150 replies to this topic

[KingCobra]

To the OP lets just clear this up right from the start the MM/ELO is a joke it does not tell you which person or team is the best all it does it try to equal the playing field for all players.The current system should only be used to train PUGS not for a true competitive type game play that many in past PC MECHWARRIOR games were used to.Once you understand this concept you know unless true 1V1-12V12 game types in private matches are made it is really not a true competition or are a reflection of a solo players skills or a teams skills in actual gameplay.

PGI promises true competition in private matches for UI2.0? but we will see if it happens without a lobby system map choices and many more multiple game types MWO is a terribly bland and uncompetitive game compared to past PC MechWarrior titles and there concept of CW will not change the fact a solo player or team cannot seethe true skill potentials or true competition with the current MM/ELO system.

[Odins Fist]

Mudhutwarrior, on 15 November 2013 - 04:45 AM, said:

I started a new account over the weekend and saw no real differnce in gameplay between my 8000 match pug account and the new one. Saw very few new players and met just as many premades as I do in this account.

^THIS^ ^ ^
I have heard this from many of my friends that tested ELO as it relates to BRAND NEW players (using alt accounts).

They all said similar things, "within 3 matches i'm running against good people I normally run against"

NOW I can't confirm or deny this, and I will test the theory, but I "think" ELO is only based on win/loss ratio and nothing more.

[Roadbeer]

You know, when I woke up this morning, I was thinking,

"God I hope there is a new Elo/MM thread on the forums, that topic just hasn't been covered enough"

Thanks OP.

[SmithMPBT]

Kunae, on 15 November 2013 - 06:30 AM, said:

It's not elitism, it's just observed facts.

If you want to talk jack-assery, then just look at the guys who are doing that "Lord ..." followed by "ggclose" at the start of matches. Those guys tend to be only useful in groups, and are not representative, at all, of the majority of folks who are in the upper Elo brackets.

The OP is merely expressing his frustration at the stupidity of PGI making the MM use low Elo folks, many still in their first 25 games, to "average out" people with high Elo and over 4000 matches. It's just dumb.

I was gonna say the OP sounds like a pompous stat loving WoT playing ***, but you're absolutely right Kunae. Balancing a match using extreme sides of the ELO scale is insane.

[Kunae]

SmithMPBT, on 15 November 2013 - 08:50 AM, said:

I was gonna say the OP sounds like a pompous stat loving WoT playing ***, but you're absolutely right Kunae. Balancing a match using extreme sides of the ELO scale is insane.

I honestly think Matt Craig was only telling us part of the story, too, when he used his example of using "1 high elo player and 1 low elo player to balance out against 2 average elo players." It's becoming my thinking, and I could be totally off on this, that it will take one very high elo player and use 2-4 low elo players to balance it with. It's also possible I am smoking bath-salts and could be getting a hankering for chewing some random person's face off, too...

[KingCobra]

RoadBeer said==You know, when I woke up this morning, I was thinking,

"God I hope there is a new Elo/MM thread on the forums, that topic just hasn't been covered enough"

Thanks OP. Roadbeer all you do I patt your forum stats and never have anything worth saying or contribute on any topics so just go away and drink your life away.But h likes to derail topics be a knight knowing PGI/DEVS will never read 99.9% of any topics and posts. Most PGI knights are===>

[Grits N Gravy]

All they have to do is give the entire 3 and 4 person group the highest Elo score of their party for the purpose of matchmaking. It would likely boost some groups outside of the of the acceptable variance of average solo players. Though You may end up in the queue a bit longer if your group though.

Lets say you're in a 4 man with a wide spectrum of skilled players, remembering that about 1250 is the median elo score and 1500 is pretty elite. Your team is composed of a 1250, 1300, 1345, 1500 thus you average out to 1348. According to Elo a 50 difference in scores equates to about a 7% advantage to win. So we can assume that the 1348 teams has a variance of at least +-50 Elo points. The average player with his 1250 and +-50 variance overlaps with 1348 team. Which can result in a team of all averages Vs a team with a 4 group and 8 average players.

This will happen quite frequently too because it's the most probable event to occur. The average player has to play up and the higher ranked players will have to play down more frequently do to the distribution of Elo scores. As time in queue increase the spread of matching gets larger. Because the scores are distributed logarithmically, over time the most likely place a a higher ranked player will find a match is from the pool of players with scores lower than his, because there are simply more people there. There are less people above him than below him to match against.

