

Whatever Happened To Weight Matching?
#1
Posted 07 April 2013 - 02:49 PM
#2
Posted 07 April 2013 - 03:02 PM
#3
Posted 07 April 2013 - 03:10 PM
codynyc, on 07 April 2013 - 03:02 PM, said:
ELO was supposed to stop us getting stuck with incompetent players...that didn't work either. Put simply, the matchmaking sucks all round.
#4
Posted 07 April 2013 - 03:12 PM
Remember when there were fewer "unable to find a match" when you were matched by weight, no matter how ricockulous your comp was?
#5
Posted 07 April 2013 - 03:13 PM
It seems really simple.
Edited by Victor Morson, 07 April 2013 - 03:13 PM.
#6
Posted 07 April 2013 - 03:23 PM
Victor Morson, on 07 April 2013 - 03:13 PM, said:
It seems really simple.
Probably part my fault. There were some of us QQing that there was no ranking system that matched you with, and against similarly skilled/competent players...little did we know the monster this would create. But, it is BETA, and we are here to test these things. If it were full release, the game would be boned on the MM alone.
#7
Posted 07 April 2013 - 03:23 PM
Make weight class matching the top priority.
Elo matching would become the second priority.
This system should be able to correct any Elo mismatches over time as the losing teams over time will have a drop in Elo ratings and face easier teams and the winning teams will have a rating increase to face more difficult teams.
The big difference with this system is that players will see equal weight class matching. This is very important as new players may have no idea Elo is used in matchmaking. All they see are the mismatched weight classes and they give up on this game because they see a very unfair matchmaker that is allowing 1 team to run 4 assaults while their team gets a single assault for example.
Even experienced players can see serious balance problems. If the match turns into a brawl the team with more firepower will generally win if equal skilled pilots are on both teams. If the map is larger the team with more lights often wins by cap.
#8
Posted 07 April 2013 - 03:40 PM
the unmatched tonnage is more realistic.
#10
Posted 07 April 2013 - 03:51 PM
#11
Posted 07 April 2013 - 03:53 PM
Tennex, on 07 April 2013 - 03:40 PM, said:
the unmatched tonnage is more realistic.
I don't mind the lack of weight class matching for 8-man drops as a group has control over the entire team.
The problem with mismatched weights in small groups or solo drops is the way players have no control over the rest of the team. New players see mismatched weight classes and give up on this game after their team with a single assault gets stomped by a team with 4 assaults in under 3 minutes.
Vrekgar, on 07 April 2013 - 03:51 PM, said:
It seems like it's doing this after even 15 to 20 seconds though.
#12
Posted 07 April 2013 - 04:00 PM
I totally agree that tonnage should be closely matched too.
As far as Elo goes generally though, I've found it works for me because I'm not good enough to get balanced by a lame team. My win/loss ratio has levelled out around 50/50 since Elo was introduced, and my allies and opponents all seem fairly competent.
#13
Posted 07 April 2013 - 04:03 PM
Hedonism Robot, on 07 April 2013 - 03:49 PM, said:
Actually the closed beta matchmaking system was pretty good in 2 areas in the later part of closed beta:
Any size group allowed.
Exact weight class matching.
The only changes that were needed:
Match a group on 1 team against a group on the other team.
Add Elo to also match player skills.
This would have resulted in a matchmaker similar in function to the closed beta system that was able to do the following:
Exact weight class matching.
Any size group allowed, matched against equal size groups when possible.
1 group per team allowed.
Teams balanced as close as possible based on Elo rating.
#14
Posted 07 April 2013 - 04:30 PM
I would definitely support having weight ton matching enabled, especially on matched groups where another could benefit from my frugality.
It only helps support diversity.
#15
Posted 07 April 2013 - 04:50 PM
#16
Posted 07 April 2013 - 05:49 PM
I think the logical thing to do is to have pre-match lobbies just like LoL and see what kind of meta evolves.
#17
Posted 07 April 2013 - 05:58 PM
Noobzorz, on 07 April 2013 - 05:49 PM, said:
I think the logical thing to do is to have pre-match lobbies just like LoL and see what kind of meta evolves.
This would be the best option with role warfare. The idea behind role warfare is each mech contributes in its own unique aspects to winning the game. 8 assaults loses in 8 mans (no one plays this unless they are trolling and want to lose).
Please remember folks that balance (decisions on what to play) is not easy, but will make for a lasting game.
#18
Posted 07 April 2013 - 06:32 PM
Vrekgar, on 07 April 2013 - 03:51 PM, said:
I did not know this... This explains why I keep getting dropped with noob team mates, despite a 3/1 win/loss ratio, and why it always seems to be my atlas vs 3 or 4...PGI really need to pull their finger out of their arse if they want to go live in summer. This game won't last 5 minutes live in its current state.
*so I went back and worked it out, and my ratio is nowhere near that good now. Suppose teaming you with gimps, and pitting you against similarly skilled people to yourself (guaranteeing a loss), is ELOs way of 'balancing' your ratio.
Edited by Mokey Mot, 07 April 2013 - 06:40 PM.
#19
Posted 07 April 2013 - 07:22 PM
Zylo, on 07 April 2013 - 03:53 PM, said:
The devs never said WHO has to wait. My running personal theory is that if any group waits more than 2min it just grabs people out of the Que immediately even if they have not hit that time. Its the only way to explain getting a near instant match that is very one sided with a large 200+ton disparity in weight.
#20
Posted 07 April 2013 - 07:25 PM
Vrekgar, on 07 April 2013 - 07:22 PM, said:
While this may be true it seems odd that I have never seen matchmaker take 2 minutes to find a match except in 8-man drops. This might be related to the fact that I always drop in a group of 2, 3 4 or 8 though.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users