Guild Wars 2

Hi,

I just had three strange matches. I was DuoQing and we faced in all three matches the same DuoQ (let’s call them Team Bob).

• Match 1: We got farmed (504 to 117; Foefire @ ~0:15 CET)
• Match 2: We farmed them (576 to 139; Foefire @ ~0:30 CET)
• Match 3: We won, but it was close (500 to 443, Nifhel @ ~0:45 CET)

Fact 1: Every team consists of players of about the same MMR (inside the pip range).
Fact 2: The opponent team is just inside the same pip-range.
Fact 3: Because of the interaction of Fact 1 & Fact 2, you will end up in a match, where there can be a huge MMR (and therefore skill) difference between the teams. This will lead to blow-out matches.

So how did the above-mentioned series happen?
In Match 1, it seems that Team Bob had the higher MMR, therefore, got the higher ranked teammates and farmed an.
BUT if they had the higher MMR, how could we farm them in the second match?
Did they had an almost equal MMR to us, because of the third match? Then how did the blow-outs happen?

The only solution I could find, was, that the MMR system does not represent our skills anymore.

Edit: The matches happened in Emerald Tier 2.

No one an idea, why those matches happened?

In the past, a blow-out match meant one of the following two things.

• There was a huge difference between the MMR of the teams. This is why we had win-chance based rewards in the last season. System is now forcing this to happen.

• There was a bug in the whole match making system. If you followed the dev posts in the past, they fixed a few a those. Maybe we have gotten a new one?

The only solution I could find, was, that the MMR system does not represent our skills anymore.
^
this…lucky average ppl have got huge winning streak because their are carried by pro player, unlucky average ppl got the opposite.

btw you are lucky, you have had a close match…yesterday on 10 matches, the looser side never gone over 300.

Score difference doesn’t always represent skill difference. Sometimes just a small mistake can lead to a big score difference (“snowball effect”). So both teams had most likely a pretty similar MMR.

vulneraria.4865:

The only solution I could find, was, that the MMR system does not represent our skills anymore.
^
this…lucky average ppl have got huge winning streak because their are carried by pro player, unlucky average ppl got the opposite.

A bit of topic, but yeah, ArenaNet managed to create Elo-Hell.

The term Elo-Hell is very popular in other games. It does not exist there and is only representing the players who are suffering from the Dunning-Kruger effect.

But GW2 has managed to create it.

Let’s take Bob as an example with a very low MMR of 300. Since his MMR is far lower than the average, he will lose more games than he will win (due to the match making algorithm explained in my post above). Because he lost more games, his MMR will suffer more and therefore, he will have the higher chance of ending up on the team with the lower skilled players. (The top players will have moved out of the range, but there will still be a skill difference between the players.)

Even if Bob would finally learn2play at the same level as an average player, Bob has no chance to escape this cycle since most of the time, his teammates will be at the bottom of the MMR for a reason.

In the past, the teams would have been equal regarding the MMR. There, it would have made a huge difference for Bob to improve his gameplay.

Teutos.8620:

vulneraria.4865:

The only solution I could find, was, that the MMR system does not represent our skills anymore.
^
this…lucky average ppl have got huge winning streak because their are carried by pro player, unlucky average ppl got the opposite.

A bit of topic, but yeah, ArenaNet managed to create Elo-Hell.

The term Elo-Hell is very popular in other games. It does not exist there and is only representing the players who are suffering from the Dunning-Kruger effect.

But GW2 has managed to create it.

Let’s take Bob as an example with a very low MMR of 300. Since his MMR is far lower than the average, he will lose more games than he will win (due to the match making algorithm explained in my post above). Because he lost more games, his MMR will suffer more and therefore, he will have the higher chance of ending up on the team with the lower skilled players. (The top players will have moved out of the range, but there will still be a skill difference between the players.)

Even if Bob would finally learn2play at the same level as an average player, Bob has no chance to escape this cycle since most of the time, his teammates will be at the bottom of the MMR for a reason.

In the past, the teams would have been equal regarding the MMR. There, it would have made a huge difference for Bob to improve his gameplay.

Well, I think Bob is kittened.

On another note, I think this is the worst system so far. What’s the point of creating unbalanced games by default? I don’t get it at all.

Yeah and we have a huge number of Bob´s now, even some that actually can play well. Many will give up … Its hardcore driving off non highly competative players from ranked.
Even some very good one´s get bored from constant 500:xx.

I’m still having a hard time believing that people who go 0-10 or worse are ONLY losing because of teammates. You can be quite impactful yourself if you truely are a great player. People are bandwaggoning on this “ITS MMR NOT ME” so hard right now.

