Guild Wars 2

Each account’s Match Making Rating (MMR) for Ranked PvP has been frozen since last Friday’s patch. Please correct this issue.

Each account’s Match Making Rating (MMR) has been frozen since last Friday’s patch. Please correct this Anet.

Really Anet, you’re not going to reply ?

What makes you think this is the case? What data do you have to support this?

What’s your Source?

They don’t use MMR anymore.

Sunshine.5014:

They don’t use MMR anymore.

Nope, they use magic pixies that get sent straight to your home and sprinkle lucky clover dust all over your computer to make sure all your matches are Super-funny-happy-time.

bump, plus some more characters to get to fifteen.

I’m assuming people think it happened because the website’s PvP leaderboards are no longer available.

Evan Lesh.3295:

The only changes we made last week is the search range for ladder position.

So you did silently change something that completely broke Ranked PvP for some people.

There are two extremes introduced since Friday: People who now win almost every single match (seriously, this is my friend’s case, he can AFK in a corner and still win matches; went from low Ruby to Legendary x1 since Friday) and people who now lose almost every single match (my case; at one point I had 47 straight losses in Ranked from queuing solo Friday to Monday and not a single close match; all ended with points differences greater than 300)

There are numerous posts about these extreme win/loss streaks all over the forum and Reddit since Friday.

Please look into the issue and into what broke from your new algorithm last Friday.

Evan Lesh.3295:

Your MMR has updated every match. Can you please elaborate?

Can you verify that my MMR has updated after every match? Personally, it is impossible for me to win a ranked match while queuing solo since last Friday. I’m sure your records show my ~40 losses in a row. Recently I found someone in the opposite situation of mine who wins matches simply through unbalanced Match Making always in her favour and we teamed up, winning 8 ranked matches in a row by more than 350 points each time.

Match Making in Ranked PvP was greatly altered in the last patch and I think the Guild Wars community deserves an explanation of how the system was changed.

That’s all I ask for; a little transparency in the patch notes. Thanks

Hmm. It doesn’t explain the huge streaks some people are experiencing. Also, average ranked queue times are significantly longer since Friday. Is the search range upwardly skewed – does it look for players above your division more than for players below your division ? or vice versa ?

Why can’t we see our MMR? We should be able to.

Constantin.5608:

Hmm. It doesn’t explain the huge streaks some people are experiencing. Also, average ranked queue times are significantly longer since Friday. Is the search range upwardly skewed – does it look for players above your division more than for players below your division ? or vice versa ?

I’ve been getting less then 1 minute queue times in Ruby the last few days as solo queue so no, I don’t think it’s skewed at all.

Valkyriez.6578:

Constantin.5608:

Hmm. It doesn’t explain the huge streaks some people are experiencing. Also, average ranked queue times are significantly longer since Friday. Is the search range upwardly skewed – does it look for players above your division more than for players below your division ? or vice versa ?

I’ve been getting less then 1 minute queue times in Ruby the last few days as solo queue so no, I don’t think it’s skewed at all.

What’s your source?

mrauls.6519:

Why can’t we see our MMR? We should be able to.

I completely agree!

mrauls.6519:

Why can’t we see our MMR? We should be able to.

This on soooo many levels

Evan Lesh.3295:

Xiahou Mao.9701:

I’m assuming people think it happened because the website’s PvP leaderboards are no longer available.

That could be. The website wasn’t accurate and shouldn’t have been visible anyways.

I would still really like to know in which way the website had inaccurate reflections…?

What I do know is that when I teamed with friends or guildies who were in a similar points range as I was the matchups were more even and results came down to skill and mistake limitation. When I teamed with players who had significantly higher points the matchups were mostly horrific against very strong opponents.

One thing is certain and that is players who reached legendary with 30-40% win rates which of course reflects poorly for a formal tournament.

I still very strongly feel that players who exploited the system for easier matchups and to avoid pip loss should have been punished nu stripping their wins, but I guess to look into each case of thousands would have been too much work.

Evan, will there be some type of “Season 1 results” board that shows the number of games played, won, lost, % win, and final division for all participants?

Hellbound.5610:

Valkyriez.6578:

Constantin.5608:

Hmm. It doesn’t explain the huge streaks some people are experiencing. Also, average ranked queue times are significantly longer since Friday. Is the search range upwardly skewed – does it look for players above your division more than for players below your division ? or vice versa ?

