Matchmaker Adjustment Idea
#1
Posted 02 December 2020 - 08:11 PM
The idea is to remove groups from the priority in the MM and instead add a handicap adjustment to pilot skill rating and mech tonnage used by the MM based on group size. Could still limit max number of people in groups on a team as done today.
Part of the thinking here is that, if you are in a group of skilled players you want to make that factor in more than a group of noobs. This still takes into account if you are in a group, just not as its own step. It also allows pilot skill rating and mech tonnage to be more impactful in the MM like previously.
So for example, if your pilot skill rating was 1000 and in a group of 3, adjust the rating to 1300 (30%), but if you had only rating of 100, only adjust to 130. If you were in a group of 4 and mech tonnage of 50, treat it as if it were 70 instead (40%), or tonnage 100 would be 140, etc.... The exact numbers aren't important, just examples, they could be changed as needed to get the desired balance.
#2
Posted 02 December 2020 - 11:00 PM
#3
Posted 03 December 2020 - 06:48 AM
MM1: Pulls solo players to make a game. If runs out of pugs it starts adding groups.
MM2: Pulls groups to make a game. If there are any gaps and no suitably sized groups to fill them it adds pugs.
This way group players primarily play groups and pugs primarily play pugs. It's only when the queues get low that the two get mixed.
Edited by VonBruinwald, 03 December 2020 - 06:48 AM.
#4
Posted 03 December 2020 - 07:40 AM
The problem is the matchmaker treats 4 top 1% players in a group the same as 4 cadets in a group. Right now the system matches a group of low PSR players against a group of high PSR players and thinks it's making a good match
A matchmaker that looked at PSR as the primary factor rather then group vs non group would work better:
1 Group of low PSR individuals - gets matched against low PSR opponents group or not group.
2. Group of very high PSR players - gets matched against high PSR opponents group or not group.
Change the primary matchmaking factor and I really believe the rest of the metrics will fall into place. That may mean one side gets two groups and the other none which may certainly cause people angst but I think would result in better matches then way it handles groups currently.
Edited by GARION26, 03 December 2020 - 12:28 PM.
#5
Posted 03 December 2020 - 09:58 AM
VonBruinwald, on 03 December 2020 - 06:48 AM, said:
MM1: Pulls solo players to make a game. If runs out of pugs it starts adding groups.
MM2: Pulls groups to make a game. If there are any gaps and no suitably sized groups to fill them it adds pugs.
This way group players primarily play groups and pugs primarily play pugs. It's only when the queues get low that the two get mixed.
except that groups have no interest in playing against other groups. It's why group queue died in the first place.
Edited by Knight Captain Morgan, 03 December 2020 - 10:03 AM.
#6
Posted 03 December 2020 - 10:11 AM
Knight Captain Morgan, on 03 December 2020 - 09:58 AM, said:
GQ died because the MM kept putting the same two teams together where one team always won and the other always lost. That's not fun. People want a fair challenge and a fair Matchmaker. https://mwomercs.com...hmaker-is-king/
Edited by Nightbird, 03 December 2020 - 10:11 AM.
#7
Posted 03 December 2020 - 04:52 PM
GARION26, on 03 December 2020 - 07:40 AM, said:
The problem is the matchmaker treats 4 top 1% players in a group the same as 4 cadets in a group. Right now the system matches a group of low PSR players against a group of high PSR players and thinks it's making a good match
A matchmaker that looked at PSR as the primary factor rather then group vs non group would work better:
1 Group of low PSR individuals - gets matched against low PSR opponents group or not group.
2. Group of very high PSR players - gets matched against high PSR opponents group or not group.
Change the primary matchmaking factor and I really believe the rest of the metrics will fall into place. That may mean one side gets two groups and the other none which may certainly cause people angst but I think would result in better matches then way it handles groups currently.
OP may have been too verbose. Yes, my idea is to use PSR as the main factor. You just treat the PSR as a higher value if in a group.
I mentioned tonnage can be adjusted, but it could just be PSR if that makes more sense. I meant this to be flexible, so it could be adjusted as needed including amount to adjust by.
Edit, I mentioned that you could still try to limit number of groups to a team, but PSR was still intended to be the #1 factor.
Edited by ParticleProjector, 03 December 2020 - 04:55 PM.
#8
Posted 03 December 2020 - 05:01 PM
Knight Captain Morgan, on 03 December 2020 - 09:58 AM, said:
except that groups have no interest in playing against other groups. It's why group queue died in the first place.
This is absolutely false. GQ barely ever lived and eventually died because it never had a functioning matchmaker thanks to no "groups of 1" being there to fill the gaps. For a match to launch you needed a combination of group sizes that adds EXACTLY to 12 not just once but twice, which made waiting times long. 7+ 6 is one too many. 8+3 is one too few. Can't even make a different arrangement out of those 4 groups. Once there is one team of 12 put together, you still need for a group of correct size to magically appear to round out the other side, and woe to those whose group was of the wrong size to get matched. They could wait for hours.
To make the matter worse, there was no information presented to queuing players about if there is anyone else in the queue or are they waiting alone hopelessly. People eventually learned to assume that they are alone, and gave up trying, and many gave up playing MWO.
Edited by Gagis, 03 December 2020 - 05:03 PM.
#9
Posted 03 December 2020 - 07:24 PM
The Problem with Groups you have in each PvP ,whe n you drop in CoD warzone as Singleplayer ,and confronted with Teams thats worked with ´Teamspeak Coordinated is a Feature of the Game ,same by War Thuner or each Other PvP Game ...when you not will fight against groups you must play MW5 or Solaris,its the simple True ...you will not biting ,swim not in a Shark pool
#10
Posted 03 December 2020 - 08:46 PM
Any "solution" that means longer queue times just means more people like me exit the game and play something else too.
#11
Posted 03 December 2020 - 09:19 PM
crazytimes, on 03 December 2020 - 08:46 PM, said:
Any "solution" that means longer queue times just means more people like me exit the game and play something else too.
I don't see this as an "exotic" system. We are talking about multiplying a number that is already being used by the MM by a constant.
I don't see how this would increase que times. Having a higher effective PSR doesn't necessarily mean that the 24 people in a match couldn't be the same, but might better distribute those people across the 2 teams for a potentially better experience which is what it tries to accomplish.
Edited by ParticleProjector, 03 December 2020 - 09:20 PM.
#12
Posted 03 December 2020 - 10:31 PM
This topic is similar to what is discussed in
Mechwarrior Online 2021: Modes
Daeron Katz, on 16 November 2020 - 01:00 PM, said:
- Review / Update Game Modes
- Solution for Faction Play
- Solution for Solaris
- General Match Maker Improvements
- 8v8 vs 12v12
- Solo / Group Queues
- Examine Match Scoring (AMS)
- Remove 2 Minute Requirement for Reconnection
- Private Lobby Updates (More Options, Maps, Host Assignment, More Spectators)
- Dailies/Weeklies
Please repost and continue the discussion there. It's not useful to scatter the discussion.
Thank you!
Thread locked.
2 user(s) are reading this topic
0 members, 2 guests, 0 anonymous users