I’ve personally steamrolled through amber with like , 10-2 or w/e. Then second day was a bit back & forth , overal gained 2-3 tiers in emerald but did not steamroll at all. then 3rd day it was hard games for the most part, BUT i managed to win 80% of them and got to saphire. Among these hard games just to paint a picture, we had an engineer who litterally ended the game on 0 score. He played all game long, died all the time, and ended on 0 score. Meanwhile our druid was also dying quite frequently but he did do something atleast. ended the game 500-470. (and to counter this, all the enemies had a score of over 50, so its not like it was balanced out by our enemies having equally trash players)

What does this tell me? If i was not considerably better than my opponents and i did not win every 1v1, get clutch finishes/revives in teamfights, then this game would have easily been lost.

Ontop of that, conquest is a gamemode that can be very snowbally, plus people are giving up a bit earlier than before, or are simply tilting (just lost 10 games in a row and now you just lost that middle fight AGAIN!!??). This alone is a big factor in being able to win games 500-100 +-.

Laraley.7695:

On another note, I think this is the worst system so far. What’s the point of creating unbalanced games by default? I don’t get it at all.

The only reason, we have the current mess, is because they tried to salvage the mess we had last season.

At the end of the season, the best players should be on the top of the league system, but last season was just a point-farm where it was actually easier to reach Legendary if you were not a top player. (With the new season, good players who play a lot of games will be on top, not the best players.)

What is every other (major) game doing:
1.) Let players play with and against players of about the same MMR.
2.) Distribute the reward (the amount of pips you gain and lose) depending on your MMR.
3.) A Challenge-League for players who have reached Legendary to really have the highest-rated players on top.

Let’s take average Carl as an example: He has a MMR of 1.000.
Now, ArenaNet says: with a MMR of 1.000 you deserve to reach Ruby.

• Till Carl reaches actually Ruby, he will gains 3 pip per win and still only loses 1 pip. (The whole time, he will fight with and against other players with about the same MMR. Therefore, his match quality will be much better than ours.)
• Now Carl has finally reached Ruby, now he gains 1 pip per win and still looses 1 pip.
• Carl finally get’s really lucky, he manages to have a win streak on Ruby. He reaches Ruby Tier 4, but he still hasn’t gotten any better (still MMR of 1.000). He will still gain 1 pip per win, but he will lose 3 pips.
  (Other games are not using 5 pips per tier, but e.g. 100 league points, therefore it is easier to grade how many points they get per win and how many points they lose.)

I really don’t understand why we don’t have such a system.

Dril.2495:

Ontop of that, conquest is a gamemode that can be very snowbally, plus people are giving up a bit earlier than before, or are simply tilting (just lost 10 games in a row and now you just lost that middle fight AGAIN!!??). This alone is a big factor in being able to win games 500-100 +-.

This is a big issue at the moment. It was already mentioned before leaguage start imminiateley after system details got shown. Currently there is a lack of motivation to fight hard to the end. This motivation comes from the following:
1.) A player must feel he has an impact on the game. (positive and negative!)
2.) A player should have challenging games (not constant steamrolls to either side).
3.) A player should be rewarded for fighting to the end like the 300+ point mark last .

We have lost all 3 points in the new system. 1 is also a spiraling result of 2 + others giving up.

Dril.2495:

I’m still having a hard time believing that people who go 0-10 or worse are ONLY losing because of teammates.

Ofc not. Most of those players are playing worse than their opponents. The point is, that once you are in that down spiral, you have a hard time to get out of it.

In the past it was average mmr vs average mmr. Now your individual goal was to improve yourself, prove, that the system was wrong and that you deserve to have a higher mmr and therefore, win the match. Now, most of the time, it will not be possible, unless the teams are almost equally rated (which they might be after in a week or two).

I personally, see the matches in the perspective of “match quality”. That means that the match was interesting and you felt, that your own gameplay made the difference for you winning. Currently, most matches do not fulfill the second part. It does not matter on which side (winning or losing) of the steam-rolling match you are. Most matches, that no matter what, it would have ended 500-<150 anyway.

Maybe in a week, the matches will get better, but is that reason enough, that I don’t call out the devs for implementing a system, where I have first to play through 100 matches to get high-qualitative matches when every other game out there manages to NOT have the same problem?

Ontop of that, conquest is a gamemode that can be very snowbally, plus people are giving up a bit earlier than before, or are simply tilting (just lost 10 games in a row and now you just lost that middle fight AGAIN!!??). This alone is a big factor in being able to win games 500-100 +-.