I’ve been getting less then 1 minute queue times in Ruby the last few days as solo queue so no, I don’t think it’s skewed at all.

What’s your source?

My source?

Soul.9280:

Evan, will there be some type of “Season 1 results” board that shows the number of games played, won, lost, % win, and final division for all participants?

This should be great.

I am in the same situation as Constantin: since last update is near impossible for me win two matches in a row. As I remember it happened only once. I played dozen of matches changing classes and the time of the day, but th normal thing is: I win 1 and lose 3~4. I can see very clear that my team is always the weakest, so I am not the cause of the defeats (not completely and not in this ratio).

Coelho Nat.4697:

I can see very clear that my team is always the weakest, so I am not the cause of the defeats (not completely and not in this ratio).

I remember my last few matches before the patch. I was queuing with this group of relatively new players just to welcome them into a guild and I was playing really well for some reason. I managed to carry them to four straight wins. In one match I managed to win two pips off a team of 5 diamonds while I and my team were Ruby or lower. I solo queued to Tier 2 Diamond in almost straight wins after and then the patch hit…

This pretty much maximized my MMR and since then it has not gone down. I lost ~40 matches in a row after the patch. Then I started scouting for people who were in the exact opposite situation as me and teamed up with a lucky player to breeze through 8 straight wins as a pair yesterday.

A few hours ago I tried one solo ranked match queue just to test my hypothesis and lost 500 to 17 (seventeen) in a match without groups; 5 solo vs 5 solo; all diamonds.

I have to agree that matchmaking broke. I now have a >50% win ratio if i afk and lose every match after one I actually tried to play despite winning 2v1s in those matches due to pugs beginning to suck with apparently extremely wild matchmaking swings. If i akitten the last match my pugs were at my skill lvl and we obliterated the poor people on the other team. It is quite obviously flawed atm.

lloydst.6985:

Constantin.5608:

Coelho Nat.4697:

I can see very clear that my team is always the weakest, so I am not the cause of the defeats (not completely and not in this ratio).

I remember my last few matches before the patch. I was queuing with this group of relatively new players just to welcome them into a guild and I was playing really well for some reason. I managed to carry them to four straight wins. In one match I managed to win two pips off a team of 5 diamonds while I and my team were Ruby or lower. I solo queued to Tier 2 Diamond in almost straight wins after and then the patch hit…

This pretty much maximized my MMR and since then it has not gone down. I lost ~40 matches in a row after the patch. Then I started scouting for people who were in the exact opposite situation as me and teamed up with a lucky player to breeze through 8 straight wins as a pair yesterday.

A few hours ago I tried one solo ranked match queue just to test my hypothesis and lost 500 to 17 (seventeen) in a match without groups; 5 solo vs 5 solo; all diamonds.

Ok so you won basicly every game up to diamond according to you and then find it strange that your mmr is punishing you for that. I doubt that you have won pretty much every game up to diamond as even esl players cant carry some very kittenty pugs so you ahve either been carried by your team have been very lucky or have exploited the system your choice

My necromancer which is what I played almost all ruby has an 80% win rate if including only the league matches. I made it through the first half of ruby in 3 hours and the 2nd half in about 5

To be fair, I finally reached division 4 tonight. I won 4 and lost 2 matches.

The only thing I still think is strange is that in 3 matches one of the teams had like 3 thieves, or 3 rangers, etc. while the another had none or only 1. It is like the class balance is not working.

I haven’t won a single game since switching from my Guard to my Necro, nor were any of the matches remotely close in terms of player skill. I think something must be up.

Ohoni.6057:

I haven’t won a single game since switching from my Guard to my Necro, nor were any of the matches remotely close in terms of player skill. I think something must be up.

I have encountered various players who speculate that class rating is still active….

They would queue on a brand new character for an easy matchup and then change to their main in the match.

If this is the case then it is very shady business indeed since it was very clearly confirmed to be disabled for the league, but with various resources working on processes of different builds and the matchmaking constantly being configured, who knows what is active and inactive.

Evan Lesh.3295:

Hellbound.5610:

I would still really like to know in which way the website had inaccurate reflections…?

The website was using the beta ladder numbers, didn’t get reset with the season, wasn’t using the same formula as the actual season, and was likely bugged as we stopped maintaining the beta algorithm.

