Guild Wars 2

It is not secret that games stacks all odds against you if your win rate is above 50% forcing you to play against good players and having really new players on own team.

This becomes even more worse with the fact that solo players face premades pretty much every match.

I just want to be done with daily but i can’t even do this (unless i run premade myself) due to the way matchmaking is atm.

Trying to make people form more teams is fine and all but pvp shouldn’t be have a team or don’t play at all :/

Suggestion:
- allow first match per day ignore 50% win rate rule :/
- separate soloq and teamq

They should apply the Matchmaking algorithm they have to finding precursors as drops and apply the stark RNG from PVE to PVP.

So you want free daily…. I’m usually over 50% winrate…. round 60% usually and I don’t seem to notice that When I’m trying to do a daily, although I could give two kittens about quests I do pvp to pvp. Also they said class mmr is in effect… didn’t they?

Separate SoloQ and TeamQ is the best choice. It was discused many times, and I’ve been alwasy repeating we need to have GW1 like ques, where:
→ after you win you can vote if you want to stay with your SoloQ team for next match;
→ once you lose your SoloQ team is disbanded;
→ for each game you win you get reward track bonus;
→ once you get X wins in a row (e.g. 10), you end in TEAM que;
→ TEAM que has higher rewards (e.g. x3).

MM where normal SoloQ teams are ‘just’ matched vs Premades is not acceptable in game aiming at fair competition…

NeXeD.3042:

So you want free daily…. I’m usually over 50% winrate…. round 60% usually and I don’t seem to notice that When I’m trying to do a daily, although I could give two kittens about quests I do pvp to pvp. Also they said class mmr is in effect… didn’t they?

it is not free, you still need to win match but at least you won’t have to try to carry (which is impossible) 4 people who can’t even gain a point whole game….

i did have teams in the past that were getting farmed at spawn and afk’d out obviously… by all means i can’t solo 5 people…

I can’t say I’ve ever queued and got a match where We got farmed like that unless we had a dc and everyone else just went afk, even then I’m still gonna keep pushing lol

Cynz.9437:

- allow first match per day ignore 50% win rate rule :/

There is no such rule. This is how a good match making works.

1.) You win vs your opponents? Your mmr-rating increases, and you will face stronger opponents.
2.) You lost? you will face weaker opponents the next time.

If you win about 50% of your matches, the match making system is working perfectly fine to find you opponents about your own strength.
If you win more than 50% → The match making can either not find suitable opponents (due to a small community); the match making process is not working or your own mmr-rating was still too low, and has therefore to increase further.

Just lost 3 followed matches against this premade, first was a fair fight, the 2nd was a 4vs5, in courtyard we cant do much vs a organized team…

Sneakier.9460:

Just lost 3 followed matches against this premade

How do you know, that they were a 5 player premade? Maybe they were 3-2.
Did you look at your team, and if you were playing with a premade? Maybe 4-1?

4th game… same premade… 1 of my team disconected…

again 4vs5 against a premade… rage quit! no daily today!

Teutos.8620:

Sneakier.9460:

Just lost 3 followed matches against this premade

How do you know, that they were a 5 player premade? Maybe they were 3-2.
Did you look at your team, and if you were playing with a premade? Maybe 4-1?

I got the same team 4 times now… look at their guild tag…

Teutos.8620:

Cynz.9437:

- allow first match per day ignore 50% win rate rule :/

There is no such rule. This is how a good match making works.

1.) You win vs your opponents? Your mmr-rating increases, and you will face stronger opponents.
2.) You lost? you will face weaker opponents the next time.

If you win about 50% of your matches, the match making system is working perfectly fine to find you opponents about your own strength.
If you win more than 50% -> The match making can either not find suitable opponents (due to a small community); the match making process is not working or your own mmr-rating was still too low, and has therefore to increase further.

Thank you! I find it odd that you don’t understand that, Cynz. Why would you want to have a ratio above 50%? If you improve gradually, you should in the long run have a win ratio a little over 50%, but certainly not over 60 or 70%. Every Glicko or Elo based MM system works like this, and that’s exactly what we want. I don’t want to win all the time. I want to play against equally skilled players.

The only possible reason to have an a win ratio well over 50% would be to queue in one of the best teams. In that case, the system couldn’t find you enough good teams to play against. This is a problem for the small minority we see in ESL, but for the rest of us, 50% is fine.

Sneakier.9460:

Teutos.8620:

Sneakier.9460:

Just lost 3 followed matches against this premade

How do you know, that they were a 5 player premade? Maybe they were 3-2.
Did you look at your team, and if you were playing with a premade? Maybe 4-1?

