41 Posted 15 August 2019 - 06:57 PM

    First Sergeant

  • Players
  • 28811 battles
  • 1,530
  • [TITAN]
  • Member since:
    09-17-2016

View PostWouldChip, on 14 August 2019 - 10:24 AM, said:

I feel like most of you are missing the forest for the trees here. WG has carefully worded this to appeal to as many of you as possible. All I see in this is WR based MM that has been easily argued against for years. You want everyone to merge to an average WR? This is how it's done. If you don't see it, let me explain.

 

Let's say I start playing and over 4 games, I average 75% (3 out of 4). WG decides I'm winning too much and places me in Advanced queue. In Advanced queue, I win 25% over 4 games (1 out of 4), and now decides I need to go to the kiddie pool. I realize that these are not necessarily the number of games or WR that WG would use, but it's just an example for this illustration.

 

Now do the math. Using the above numbers, I would average 50% (4 games won out of 8). Extrapolate any period length you like, it will always come out to 50%. WG has just worded this differently to hopefully gain some acceptance for some new MM rules. If this is what you want, you're nuts.


wouldchip is right about this.   This is a clever form of winrate based mm.   It’s not as extreme as a straight up wr mm queue, but it’s going to slightly skew a percentage of winrates closer to the total average winrate over time.  


 

"Mm sucks just like u r face.    Now Go cry on tissue not here." ~ raj

 


twirlingtowardsfreedom #42 Posted 15 August 2019 - 07:48 PM

    Sergeant

  • Players
  • 29859 battles
  • 270
  • Member since:
    12-04-2014

View Postzeur0sis, on 15 August 2019 - 06:57 PM, said:


wouldchip is right about this.   This is a clever form of winrate based mm.   It’s not as extreme as a straight up wr mm queue, but it’s going to slightly skew a percentage of winrates closer to the total average winrate over time.  

 

There is undeniably some winrate based MM with this change, but we don't know enough to assume it will trend towards 50%. That will depend on how many queues there are, and the size of those queues. For instance, if there are only two queues - group A that has been beyond awful and relegated to tier 2 for X amount of battles, and group B that has just been better than 'beyond awful', then the size of the second will be massively larger and allow for a wide range of winrates. However, if there are three or even more groups like wouldchip says, and they are kept at similar sizes, then the winrates would absolutely begin to blend.



HugGigolo #43 Posted 15 August 2019 - 09:03 PM

    Senior Sergeant

  • Players
  • 21865 battles
  • 836
  • [AGILE]
  • Member since:
    07-28-2016

View PostEricOtown, on 14 August 2019 - 04:31 PM, said:


I wonder why doesn't the new MM apply to NA? Is WG testing it first on the other country servers? Should we expect NA to get the new MM algo sometime down the road? 

 

Last time WG changed MM I think it was limiting the number of tank types, but it only kicks in when the server pop is above 8k. Go look at how often this happens on NA now. If the new MM uses the same 8k threshold we'll rarely see the change anyway.







Also tagged with Matchmaking, New, MM

1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users