I personally never played in the ladder.

From my point of view the leaderboard was functioning just fine when analyzing my results and that of my friends and guildies. The ranking adjusted as matches were won or lost, points improved as progress was gained and of course the win ratio percentage was self explanatory.

Perhaps it may not have been 100% accurate relative to the league formulas, but it gave a reasonable reflection of results and progress through the league.

I do understand that legendary players with low win ratios sheds a bad image on the integrity of the tournament and the dream of many to get into eSports, but instead of just cementing over the flaws, the issue should have been dealt with, because currently there are a great many inept players stuck in ruby and diamond in particular that greatly affect the matches for others due to being carried by some for reduced MMR and easier matchups.

Hellbound.5610:

Evan Lesh.3295:

Hellbound.5610:

I would still really like to know in which way the website had inaccurate reflections…?

The website was using the beta ladder numbers, didn’t get reset with the season, wasn’t using the same formula as the actual season, and was likely bugged as we stopped maintaining the beta algorithm.

I personally never played in the ladder.

From my point of view the leaderboard was functioning just fine when analyzing my results and that of my friends and guildies. The ranking adjusted as matches were won or lost, points improved as progress was gained and of course the win ratio percentage was self explanatory.

Perhaps it may not have been 100% accurate relative to the league formulas, but it gave a reasonable reflection of results and progress through the league.

No it didnt. A leaderboard should reflect all matches played in my oppinion and not just matches played from 1-2 seasons combined together. The leaderboard was ONLY based on the matches played from league and last season before that. They who played the most matches would automaticly have more points than people who wouldnt have so much time to play. To mention one example…a guy i know in the game, belong to top tier mmr where he gets matched with top players, yet his percentile from this league season & last season shows 66% because he didnt play many matches during those seasons. If he enters pvp he will still face high mmr players from esl. This is how the leaderboard is so far from accurate its not even funny…good they removed it.

However it would be nice to see our winrate…

sanctuary.1068:

No it didnt. A leaderboard should reflect all matches played in my oppinion and not just matches played from 1-2 seasons combined together. The leaderboard was ONLY based on the matches played from league and last season before that. They who played the most matches would automaticly have more points than people who wouldnt have so much time to play. To mention one example…a guy i know in the game, belong to top tier mmr where he gets matched with top players, yet his percentile from this league season & last season shows 66% because he didnt play many matches during those seasons. If he enters pvp he will still face high mmr players from esl. This is how the leaderboard is so far from accurate its not even funny…good they removed it.

However it would be nice to see our winrate…

That is not the leaderboard, that is the matchmaking system.

This should be obvious. You just recently went from Ruby to Diamond, and all of a sudden you are loosing matches. Maybe you are just playing against better players now.

Instead of the weird conspiracy theory were ANet is out to get you with faulty programming. Understand that one player making poor decisions at Diamond/Legendary league is more than enough to lead to a loss.

Sure, you will encounter players in these leagues that are not the greatest, but you may be one as well. Especially if you are going on 40 game loosing streaks. Maybe, evaluate what you are doing? Watch a few ESL matches, review MetaBattle, etc…

Archon.6480:

Watch a few ESL matches, review MetaBattle, etc…

I’ve competed in a few ESL Weekly Go Cups and was invited to a monthly once. I don’t understand why you’re attacking me instead of contributing to the discussion. Do I intimidate you or do you troll forums as a hobby ?

Anyone who plays 10+ ranked matches per day will have noticed that Match Making changed since the last patch. As many posters above have mentioned. The purpose of this thread is to discover what changed, not to harass others simply because they queue for Ranked matches and you don’t.

Hellbound.5610:

sanctuary.1068:

No it didnt. A leaderboard should reflect all matches played in my oppinion and not just matches played from 1-2 seasons combined together. The leaderboard was ONLY based on the matches played from league and last season before that. They who played the most matches would automaticly have more points than people who wouldnt have so much time to play. To mention one example…a guy i know in the game, belong to top tier mmr where he gets matched with top players, yet his percentile from this league season & last season shows 66% because he didnt play many matches during those seasons. If he enters pvp he will still face high mmr players from esl. This is how the leaderboard is so far from accurate its not even funny…good they removed it.

However it would be nice to see our winrate…

That is not the leaderboard, that is the matchmaking system.

