33 replies to this topic

[Yuri Kovalesky]

Not sure if is only with me and some friends but after this new patch i have noticed the ELO is not working as it should be.

Every single match alone or with friends i see my team entirelly with random pugs ( Not even part of a unit ) and the enemy side ALWAYS with premade units and people with groups.

This is happening after this new patch and i am starting to believe it is not a coincidence.

I have made one print and it is not uncommon.

Im not here to argue about the units i faced but to this Matchmaking mechanics.

Theres a difference between find one and maybe some premade teams in the enemy side while you face pugs in yours but 8/10? Every day?

What is happening?

[Wintersdark]

Ok, there are no solo pugs here. It's all teams vs. teams. The number of big teams vs. little teams has absolutely nothing to do with Elo either.

And group queue matches tend to have a larger Elo spread than solo queue matches, simply because of the nature of the beast. That's got nothing to do with the matchmaker either.

[Yuri Kovalesky]

Wintersdark, on 09 May 2015 - 01:51 PM, said:

Ok, there are no solo pugs here. It's all teams vs. teams. The number of big teams vs. little teams has absolutely nothing to do with Elo either.

And group queue matches tend to have a larger Elo spread than solo queue matches, simply because of the nature of the beast. That's got nothing to do with the matchmaker either.

Still, have my point.

Even if you are right, it is unbalanced because the LFG dont count as ELO but the Matchmaking throws Premade in the other side.

Its all about one thing not being used with another and it wasnt like that before.

[Wintersdark]

Yuri Kovalesky, on 09 May 2015 - 02:03 PM, said:

Still, have my point.

Even if you are right, it is unbalanced because the LFG dont count as ELO but the Matchmaking throws Premade in the other side.

Its all about one thing not being used with another and it wasnt like that before.

Everyone in both teams is in a premade. All of you are. You can't expect differentiation between LFG groups and other groups; that level of queue fracturing would make both queues even worse than the current group queue.

LFG *is* counted towards Elo calculations just like any other group. No matter how you made the group, as far as the MM is concerned you're just a group like any other.

So, your group (be it LFG sourced, or friends list, or unit sourced) is simply a collection of players with an Elo score equal to the average of all the component players.... just like it has always been.

Edited by Wintersdark, 09 May 2015 - 02:10 PM.

[Ironwithin]

Yuri Kovalesky, on 09 May 2015 - 01:45 PM, said:

Not sure if is only with me and some friends but after this new patch i have noticed the ELO is not working as it should be.

Every single match alone or with friends i see my team entirelly with random pugs ( Not even part of a unit ) and the enemy side ALWAYS with premade units and people with groups.

This is happening after this new patch and i am starting to believe it is not a coincidence.

I have made one print and it is not uncommon.

Im not here to argue about the units i faced but to this Matchmaking mechanics.

Theres a difference between find one and maybe some premade teams in the enemy side while you face pugs in yours but 8/10? Every day?

What is happening?

I think you misunderstand how the matchmaker works.

If you drop with a group of friends, no matter how large (as long as you are at least 2 people) you will always drop in the group-queue, where both sides are made up entirely of groups.
Supposedly the matchmaker tries to find somewhat equally sized groups to put together/against each other but the longer the search goes the more the restrictions are eased. The same goes for Elo-scores.

So yeah ...worst case you end up with something like a 10man+2man group of really skilled and coordinated players facing off against a team of six 2-man groups that are not coordinated at all.


If you drop alone you will always drop in solo-queue where there are no groups at all on either side.

Also: Elo is a name, not an acronym.

Edited by Ironwithin, 09 May 2015 - 02:12 PM.

[Void Angel]

[Hydrocarbon]

One "tell" about that screenie is the 12DG guys, who are mostly German time-zone. My guess is you drop when the server population is very low. Not much can be done when only a few dozen mixed-size groups are roaming the server.

BTW I've seen random pugs in CW have better scores than most of the unit blobs, but they are certainly the exception to the rule. Either scalp other unit members to switch to your unit or deal with it.