Treating this group as the highest Elo largely fixes this problem. As it will result in more matches for high ranked players. It ends up populating more of the higher range of Elo scores, thus you don't end up with every other game being a pug team VS 4 man and 8 pugs.

Edited by Grits N Gravy, 15 November 2013 - 09:05 AM.

[Clit Beastwood]

DeathofSelf, on 15 November 2013 - 06:58 AM, said:

Not going to claim to know my ELO but I will say that I know my ELO should be high enough to not be in groups with new players. I am quite often on teams with players who are very new, always a bad sign when someone asks something like "How do I change weapon groups?"

Well, I ask stuff like that in all chat all the time thinking maaaaaybe someone on the other team will get overconfident. It amuses me, and I know I'm not the only one that does it. It's especially fun to type it in all caps. Then you get a bunch of people typing "Press alt+f4!". This is especially funny in matches where the asker racks up several kills

KingCobra, on 15 November 2013 - 08:57 AM, said:

RoadBeer said==You know, when I woke up this morning, I was thinking,

"God I hope there is a new Elo/MM thread on the forums, that topic just hasn't been covered enough"

Thanks OP. Roadbeer all you do I patt your forum stats and never have anything worth saying or contribute on any topics so just go away and drink your life away.But h likes to derail topics be a knight knowing PGI/DEVS will never read 99.9% of any topics and posts. Most PGI knights are===>

omg that looks like my friends baby.

[Roadbeer]

KingCobra, on 15 November 2013 - 08:57 AM, said:

RoadBeer said==You know, when I woke up this morning, I was thinking,

"God I hope there is a new Elo/MM thread on the forums, that topic just hasn't been covered enough"

Thanks OP. Roadbeer all you do I patt your forum stats and never have anything worth saying or contribute on any topics so just go away and drink your life away.But h likes to derail topics be a knight knowing PGI/DEVS will never read 99.9% of any topics and posts. Most PGI knights are===>

Awww, did I crush some nonsense of yours in the past?
There are currently 3 threads on this topic on the front page of THIS sub-forum alone.

[DeathofSelf]

Oh good, here comes KingCobra with his stupid big *** images that contribute nothing to the conversation (not that anything you say does either)

[Sug]

Odins Fist, on 15 November 2013 - 08:46 AM, said:

I have heard this from many of my friends that tested ELO as it relates to BRAND NEW players (using alt accounts).

They all said similar things, "within 3 matches i'm running against good people I normally run against"

NOW I can't confirm or deny this, and I will test the theory, but I "think" ELO is only based on win/loss ratio and nothing more.

Elo, not ELO, is based on wins and losses only yes. So you may have 800 damage done and 5 kills but if your team lost, you lost. Your score goes down.

A brand new player is assigned an Elo score that is about the average score among players, 1300. Since the matchmaker attempts to average out the scores a pure pugger will end up with a W/L ratio of about 50/50. Since Elo doesn't change that much over time, because you have a 50/50 win ratio, most players never get too far away from the average.

Again this is for pure puggers. If you do premades of any kind you've skewed your score a bit. Unless you're awful.

Read up on the matchmaking magic here http://mwomercs.com/...79-matchmaking/

Edited by Sug, 15 November 2013 - 09:13 AM.

[Kunae]

Sug, on 15 November 2013 - 09:10 AM, said:

A brand new player is assigned an Elo score that is about the average score among players, 1300.

No, a brand new player, for their first 25 missions, starts with a score of 1100. On the 26th mission they get a bump of +200pts to whatever their modified score is, at that point.

[Odins Fist]

Sug, on 15 November 2013 - 09:10 AM, said:

Elo, not ELO,

ELO, not elo

Cause I like to POST it that way...

I know, I know, Arapad Elo

Edited by Odins Fist, 15 November 2013 - 09:15 AM.

[Kunae]

Odins Fist, on 15 November 2013 - 09:13 AM, said:

ELO, not elo

Elo not ELO.

[Sug]

Having trouble finding the posts but these are the graphs that show the distribution of our Elo scores at some point like a year ago. I believe the red graph shows when we were given Elo scores but no match making, the lighter red shows where we started and the darker is weeks later.

The blue graph shows our Elo scores after match making was implemented. The lighter area again is where we started.