I got the same team 4 times now… look at their guild tag…

You can see 4 opponents and two different tags on the screenshot.
Edit: On a sidenote, I’d love to recommend you to use the thief-meta build.

Sorel.4870:

[snip]

Winrate and match quality are two different thing.
OP isn’t complaining about winrate actualy, but horrible MM giving him constantly not equaly skilled teammates pulling him down instead of stronger enemies.

Sorel.4870:

Teutos.8620:

Cynz.9437:

- allow first match per day ignore 50% win rate rule :/

There is no such rule. This is how a good match making works.

1.) You win vs your opponents? Your mmr-rating increases, and you will face stronger opponents.
2.) You lost? you will face weaker opponents the next time.

If you win about 50% of your matches, the match making system is working perfectly fine to find you opponents about your own strength.
If you win more than 50% -> The match making can either not find suitable opponents (due to a small community); the match making process is not working or your own mmr-rating was still too low, and has therefore to increase further.

Thank you! I find it odd that you don’t understand that, Cynz. Why would you want to have a ratio above 50%? If you improve gradually, you should in the long run have a win ratio a little over 50%, but certainly not over 60 or 70%. Every Glicko or Elo based MM system works like this, and that’s exactly what we want. I don’t want to win all the time. I want to play against equally skilled players.

The only possible reason to have an a win ratio well over 50% would be to queue in one of the best teams. In that case, the system couldn’t find you enough good teams to play against. This is a problem for the small minority we see in ESL, but for the rest of us, 50% is fine.

ideally, if you win, yes you should be fighting tougher opponents… but it would also make sense to have teammates of similar skill level

what happens in gw2:
→ you get abogve 50% win rate
→ game puts you against better opponents but actually worse teammates than you had before… this is my actual issue with system

i don’t mind facing tougher opponents, i just mind trying to fight esl teams while pugs on my team can’t get a single point whole game…

Teutos.8620:

Sneakier.9460:

Teutos.8620:

Sneakier.9460:

Just lost 3 followed matches against this premade

How do you know, that they were a 5 player premade? Maybe they were 3-2.
Did you look at your team, and if you were playing with a premade? Maybe 4-1?

I got the same team 4 times now… look at their guild tag…

You can see 4 opponents and two different tags on the screenshot.
Edit: On a sidenote, I’d love to recommend you to use the thief-meta build.

I use meta on conquest, they were a premade, i did 4 games followed against the same team, with the same ppl.
2 Mesmer Shater, 1 well necro, 1 D/D cele and 1 hambow

stop trolling plz…

Sorel.4870:

Teutos.8620:

Cynz.9437:

- allow first match per day ignore 50% win rate rule :/

There is no such rule. This is how a good match making works.

1.) You win vs your opponents? Your mmr-rating increases, and you will face stronger opponents.
2.) You lost? you will face weaker opponents the next time.

If you win about 50% of your matches, the match making system is working perfectly fine to find you opponents about your own strength.
If you win more than 50% -> The match making can either not find suitable opponents (due to a small community); the match making process is not working or your own mmr-rating was still too low, and has therefore to increase further.

Thank you! I find it odd that you don’t understand that, Cynz. Why would you want to have a ratio above 50%? If you improve gradually, you should in the long run have a win ratio a little over 50%, but certainly not over 60 or 70%. Every Glicko or Elo based MM system works like this, and that’s exactly what we want. I don’t want to win all the time. I want to play against equally skilled players.

The only possible reason to have an a win ratio well over 50% would be to queue in one of the best teams. In that case, the system couldn’t find you enough good teams to play against. This is a problem for the small minority we see in ESL, but for the rest of us, 50% is fine.

You are confusing solo queue and team queue here. It is very easy to have win rates of 60% or 70% if you queue as part of a team. It is not easy to have them as a solo player. The higher your MMR as a solo player, the more likely you’ll run into a premade, which at higher levels is pretty much a guaranteed loss.

The solution is a return to solo and team queue.

Sneakier.9460:

stop trolling plz…

Why am I trolling? I am asking you serious questions, and you have barely answered them. How do you want to be treated on the forum, should I fall on my knees and kiss you?

You my friend, sounds like the saying “mad cuz bad” is perfectly for you.

Everyone with a 50% win rate will virtually have similiar mmr. The average win rate on the leaderboards is 50%.

Half the time, your mmr is no different than anyone else.

