25 replies to this topic

[Knight Captain Morgan]

Thanks to mixed group vs solo que you win or lose based on wether or not the "group" match maker assigns to your team sucks or not so good luck! At the beginning of the match, just look for any 2-4 group of players who share the same unit tag. If they suck then congrats! you lose! so be sure to credit their suckiness with your loss in general chat...

Edited by Knight Captain Morgan, 04 January 2021 - 04:55 PM.

[VonBruinwald]

Yep.

[Heavy Money]

I've got some bad news for ya, but you've always won or lost based on who you get matched with, regardless of if they're a group or not.

[JediPanther]

Should I even show a Founder in t5 with the cptl-1f screen shot my brother on his act took? In t5-4 it very much depends on the 2-4man sucking less than the enemy from watching the matches he's had.

[PocketYoda]

I win either way as i'm playing Virtual Battletech no matter which team loses i get cbills.

Edited by Samial, 05 January 2021 - 09:13 PM.

[Elizander]

It's not just the group, it's the players in your team in general and it's always been that way. Losing when you do 800 damage done happened before the merge and winning with under 100 damage done also happened.

Let's just accept that people are gonna be salty and give reasons why they are losing regardless of what system PGI or any company puts in to place. Is this current system perfect? Hell no, but at least admit that no system is gonna be perfect for you unless you get 100% win rate.

People log in and log off all the time. The players that best suit your playstyle aren't gonna be robots that are on 24/7 to give you the perfect match each time.

[RockmachinE]

Elizander, on 05 January 2021 - 02:34 AM, said:

Let's just accept that people are gonna be salty and give reasons why they are losing regardless of what system PGI or any company puts in to place.

This. I've been playing this game for years and public play has more or less always been the same.

[General Solo]

Knight Captain Morgan, on 04 January 2021 - 04:51 PM, said:

Thanks to mixed group vs solo que you win or lose based on wether or not the "group" match maker assigns to your team sucks or not so good luck! At the beginning of the match, just look for any 2-4 group of players who share the same unit tag. If they suck then congrats! you lose! so be sure to credit their suckiness with your loss in general chat...

Its pretty much the same as old solo queue expect we used the term team mates rather than group

As always a working match maker is the answer

Doh premads don't agree, its not match maker say guys in a group, as I was told in game last night, both red and blue groups

What a load
Maybe they never hear of dead group que

[GARION26]

Do you really think at most 1/3rd of the team has more match impact then then 2/3rd of the team?

If you've got 2-4 bad or really good players on yoru team they will shift your match whether they are or aren't in a group particularly if they aren't matched on the other side in terms of skill.

Matchmaker treats all groups the same currently - that's a problem certainly. But it's not the main reason most matches are won or lost.

[Anomalocaris]

Just loving all these people saying "it isn't groups"

Let's be clear. PGI doesn't do matchmakers well. But what groups do is make the matchmaker even _worse_

Please imprint that on your brain. Fixing groups will not fix all PGI's matchmaking woes, but it will make things _better_.

And yes, groups do have a disproportionate impact on the W/L of the team. This effect gets worse the more their skill level deviates from the mean skill in the match. In the old solo matchmaker, you'd have high skill and low skill players _more_ equally distributed between teams. Yes, you'd get imbalances. But they would _rarely_ equate to the imbalance we see when 4 potatoes in lights get dropped against 4 elite players in RFL-IICs.

This is really not even a debatable point, as PGI's own data showed that match quality declined when the queues were merged.

So, yes, please, keep pestering PGI about improving matchmaking (they need to start by fixing PSR). But if you want better matches now, separating queues is the quick and easy way to do it.

So yes, dumping groups from solo queue would be an improvement. And its a logical step since that was the last major change made to matchmaking/queues.

[PocketYoda]

I agree somewhat now you can predict almost when a match is going to lose just by looking at the groups tags.. I do not know how to fix it though as the match maker is really bad now. Honestly i've never seen a worse matchmaker than MWO.

One thing i do know is all tiers should not be fighting each other, tier 1 and 2 should not be farming Cadets and tier 5.

Edited by Samial, 05 January 2021 - 04:50 PM.

[Zordicron]

The only time I mind a group on my team is when they decide they are going to play as a group, and completely ignore the rest of their team. you know, the other 2/3's of it.
Example, polar highlands, 4 man in Orion IIc's runs off down the sideline.... to go cap.... the enemy cap point on the far side of the map.
Or when the group is "mad" that someone voted for assault or some other non skirmish/domination game mode and base rushes, by themselves.
Tonight I had a premade of 3 all in Stalkers decide to base rush in Crimson on an incursion game. I mean, define "rush" in a Stalker I guess. All that happened was the rest of the team reluctantly went to their aid when the enemy returned to their base in force, and the enemy basically just surrounded us and a slugfest happened briefly, but we lost 12-5. Half thr team got cored in the back because there was no place to go to regroup or use for cover. Brilliant.