What im saying is, its not accurate for people who dont belong in top 1000 or higher percentile to have high spots on the leaderboards during seasons. If a person have high mmr but has 66% on the leaderboard thats highly inaccurate in general meanwhile a person who fights low mmr can still be put into the 95+ percentile on leaderboards because its based on points gained during season. So that means, you cant keep track of your friends on the leaderboard. They might belong to a higher mmr and have alot harder fights, or they belong to lower mmr and have easier fights and end of the day both have 95% on the current leaderboards because its all based on points gained aka wins/losses.

If you look at the leaderboards and it says you have 93% but you fight lower mmr, and then you have your friend who has to fight esl players and have alot harder matches than you who gained less points because of that during season and therefore belong to 90%. Based on the leaderboards that would mean you are a better player than your friend bc of the difference in percentile, while in reality thats not the case hence its not accurate the way its built right now.

This should be obvious. You just recently went from Ruby to Diamond, and all of a sudden you are loosing matches. Maybe you are just playing against better players now.

The thing is, moving up in the rankings could explain losing more 1v1 scenarios, but it shouldn’t completely destroy your W/L ratio. I mean, if you were doing well before, like a 5-1 average, then you shouldn’t drop to a 0-12 record over night. Even if the players on the other team are much better than you are and you lose every 1v1 you end up in, the players added to your team should be just as good as those on the other team, and you should still win matches occasionally just from them dragging you across the finish line.

Even if you just camp the starting area AFK you should win more matches than a lot of people are reporting.

sanctuary.1068:

What im saying is, its not accurate for people who dont belong in top 1000 or higher percentile to have high spots on the leaderboards during seasons. If a person have high mmr but has 66% on the leaderboard thats highly inaccurate in general meanwhile a person who fights low mmr can still be put into the 95+ percentile on leaderboards because its based on points gained during season. So that means, you cant keep track of your friends on the leaderboard. They might belong to a higher mmr and have alot harder fights, or they belong to lower mmr and have easier fights and end of the day both have 95% on the current leaderboards because its all based on points gained aka wins/losses.

If you look at the leaderboards and it says you have 93% but you fight lower mmr, and then you have your friend who has to fight esl players and have alot harder matches than you who gained less points because of that during season and therefore belong to 90%. Based on the leaderboards that would mean you are a better player than your friend bc of the difference in percentile, while in reality thats not the case hence its not accurate the way its built right now.

That is where the points section comes in, it is very clear, the better you are, the more you progress, the higher the points.

I definitely do not agree with the high MMR low ranking perception.

It was very easy to keep track of your friends, simply just log into the official site, under the leaderboard you can filter according to friends or guilds.

Ohoni.6057:

The thing is, moving up in the rankings could explain losing more 1v1 scenarios, but it shouldn’t completely destroy your W/L ratio. I mean, if you were doing well before, like a 5-1 average, then you shouldn’t drop to a 0-12 record over night. Even if the players on the other team are much better than you are and you lose every 1v1 you end up in, the players added to your team should be just as good as those on the other team, and you should still win matches occasionally just from them dragging you across the finish line.

Even if you just camp the starting area AFK you should win more matches than a lot of people are reporting.

This guy gets it!

Hellbound.5610:

sanctuary.1068:

What im saying is, its not accurate for people who dont belong in top 1000 or higher percentile to have high spots on the leaderboards during seasons. If a person have high mmr but has 66% on the leaderboard thats highly inaccurate in general meanwhile a person who fights low mmr can still be put into the 95+ percentile on leaderboards because its based on points gained during season. So that means, you cant keep track of your friends on the leaderboard. They might belong to a higher mmr and have alot harder fights, or they belong to lower mmr and have easier fights and end of the day both have 95% on the current leaderboards because its all based on points gained aka wins/losses.

If you look at the leaderboards and it says you have 93% but you fight lower mmr, and then you have your friend who has to fight esl players and have alot harder matches than you who gained less points because of that during season and therefore belong to 90%. Based on the leaderboards that would mean you are a better player than your friend bc of the difference in percentile, while in reality thats not the case hence its not accurate the way its built right now.

That is where the points section comes in, it is very clear, the better you are, the more you progress, the higher the points.

I definitely do not agree with the high MMR low ranking perception.