[Modo44]

The matchmaker builds a team of players with similar skill. If they happen to be very high or very low Elo, there may not be enough people in the queue to build a matching team. After 4 minutes, it throws in the towel, and drops the way over/understacked team against whatever is there. The issue gets worse when there are fewer people online. This is working as designed. It will not change unless the matchmaker is rewritten to 1) build the entire match combined in one go, and 2) not allow such large groups.

[ShinVector]

Yesterday was a bad night for me... Very long search times and multiple losses. (Yeah some by fault for not being able to carry.)
Anyway.. Kinda weird normally I win more than 50% of my matches.. Oh well..

[B0oN]

Pardon me, I thought I misread ...
ELO was working at some time and not just a fabricated mystery ?

Blasphemy !

/jk

[Rushin Roulette]

Rad Hanzo, on 13 May 2015 - 01:24 AM, said:

Pardon me, I thought I misread ...
ELO was working at some time and not just a fabricated mystery ?

Blasphemy !

/jk

Yes. Elo deffinitely worked at some time. However, Elo is deffinitley not working since 05. November 1992.

http://en.wikipedia.org/wiki/Arpad_Elo

[Ironwithin]

Rushin Roulette, on 13 May 2015 - 01:45 AM, said:

Yes. Elo deffinitely worked at some time. However, Elo is deffinitley not working since 05. November 1992.

http://en.wikipedia.org/wiki/Arpad_Elo

But he was talking about ELO and according to wikipedia (http://en.wikipedia....Light_Orchestra) they somwhat recently reformed and might even go on tour.

[Rushin Roulette]

Ironwithin, on 13 May 2015 - 01:50 AM, said:

But he was talking about ELO and according to wikipedia (http://en.wikipedia....Light_Orchestra) they somwhat recently reformed and might even go on tour.

Either that or Rad is a fan of certain Canine species

http://en.wikipedia.org/wiki/Elo_(dog)

[Michal R]

Elo is only use when you play solo. When you play premade vs premade there is only 3/3/3/3 rule.
Simple?

[PhoenixFire55]

Elo is working Fine^TM. You are on an Island^TM.

On a serious note ... we've had this discussion hundreds of times already. Elo doesn't work, it never did and it never will be as long as PGI does things "Our way^TM".

Solo queue at times have at least some resemblance of balance, group queue has none whatsoever. The quality of matches is same as produced by random MM before Elo idiocy started, but wait times are 3-5 mins instead of 3-5 secs. CW matches ... don't even have any MM in the first place.

The screenshot you posted is yet another proof of what I've been saying dozens of times. One team has a highly skilled 5-man and a highly skilled 7-man. Other team has (at best) a not-so-skilled 7-man, a 3-man and a pair. Instead of putting Donegals and Spikes on different teams for some degree of balance (which could have been easily done by swapping all 12DG guys for INEX+WBH), our precious Elo-MM that is Working as Intended^TM puts them on the same team to ensure a victory for that team. I bet the next game both these groups are placed on a "bound-to-lose" team in the same fashion. Thus over the extended period of time everyones W:L ratio will become 1:1 which will lead all clueless people to believe in the balanced MM, while in reality none of the matches played have been balanced (at least not intentionally) by Elo-MM.

All we keep hearing are excuses why this and that doesn't work and why such and such suggested improvements can't be made, but as long as the moneyflow from yet another package or sale keeps going they won't do crap.

[Tarogato]

Michal R, on 13 May 2015 - 02:26 AM, said:

Elo is only use when you play solo. When you play premade vs premade there is only 3/3/3/3 rule.
Simple?

Absolutely false. Elo rating is shared between group and solo queue and it is used by matchmaker in both queues to group players of similar skill level as best it can under the constraints of the players available to pool from.

@OP et al:
That last bit is critical when talking about group queue matches. There might be one single uber-high Elo premade group playing, but no uber-high Elo premade group to go up against them. At some point matchmaker is going to have to decide "screw it, I need to make a game happen eventually, let's do this."

Take a look at your screenshot. There is a 5-man premade of [12DG], and presumably a 7-man premade of Jade Falconers (counting the FRR guy). On your team there was your group of 3, what looks like a group of 3 Steiners, and presumably a group of 6 lonewolfs/notaggers.