Saiyan.1704:

Everyone with a 50% win rate will virtually have similiar mmr. The average win rate on the leaderboards is 50%.

Half the time, your mmr is no different than anyone else.

That’s still not how the MMR is working.

Teutos.8620:

Sneakier.9460:

stop trolling plz…

Why am I trolling? I am asking you serious questions, and you have barely answered them. How do you want to be treated on the forum, should I fall on my knees and kiss you?

You my friend, sounds like the saying “mad cuz bad” is perfectly for you.

/facepalm
gota check ur post history

Nope, you just turned off your brain…

4 matches in a row against the same 5 ppl, the print screen was taken at third
4 of those ppl were from the same guild
it was a premade

the only match we had a chance to win was first one but we lost, they rushed our lord but fair fight
The 2nd and 4th were 4vs5…
the 3rd was slaughter house, they had superior strategy and build synergi (remenber premade?)

question?

PS: i wasn’t in a party, i never party in sPvP, i play mostly to try out new builds and have fun

Teutos.8620:

Saiyan.1704:

Everyone with a 50% win rate will virtually have similiar mmr. The average win rate on the leaderboards is 50%.

Half the time, your mmr is no different than anyone else.

That’s still not how the MMR is working.

It sure feels like that’s how it’s working.
Why should I not believe that otherwise when you/we have no source data to go off of?

Sneakier.9460:

Nope, you just turned off your brain…

Now you have to be offensive? What happened to being friendly and constructive?

4 matches in a row against the same 5 ppl

I had 3 matches in a row playing with 2 friends, where we got the same two player premade into out team. On another day I was in a premade with 3 friends, and over the whole evening we got a rotation of three solo-players.

question?

Yes, have you checked your own team? As often as you play f.e. vs a 4 player premade, you end up with being a 4 player premade on your team.

Don’t know how much you know, so please don’t feel offensive, just gonna repeat how the match making system works:

• Everyone has a rating. The rating system in gw2 is based on http://en.wikipedia.org/wiki/Glicko_rating_system
  So f.e. a player with rating of 3.000 is good, a player with 1.500 mediocre.

• The match making system tries to throw people with about the same rating together in a team and tries to distribute the ratings among the two teams as even as possible. F.e. the system finds 10 player with a rating about 2.000 and throws them together into a match

• If multiple people are queuing together (=premade) the system thinks, that they have therefore an advantage, and their rating will get adjusted higher.
  F.e. 4 players with a 1.600 rating are queuing together, so the system will place them as a group at a 2.000 rating. So the individual skill of the premade-players is not as high as the solo queue players they are facing / playing with.

Therefore you will end up with 4 players playing as a group with a 1.6k rating and 6 players with a 2k rating. Who will win?

Now to some of the most common statements:
“The higher your rating, the higher the chance you will face premades.” Will, yes, due to the rating adjustment, f.e. a player rated with 500 will most likely not face a premade since once their group rating is adjusted, they will most likely have a rating > 500.

“Premades will always win”: Well no, I have won against them and lost with them. Since the individual skill of the players is much lower.

Just for completeness:
Yes, the system has problems. F.e. against which teams should a full Abjured / TCG / oRNG play?
The playerbase is small, so you can end up with a match from player with 1k to 3k rating in the same team (ofc after waiting for 8+minutes).
The system takes some time to rank players. New players will almost always be not ranked at their appropriate ranking.

Saiyan.1704:

It sure feels like that’s how it’s working.
Why should I not believe that otherwise when you/we have no source data to go off of?

You should start with http://en.wikipedia.org/wiki/Glicko_rating_system

There is a small example if you don’t want to read about the glicko 2 rating system:
Every player has his own individual rating.

• Player A with a 2k rating will face opponents with a 2k rating and play with players with a 2k rating. If he is matches perfectly at 2k, the will end up with a 50% win chance.
• The same is true for player B with a 1k rating. Opponents and teammates ~1k, perfectly matches = 50% win chance.
• Player C plays at a 1k rating, but he now uses a meta build. He wins against his current 1k opponents, climps up the rating-ladder and therefore has a winchance okitten0%

just was in match, it was 5v5 before people quit

yep, matchmaking works ~

I think GW2 must be the only game which doesn’t increase rewards for better play. I mean…the system is designed to keep you at a 50% win rate ideally. Fine, so far so good. But the astounding part is that the progression rewards are the same regardless of your MMR!

Inexplicably, it means that being a more skilled player earns you the same rewards per unit of time than a lesser skilled one, given the ideal 50% win rate! It would make sense if progress in your track was affected by your MMR – but no. It doesn’t work that way.