Don't quote me on this, this was from like last year so my memory isn't 100%

Edited by Sug, 15 November 2013 - 09:23 AM.

[JudgeDeathCZ]

Imho if ELO/MM works fine there should be no 12-0/1/2 etc stomps...
If it is balanced how it should be then there will be only 1 to 4 mechs alive on winning side.
Problem is that our Elo is average Elo so there are games like today where I lost game 12-2 and I was only one with over 150 dmg(I did over 500)...yeah team full of noobs.

[Grits N Gravy]

Sug, on 15 November 2013 - 09:21 AM, said:

Having trouble finding the posts but these are the graphs that show the distribution of our Elo scores at some point like a year ago. I believe the red graph shows when we were given Elo scores but no match making, the lighter red shows where we started and the darker is weeks later.

The blue graph shows our Elo scores after match making was implemented. The lighter area again is where we started.


Don't quote me on this, this was from like last year so my memory isn't 100%

Lighter area is less than 50 games. The red graph is the distribution of scores during a bug which caused winners to be awarded to many Elo points against lower ranked opponents. The blue graph shows scores after fixing the bug and re-calculating scores.

source: http://mwomercs.com/...ted-april-19th/

Edited by Grits N Gravy, 15 November 2013 - 09:30 AM.

[Sug]

Grits N Gravy, on 15 November 2013 - 09:28 AM, said:

source: http://mwomercs.com/...ted-april-19th/

Ah ty


"This caused Elo values to increase by a lot when they should have increased by only a little, skewing the player Elo distribution so that it became top-heavy."

Oh noes! The people that win more games actually increase their score and play against other people with high scores! The horror! Our population of frittata players will never be able to deal with that. Better hold their hands and implement Elo communism so bads can win a game sometime instead of just becoming better players.

Edited by Sug, 15 November 2013 - 09:52 AM.

[Ryan Steel]

Kunae, on 15 November 2013 - 06:30 AM, said:

If you want to talk jack-assery, then just look at the guys who are doing that "Lord ..." followed by "ggclose" at the start of matches. Those guys tend to be only useful in groups

You can tell the population of this game thrives on not having lobbies for private matches. Right now, anyone who runs their mouth like Kunae does can't be instantly challenged to a 1v1 or, 2v2 or what have you to put their money where their mouth is. Why don't you watch MavRCK's stream: http://www.twitch.tv/kaosmavrck and watch me go 1v12 in my Misery over the course of a night? Or watch videos of Villz (http://www.youtube.com/user/VillzMWO) and Kaffeangst (http://www.twitch.tv/kaffeangst) solo pug? We're pretty much 6 -- one half dozen dozen of the other in terms of skill level so I'm not too sure what you mean by "we're only useful in groups." Ok bud: http://mwomercs.com/...nt-performance/
Not a single one of us spent more than 3 hours playing that tourney.

Right now we play with a diceroll Elo system that handicaps us and matches us against a proportionately greater number of good players time and time again:
[Image Redacted]

It's just gg1ose and I vs Profiteer and his lance and a full lance of SJR and a Liao lance. We do have Heimdelight who is a good player and that's it. Common. This is all we face every game we play. To top it off I can't find matches in my assaults unless we have one of our guys in an alt account. We have send a ticket to PGI about waiting 55 minutes with 9 failed to find matches in our assaults. I pug in mediums now because it's all I can play alone if i want to find a match, and forgive me if there's some of you who have witnessed me fail to carry in 50-55 ton mechs when I actually get my hands dirty on the front lines while the rest our our tardo pugs farm damage from the back. You'll get these tardos in DPS mechs hanging all the way back to stat ***** with the worst aim pelting rounds spread all over the enemy mechs and get ~700 dmg and 0 kills and go (LOLOLOLOLO ROFL I GUESS THE LORDS CAN'T CARRY OUR MATCH LOL LOOK HOW MUCH DMG I DID 1111!!!!!!). If you think we can't hang with the top without a group then just challenge us to a 1v1 already.

To the OP, yeah we all think Elo has been tweaked in the last ~3 weeks because our matches are pretty messed up right now. Because we know this game to the soul, we know when something is amiss.

Edited by John Wolf, 16 November 2013 - 07:58 AM.
Name & Shame

[Lightfoot]

So long as no teammates shoot me in the back for practice, I give them a passing grade.

I guess I will start to care about wins and stats when I stop Beta Testing.