5+7 = 12
3+3+6 = 12

You can't have 7+6,
You cant have 5+6
You can't have 7+3+3

... matchmaker has to play Tetris with group sizes, AND try to uphold 3/3/3/3 when possible, AND try to keep the Elo disparity low. It literally does not have the power to make players appear out of thin air in the correct group sizes just to play against your group at the appropriate skill level. Concessions have to be made.

Edited by Tarogato, 13 May 2015 - 04:55 AM.

[Kyynele]

PhoenixFire55, on 13 May 2015 - 03:35 AM, said:

The screenshot you posted is yet another proof of what I've been saying dozens of times. One team has a highly skilled 5-man and a highly skilled 7-man. Other team has (at best) a not-so-skilled 7-man, a 3-man and a pair. Instead of putting Donegals and Spikes on different teams for some degree of balance (which could have been easily done by swapping all 12DG guys for INEX+WBH), our precious Elo-MM that is Working as Intended^TM puts them on the same team to ensure a victory for that team. I bet the next game both these groups are placed on a "bound-to-lose" team in the same fashion. Thus over the extended period of time everyones W:L ratio will become 1:1 which will lead all clueless people to believe in the balanced MM, while in reality none of the matches played have been balanced (at least not intentionally) by Elo-MM.

12DG is a huge unit. And to me, these guys don't seem to be their #1 competitive guys (do correct me if I'm mistaken, and no offense meant to anyone, maybe I've just forgotten some names.) This match would've been a stomp any way you would've mixed these groups into teams. It might've been less of a dramatic screenshot, but all you can really say from this screenshot is that the Spikes got matched against and with less skilled people. After enough waiting, MM just has to put these groups in some matches.

Also, lol @ everyone's W:L ratio becoming 1:1.

[Idealsuspect]

Yuri Kovalesky, on 09 May 2015 - 01:45 PM, said:

What is happening?

Nothing you just face a full 12' ...

Well you don't know that MM is broken for little groups? dont play in little group if you are less than 6 or 7 ...

[PhoenixFire55]

Kyynele, on 13 May 2015 - 05:20 AM, said:

12DG is a huge unit. And to me, these guys don't seem to be their #1 competitive guys (do correct me if I'm mistaken, and no offense meant to anyone, maybe I've just forgotten some names.) This match would've been a stomp any way you would've mixed these groups into teams. It might've been less of a dramatic screenshot, but all you can really say from this screenshot is that the Spikes got matched against and with less skilled people. After enough waiting, MM just has to put these groups in some matches.

Also, lol @ everyone's W:L ratio becoming 1:1.

What is your point? That MM is working just fine? This 12:0 roflmaostomp disagrees. This is a match from EU prime time, there are plenty people online.

As for 1:1 ratio you know well enough yourself that for most people it is just about 1:1. And it was achieved by placing them in such particular situation more or less every match, on either a bound-to-win or bound-to-lose team with equal probability.

[SilentScreamer]

Since the matchmaker difficulties and mystery of Elo have already been discussed I am going to add two more ideas as to why the result was 12-0.

1) What plan of attack and tactics did each team use?
2) How well did each team execute that plan?

I have the good fortune of playing with two seperate pilots that have placed in the top 10 players during events. Even with those pilots, we take 12-0 losses sometimes. The result is usually because half our team spreads out, allowing the enemy to pick us off one at a time. That loss is lack of leadership and a plan and not the fault of the matchmaker or Elo.

It does not matter if you PUG, drop as a 2-10 man, fill up to 12 using LFG or join a Teamspeak server to find teamates. All carry the risk of getting low quality teamates who 1) have less than optimal mech builds, 2) do not have good skills, 3) choose not work with the team.

I think #3 results in more 12-0 matches than the other problems. The only way to prevent it is to form a 12 man and KICK players who do not work with the team. Yes, it could get you a nasty reputation, but I think it will get you better results too. Just a side note, 8-10 players is probably better than 12 to avoid a true 12 man premade vs your improvised premade. During the LFG event I was on several improvised teams that beat a 12 man clan/unit group. We had to come up with a plan AND execute it well.