107 replies to this topic

[blood4blood]

I went searching for official info...

Found this old thread with math on computing ELO and how it is adjusted match-to-match:
http://mwomercs.com/...79-matchmaking/

Some stuff about how the matchmaker uses ELO:
http://mwomercs.com/...-making-update/

More recent adjustment to threshold info:
http://mwomercs.com/...old-adjustment/

There were other threads about match making as well, but mostly concerning weight classes/tonnage and premade vs PUG.

After reading the little info available that's official, the ELO that counts in PUG games appears to be a "closest-to-target" "team average" rating matched against a threshold value (most recently announced at 1400). If you're very high on the ELO scale (which runs 0 to 2800) it would naturally appear to team you with more newbies to reach that average team score of 1400. But the whole point was to match players of roughly equal skill level to get more competitive matches...I guess I just don't get how team average ELO scores are supposed to do that. They seem designed to create the situations OP is experiencing instead - wildly unbalanced skill levels within a PUG team.

[Taffer]

Weird I've noticed this too. A lot more people saying this is their first match and stuff.

I thought somehow, someway, in the last couple of weeks mwo got a lot of new players.

[wanderer]

Basically, you've got the gist of it.

The better you are, the higher your odds of being "balanced" by newbies.

Of course, the worse and larger the numbers of said easy targets become, the easier it becomes for even the derpiest of PUGs to focus fire (naturally, a smaller number of live targets means more people are likely to fire at them) and hence the steamroll gets it's smash on.

It really doesn't take much. 2-3 easy targets means a significant early advantage, those people you see piloting an assault/heavy that died with under 100 damage dealt and making your life that much shorter.

[Yokaiko]

Hence why I said using individual ELO in a TEAM game was pretty stupid.
Also why 3/3/3/3 was worse
...as well as ELO by class
and matches being determined by TEAM elo average.


If you had the population of LoL or WoT that is one thing, MWO doesn't, nor will it.

[fyurian]

I know your feels bro
tho only mostly in heavy's because that elo is considerably higher, in lights im a complete derp

I rather see a ranking system that goes trought atleast 5levels kinda like L.O.L.
and just keeps u in the same skill group, so u know u can depend on them when sh*t hits the fan.
from 1 being new and a complete "insert red foreman quote"
to 5 pure godlike player

[Oderint dum Metuant]

mogs01gt, on 07 January 2015 - 12:27 PM, said:

Im sorry but people cannot be this bad. Hell I've even posted about them being bots because they behave like botss and not like humans.

No human would simply walk into 1v12.

see sig

[Fate 6]

mogs01gt, on 07 January 2015 - 12:27 PM, said:

Im sorry but people cannot be this bad. Hell I've even posted about them being bots because they behave like botss and not like humans.

No human would simply walk into 1v12.

New players don't really get how this game works. It's a very different kind of game from most shooters, and for a lot of people that is a difficult concept to grasp.

[Yokaiko]

Fate 6, on 07 January 2015 - 03:38 PM, said:

New players don't really get how this game works. It's a very different kind of game from most shooters, and for a lot of people that is a difficult concept to grasp.

There are still founders crying that LRMs are OP........

[PurpleNinja]

ELO is OP.

[Mudhutwarrior]

I am still not seeing many newbies in solo but some of my friends say they have had a lot. Maybe thats ELO or MM who knows?

Seeing a whole ton of flounders in solo now though. I guess they all came back tried CW and gave up.

[Gyrok]

FupDup, on 07 January 2015 - 12:21 PM, said:

Ok, I LOL'ed at the upside down stalker at the bottom of the pack...hahaha

Mudhutwarrior, on 07 January 2015 - 04:05 PM, said:

I am still not seeing many newbies in solo but some of my friends say they have had a lot. Maybe thats ELO or MM who knows?

Seeing a whole ton of flounders in solo now though. I guess they all came back tried CW and gave up.

Or...you know...they pug while waiting for their group to form up...because nobody drops while waiting...

Edited by Gyrok, 07 January 2015 - 04:22 PM.

[Impossible Wasabi]

Yokaiko, on 07 January 2015 - 02:40 PM, said:

Hence why I said using individual ELO in a TEAM game was pretty stupid.
Also why 3/3/3/3 was worse
...as well as ELO by class
and matches being determined by TEAM elo average.


If you had the population of LoL or WoT that is one thing, MWO doesn't, nor will it.