In theory, you are totally right, but my experience in last season tells me a completely different story. ArenaNet has already proven, that they can generate a match making, where most matches are high-qualitatively and NOT a steam roll.

It happened a similar thing to me yesterday, and 3 times.
- In the morning, I got against a guy, and next 2 consecutive with him. First lost, but it was close. Second and third we destroyed the enemy.
- Then, in the afternoon I got with a duo twice and a consecutive third against them. First 2 we lost, and third they won.
- But even more, I later got with another duo twice consecutive, losing both matchups.
The amount of consecutive fights you get paired with some people is unreasonable.

Then there’s the fact that your overall performance doesn’t mean anything in pips. You can play an excellent match, but as your teammates play horribly you still lose a pip. You can be horrible, but as you’re carried by some duo or pro player in your team you get a pip….

Ansau.7326:

The amount of consecutive fights you get paired with some people is unreasonable.

While this is also a problem, that’s not what I’m talking about. I am just using the three consecutive fights to raise the question if the MMR behind grouping you up with your teammates is actually working properly.

I beat a certain individuum 500:127 or so, next game i had him on my own team calling me bad and going afk, so i lost ~280:500 in 4v5 :d
On topic: given that the vast majority is amber or emerald, it is pretty hard for match making to form teams with high skill differences.
Blow out matches have and will always happen even between similar teams (if you dont believe me, look at pro league or the pro league finals, especially foefire which is in my experience the most played map in ranked queues). It is a result of the gamemode/game design.contrary to mobas or gw1 there is no possibility to retreat into npcs or a tower if you are pressured, as such it is fairly likely that the match snowballs, especially in unorganized play where people just run to one of the three nodes after respawning while some of their teammates are still dead.
Additionally you have good and bad players in the same divisions because the season just started and everyone starts in amber regardless of their division in the previous season. If you have a player on one team that constantly binds 2 of the opposing players, chances are that the team with the said player is going to win. Same thing if you have someone who is great at supporting teamfights through good focus and cleave/stomps or heals and rezzes.

Erzian.5218:

Additionally you have good and bad players in the same divisions because the season just started and everyone starts in amber regardless of their division in the previous season.

Again: The problem here lies in the above-mentioned Facts.

Your teammates have about your own mmr while you are facing a team (which in itself has about the same MMR) inside your pip range.

Since good and bad players are currently in the first few divisions, most of the time there is a huge MMR-gap between those two teams fighting each other leading to one-sided matches. Each team does not consist of good&bad players. Good players are fighting bad players. His is how the matchmaking is currently working and why we currently have so many one-sided matches.

Those three consecutive matches are just an example, that the theory does not always happen in practice. The question that remains is why. The most likely solution is, that we were rated about equal, but it just snowballed out of hand. Being in those matches, I personally don’t think, that was the case. It felt more, that there was a bug in the match making process.

According to the season 2 announcement, matchmaking is primarily based on pips not mmr apart from that, if the amount of players available at a certain time is so low at you have the same players over and over again that is not an issue of matchmaking but in that case i’d rather have teams in which the players are of equal skill instead of forcing ‘balanced’ teams and a 50:50 win/lose chance for both teams, because that is exactly what we had last season and is exactly what made the last season a bad joke. In a league system that is intended to represent player skill, it is fundamentally wrong to put the best players on a team with the worst ones that are currently available in order to obtain a 50:50 win/loss ratio. If there are only 10 players available, then the best 5 of them should be put into a team and win to get to a higher division. I know, that sounds arrogant (and frustrating for the bad players), but it is necesarry if you want leagues to represent player skill.

Erzian.5218:

According to the season 2 announcement, matchmsking is primarily based on pips not mmr

Please don’t spread wrong information.

Gaile Gray

While we’ll be using divisions to match you against your immediate opponents, we’ll still use your MMR to place you on teams with similarly skilled players (from your division point range) to help ensure that you’re not forced to play with individuals that are of a much higher or lower skill than you.

Source: https://forum-en.gw2archive.eu/forum/game/pvp/Upcoming-Changes-for-PvP-League-Season-2

Erzian.5218:

According to the season 2 announcement, matchmsking is primarily based on pips not mmr apart from that, if the amount of players available at a certain time is so low at you have the same players over and over again that is not an issue of matchmaking but in that case i’d rather have teams in which the players are of equal skill instead of forcing ‘balanced’ teams and a 50:50 win/lose chance for both teams, because that is exactly what we had last season and is exactly what made the last season a bad joke. In a league system that is intended to represent player skill, it is fundamentally wrong to put the best players on a team with the worst ones that are currently available in order to obtain a 50:50 win/loss ratio. If there are only 10 players available, then the best 5 of them should be put into a team and win to get to a higher division. I know, that sounds arrogant (and frustrating for the bad players), but it is necesarry if you want leagues to represent player skill.

totally wrong sorry….
i have 10 player i put the top 5 in one team the other in second team, they play 100 times, score 100-0.
even if one or two of the second team have improved or have thinked a better build. (or simply in last game of season 1 they have tested some wrong build and they have decreased personal mmr)they are dead.
if i mix the team then i can say player 1 win 60%
player 2 win 56%
player 3 win 55%
and so on.