It was very easy to keep track of your friends, simply just log into the official site, under the leaderboard you can filter according to friends or guilds.

wow you clearly dont understand the system. Im not even gonna try harder to explain to you because i honestly dont think you will understand :/

And btw to answer the op question, they spread the range of the ladder in the matchup thats why you maybe experience more losses or more wins bc you might face/get teamed up with better or worse players.

sanctuary.1068:

wow you clearly dont understand the system. Im not even gonna try harder to explain to you because i honestly dont think you will understand :/

And btw to answer the op question, they spread the range of the ladder in the matchup thats why you maybe experience more losses or more wins bc you might face/get teamed up with better or worse players.

It is quite obvious that you are trying really hard to be very clear here, but Ill be very honest with you, your explanations make zero sense, in fact, it appears to some extent contradictory. It must be a language barrier issue….

Hellbound.5610:

sanctuary.1068:

wow you clearly dont understand the system. Im not even gonna try harder to explain to you because i honestly dont think you will understand :/

And btw to answer the op question, they spread the range of the ladder in the matchup thats why you maybe experience more losses or more wins bc you might face/get teamed up with better or worse players.

It is quite obvious that you are trying really hard to be very clear here, but Ill be very honest with you, your explanations make zero sense, in fact, it appears to some extent contradictory. It must be a language barrier issue….

k…..

Had the same problem here. Was losing like hell on friday/saturday. Then tried changing classes and noticed, that my ranger is having Super Easy matches, unlike my revenant, who always had weak team and ultra strong enemies.

Well…my solution was finding games as ranger and then swapping on my revenant. Almost never lost. Had, like, 1 lose to 10 wins.

And it wasnt because I had bad winrate on my ranger. Tried doing the same on my engineer, who had even worse winrate…and nope. Again had VERY bad matches with him.

Only my ranger was golden.

Sergoros.4398:

Then tried changing classes and noticed, that my ranger is having Super Easy matches, unlike my revenant, who always had weak team and ultra strong enemies.

Ok, this is weird. I played my Ranger last night for the first time in months just for kicks and won 6/7. On my Engi, Necro, Mesmer I have a maybe a 1/10 win rate.

I think we’re getting trolled by some intern at Anet.

Constantin.5608:

Sergoros.4398:

Then tried changing classes and noticed, that my ranger is having Super Easy matches, unlike my revenant, who always had weak team and ultra strong enemies.

Ok, this is weird. I played my Ranger last night for the first time in months just for kicks and won 6/7. On my Engi, Necro, Mesmer I have a maybe a 1/10 win rate.

I think we’re getting trolled by some intern at Anet.

Probably some script, that they forgot to check, when they were making their changes. At least, I found this solution in time to get to the legend.

Constantin.5608:

Sergoros.4398:

Then tried changing classes and noticed, that my ranger is having Super Easy matches, unlike my revenant, who always had weak team and ultra strong enemies.

Ok, this is weird. I played my Ranger last night for the first time in months just for kicks and won 6/7. On my Engi, Necro, Mesmer I have a maybe a 1/10 win rate.

I think we’re getting trolled by some intern at Anet.

So you won 40 games last week and won 10 games this week… sorry but that in no way proves anything, especially after Anet says they’ve only increased Matchmaking’s breather room.

3 thieves 2warrior comp vs a more viable stacked team
Matchmaking doesn’t care if they placed unviable classes together, it just sees numbers on a white board. If it sees a double stack comp on 1 team, it’s going to place another double stack comp on another, ignoring what class they are.

Matchmaking needs to be smarter in that certain comps, like 3 Dh, 3 Necros, 3 Thieves, etc should Never be on the same team vs a combination of 3 Revs, 3 Eles, or 3 Mesmers. We all know which side is going to win, so split that Solo Guardian away from that duo dh crowd and move him to the other team.

Who cares if mmr is suddenly unbalanced on both sides, it’s IS balanced because of team composition.

Who cares if your matchmaking algorithm can’t move players around once they’re picked for the match. It’s called programming, you obviously have the “tech” for it.

Saiyan.1704:

especially after Anet says they’ve only increased Matchmaking’s breather room.

To be fair, same happened in the past. Anet makes some unimportant and small change, but it affects another script, that no one ever thought it would break…

Like Chrono time split breaking in 1 way, while they were trying to fix other aspect of it.