The Space Pope agrees, ELO works best in Chess due to the nature and rules of said game, it also works ok (but not great) in WoT or LoL because you have huge populations.

In MWO it just blows.

fyurian, on 07 January 2015 - 02:53 PM, said:

I know your feels bro
tho only mostly in heavy's because that elo is considerably higher, in lights im a complete derp

I rather see a ranking system that goes trought atleast 5levels kinda like L.O.L.
and just keeps u in the same skill group, so u know u can depend on them when sh*t hits the fan.
from 1 being new and a complete "insert red foreman quote"
to 5 pure godlike player

The Space Pope also agrees with this, he'd rather know for sure that people he was playing with were at least somewhat capable and face similarly skilled players on the other team than get stuck on a bad team in order to balance ELO.

It would also be more fun as well, because it would mean that as you went up in level you were more likely to find good opponents.

Edited by The True Space Pope, 07 January 2015 - 04:29 PM.

[Gyrok]

blood4blood, on 07 January 2015 - 02:16 PM, said:

I went searching for official info...

Found this old thread with math on computing ELO and how it is adjusted match-to-match:
http://mwomercs.com/...79-matchmaking/

Some stuff about how the matchmaker uses ELO:
http://mwomercs.com/...-making-update/

More recent adjustment to threshold info:
http://mwomercs.com/...old-adjustment/

There were other threads about match making as well, but mostly concerning weight classes/tonnage and premade vs PUG.

After reading the little info available that's official, the ELO that counts in PUG games appears to be a "closest-to-target" "team average" rating matched against a threshold value (most recently announced at 1400). If you're very high on the ELO scale (which runs 0 to 2800) it would naturally appear to team you with more newbies to reach that average team score of 1400. But the whole point was to match players of roughly equal skill level to get more competitive matches...I guess I just don't get how team average ELO scores are supposed to do that. They seem designed to create the situations OP is experiencing instead - wildly unbalanced skill levels within a PUG team.

Exactly this...

These 2 teams are "equal" in the eyes of the MM:

Team 1:
1000
1000
1200
1200
1400
1400
1600
1600
1800
1800

Average ELO: 1400

2800
2800
2800
940
940
940
930
930
930
930
930
930

Average ELO: 1400

Which team will get rolled just by looking?

[InspectorG]

With great KDR comes great responsibility frustration.

[nehebkau]

ELO carry-harder hell.... your ELO gets to a point where you are expected to be able to telepathically guide the window-lickers to become uber-gamers.

[F4T 4L]

Sethliopod, on 07 January 2015 - 01:01 PM, said:

It means it's time for you to thank all those who had been carrying you for months. Your turn.

QFT.

[mark v92]

been there. you spawn and look next to you to see 4 guys running in 3pv with armlock on.
you will be doing more to the enemy than those 4 guys combined.

they have their uses tho:

distraction for your flanking
bullet sponge
walking cover

time to start running mechs that can carry hard

either flank and surprise kill as many high elo enemies as you can before they start chewing up you teamates or soften up all enemies before they reach you team so that your teamates might have a chance at killing them

[CMDR Sunset Shimmer]

CliffhangerJaeger, on 07 January 2015 - 11:45 AM, said:

I thought that we are grouped with similarly skilled players.

Clearly you've not been paying attnetion.

The ELO system clearly puts you with less skilled players on your team so that you can carry them... always has, always will.

IDEALY, ELO would be placing us with similarlly skilled players on teams... but apparently, that's not how it works.

[Ano]

In case no-one's said it: I've noticed this too, and while I'm certainly a more effective player than I was three months ago, it's not like my inbox is flooded with messages from people begging me to join their unit.

My suspicion: Given that a lot of "good" players are playing CW, the hump of the skill/population bell curve has been pushed to the left, moving players like me (who were most likely still working to get to "average") from the left slope of the curve to the right slope of the curve. Meaning there are a lot of comparatively easy targets for me to shoot at.

That's not to say that CliffhangerJaeger is in the same position -- he/she might well have significantly improved, or found the mech/playstyle that suits -- but in my case at least I'm a little suspicious when within a couple of weeks I go from "probably not a total waste of tonnage" to 4 kills/match not being a reason to go pour myself a celebratory drink.

[Clydewinder]

This usually only happens when I run mediums. I have screenshot after screenshot of derp teams with 40 damage Direwolves where I'm getting #1 in damage and kills in a Vindicator or Blackjack.

Something is definitely weird.

Seems to depend on the time of day as well. Morning/afternoon US time is maximum derp fest.