Erzian.5218:

According to the season 2 announcement, matchmsking is primarily based on pips not mmr apart from that, if the amount of players available at a certain time is so low at you have the same players over and over again that is not an issue of matchmaking but in that case i’d rather have teams in which the players are of equal skill instead of forcing ‘balanced’ teams and a 50:50 win/lose chance for both teams, because that is exactly what we had last season and is exactly what made the last season a bad joke. In a league system that is intended to represent player skill, it is fundamentally wrong to put the best players on a team with the worst ones that are currently available in order to obtain a 50:50 win/loss ratio. If there are only 10 players available, then the best 5 of them should be put into a team and win to get to a higher division. I know, that sounds arrogant (and frustrating for the bad players), but it is necesarry if you want leagues to represent player skill.

You are right, but without dividing known good players from the rest and ranking new players average so they can drag down others, this leads to a lot of frustration and booring matches. This is a very bad thing for player motivation an will lose players.

And in addition this will continue in all divisions. It results in a challange system and is focused on teams. High MMR will always be at an advantage especially in solo Q.

more or less this :

anet have a theory (mmr) and have build an experiment (system matchmaking) that force the reality to confirm his theory.
instead you need an experiment that confirm or adjust your initial theory.

Teutos.8620:

Erzian.5218:

According to the season 2 announcement, matchmsking is primarily based on pips not mmr

Please don’t spread wrong information.

Gaile Gray

While we’ll be using divisions to match you against your immediate opponents, we’ll still use your MMR to place you on teams with similarly skilled players (from your division point range) to help ensure that you’re not forced to play with individuals that are of a much higher or lower skill than you.

Source: https://forum-en.gw2archive.eu/forum/game/pvp/Upcoming-Changes-for-PvP-League-Season-2

“from your division point range”
hence the word primarily but thank you for a more detailed depiction of the matchmaking

Teutos.8620:

Erzian.5218:

According to the season 2 announcement, matchmsking is primarily based on pips not mmr

Please don’t spread wrong information.

Gaile Gray

While we’ll be using divisions to match you against your immediate opponents, we’ll still use your MMR to place you on teams with similarly skilled players (from your division point range) to help ensure that you’re not forced to play with individuals that are of a much higher or lower skill than you.

Source: https://forum-en.gw2archive.eu/forum/game/pvp/Upcoming-Changes-for-PvP-League-Season-2

“from your division point range”
hence the word primarily but thank you for a more detailed depiction of the matchmaking

anyways: the system is flawed either way as you cannot prevent abuse without punishing players who play “at the wrong time” (= when there are not enough people available to match players of equal skill with and against each other), if your goal for leagues is to represent skill.

Erzian.5218:

Gaile Gray

While we’ll be using divisions to match you against your immediate opponents, we’ll still use your MMR to place you on teams with similarly skilled players (from your division point range) to help ensure that you’re not forced to play with individuals that are of a much higher or lower skill than you.

“from your division point range”
hence the word primarily but thank you for a more detailed depiction of the matchmaking

The way they are currently using the MMR makes a huge difference to how matches worked in the past.

E.g. 50 players are currently queuing inside your division point range. You are the top player who is queuing: your teammates will be the top 2-5th. Whom are the top 5 players now facing? Anyone inside the division point range, maybe the 5th worst player of those 50. -> 1 sided match = bad / low-qualitative match.

Erzian.5218:

anyways: the system is flawed either way as you cannot prevent abuse without punishing players who play “at the wrong time” (= when there are not enough people available to match players of equal skill with and against each other), if you goal for leagues is to represent skill.

No, it is not. Every other (major) game manages to solve that problem. Why do we not deserve such a system too? Why do we have to suffer through bad matches?

Seems pretty simple to me lower and middle MMR players are not getting higher MMR teammates anymore like they sometimes did with MMR averaging and with the 50/50 outcome A-net tried to force on people blunted because of opponents being pip placement dependent some people are getting a rude awakening. That being said just give it a few days and I would imagine most people will go back to winning 45-55% of there games again.