So yeah, it's not really the skill level i think of the groups, it is if they decide to play the team based game as a team, or as their own little group. IMO, pugs have better "unspoken" team support skills in general, teams of 12 pugs may be chaotic, but in a herd of cats sort of way. The small groups that get dumped in let the chaos ball to their own devices and in doing so weaken both themselves and the pugs. In short, the pugs are looking at their mini map and radar and watching movements to decide their own, the group is not, and the signals get lost. Groups move one way against the instincts of the pugs on their team and at that point it turns into a turkey shoot for the enemy.

Honestly, if it was 8v8, pug teams could absorb a group a little easier IMO. Sounds counter intuitive, but when half the team is ignoreing the pugs, the pugs can adapt easier to the team. When there are 12, and only 1/3 are in their own world, it gets chaotic to adjust, if that makes sense.

[Elizander]

Zordicron, on 05 January 2021 - 08:08 PM, said:

The only time I mind a group on my team is when they decide they are going to play as a group, and completely ignore the rest of their team. you know, the other 2/3's of it.
Example, polar highlands, 4 man in Orion IIc's runs off down the sideline.... to go cap.... the enemy cap point on the far side of the map.

I've had this happen a few times. We had 4 same-tags and the enemy had none, but we still lost. The groups either brought all light mechs, went their own way or just didn't communicate with the team and died in isolation.

To be honest, I value getting a good drop caller on VOIP more than a 4-man with the same tags. Mech Commanders are OP.

Edited by Elizander, 06 January 2021 - 12:11 AM.

[Knight Captain Morgan]

Been watching streamers lately. They get matches in 10-15 seconds and there’s a 4 man premade every single time. It only takes 6 4-man groups to have an all group vs group match. 6 times 10-15 seconds is only 1-1.5 minutes. Let matchmaker b required to spend a minimum of 4 minutes trying to create group vs group matches before filling in any gaps with solos.

[VonBruinwald]

Knight Captain Morgan, on 06 January 2021 - 07:54 AM, said:

Been watching streamers lately. They get matches in 10-15 seconds and there’s a 4 man premade every single time. It only takes 6 4-man groups to have an all group vs group match. 6 times 10-15 seconds is only 1-1.5 minutes. Let matchmaker b required to spend a minimum of 4 minutes trying to create group vs group matches before filling in any gaps with solos.

Schrödinger's group players. They only play as a group when there's no group queue.

[Knight Captain Morgan]

VonBruinwald, on 06 January 2021 - 08:00 AM, said:

Schrödinger's group players. They only play as a group when there's no group queue.

The truth that none of the premades nor PGI are willing to admit to.

[Anomalocaris]

Knight Captain Morgan, on 06 January 2021 - 07:54 AM, said:

Been watching streamers lately. They get matches in 10-15 seconds and there’s a 4 man premade every single time. It only takes 6 4-man groups to have an all group vs group match. 6 times 10-15 seconds is only 1-1.5 minutes. Let matchmaker b required to spend a minimum of 4 minutes trying to create group vs group matches before filling in any gaps with solos.

I don't think it quite works that way. However, when we tried 8v8 group queue before the merge matches kicked off pretty quick during prime time. And that was with less population than we have now. 8v8 group queue only requires four 4-man groups which should be very easy to pull off.

I say reduce max groups in soup queue to 2-mans (introduce a friend to the game) and put 4-mans in 8v8.

Of course, if these groups in soup are afraid of playing against other groups, that won't work. Only one way to find out.

[SCRAPMETAL99]

i personally think match make ris rigged to keep a certian amount of w/l ratios i mean really 25l oss streak then all of a sudden 20 odd winning streak something is very fishy if you ask me

[LordNothing]

SCRAPMETAL99, on 06 January 2021 - 09:38 AM, said:

i personally think match make ris rigged to keep a certian amount of w/l ratios i mean really 25l oss streak then all of a sudden 20 odd winning streak something is very fishy if you ask me

there's definitely some control theory in play. but the game can only make predictions as to who is going to win. human behavior is one hell of an rngesus. the game then would use those predictions to figure out which team will win. then swap similar skilled players across teams so the players with wlr<1 are on the winning side. the predictions would only really need to be more than 50% accurate for that to work. the very good and very bad players would probibly see less of this, as they would be working against the control system rather than with it. its why this graph peaks instead of curves.

Edited by LordNothing, 06 January 2021 - 05:48 PM.

[Gnash Moore]

The matchmaker definitely does some pretty odd things. I just had a match where I was on the losing end of a game where the opposing team had five assaults, five heavies, one medium and one light and we had one assault, five heavies, five mediums and one light. 870 tons vs 695 tons.

Naturally they just walked right on through us and we didn't even look like stopping them for five seconds. I screwed up and got shot by most of their team but we were already five down at that point and it didn't really matter. Matches that are completely lopsided happen when one team plays well and one plays poorly, that's fine. Matches that are completely lopsided before a shot's even fired are stupid.