Inb4 people start saying “Rewards in PvP don’t matter” or “Skilled players don’t care for rewards”. If rewards don’t matter, then Anet shouldn’t give them at all. No reward track progression, no gold…nothing. And while you’re at it, take out rewards for PvE as well. Just make everything free. Because…we’re all supposed to play just for the pure joy of it right? Right?

Bottom line: The reward structure for PvP is deeply messed up. If you’re going to have rewards, then they must be rationalized. Their very presence is enough of an indication that Anet feels “just the play” isn’t enough.

Now why don’t they do an actual good job of it and increase the rewards for those with higher MMRs?

Teutos.8620:

Saiyan.1704:

It sure feels like that’s how it’s working.
Why should I not believe that otherwise when you/we have no source data to go off of?

You should start with http://en.wikipedia.org/wiki/Glicko_rating_system

There is a small example if you don’t want to read about the glicko 2 rating system:
Every player has his own individual rating.

• Player A with a 2k rating will face opponents with a 2k rating and play with players with a 2k rating. If he is matches perfectly at 2k, the will end up with a 50% win chance.
• The same is true for player B with a 1k rating. Opponents and teammates ~1k, perfectly matches = 50% win chance.
• Player C plays at a 1k rating, but he now uses a meta build. He wins against his current 1k opponents, climps up the rating-ladder and therefore has a winchance okitten0%

Appreciate the example but I understand that. What i’m taking about is how do we know whether or not our individual rating is working as intended? It could be inaccurate, much so, that it directly correlates to win percentages. Theoretically, we should have what the algorithm states but we literally have no idea how accurate it is because there are no visual rating numbers, other than win percentages, for us to go off of.

I know that’s speculation but I gathered that through visual experience. If every person with a 54% win rate has greater individual rating than everyone with a 53% win and lower, then the algorithm is not working as intended. My question is, exactly how often does a low win percentage player have a higher individual rating than a higher win percentage player? Does a gain in rank points govern a players individual rating? I certainly hope that’s not the case.

Example
On the current leaderboard, does Rank #41 with a 70% win rate have a higher individual rating than Rank player #5 with a 60% win rate? For that matter, does the #41 player have a rating higher than anyone in the top 10? If he does then that further strengthens my theory about win percentages having a direct correlation to a player’s individual rating.

Cynz.9437:

what happens in gw2:
-> you get abogve 50% win rate
-> game puts you against better opponents but actually worse teammates than you had before… this is my actual issue with system

i don’t mind facing tougher opponents, i just mind trying to fight esl teams while pugs on my team can’t get a single point whole game…

Three factors are identifiable here:

1) The match-making system tries to have an equal average MMR between teams. You can very well have bad team mates and still have the same average MMR as the other team.

2) GW2 pvp community is small, and therefore the skill variance in a team can vary greatly depending on what time you play, which queue (people are trying out professions in unranked), and MMR.

3) The “OMG-my-teammates-are-scrubs-and-my-opponents-are-pros-with-cheese-builds” factor, which is a normal and harmless delusion everyone experiences in competitive games.

Cynz.9437:

It is not secret that games stacks all odds against you if your win rate is above 50% forcing you to play against good players and having really new players on own team.

Well if that is not secret you can surely share some evidence for that with us.

bhagwad.4281:

I think GW2 must be the only game which doesn’t increase rewards for better play. I mean…the system is designed to keep you at a 50% win rate ideally. Fine, so far so good. But the astounding part is that the progression rewards are the same regardless of your MMR!

Inexplicably, it means that being a more skilled player earns you the same rewards per unit of time than a lesser skilled one, given the ideal 50% win rate! It would make sense if progress in your track was affected by your MMR – but no. It doesn’t work that way.

Inb4 people start saying “Rewards in PvP don’t matter” or “Skilled players don’t care for rewards”. If rewards don’t matter, then Anet shouldn’t give them at all. No reward track progression, no gold…nothing. And while you’re at it, take out rewards for PvE as well. Just make everything free. Because…we’re all supposed to play just for the pure joy of it right? Right?

Bottom line: The reward structure for PvP is deeply messed up. If you’re going to have rewards, then they must be rationalized. Their very presence is enough of an indication that Anet feels “just the play” isn’t enough.

Now why don’t they do an actual good job of it and increase the rewards for those with higher MMRs?

until anet makes you weapons that all look like signs that say “I need this weapon to show other people I’m good because my skill alone isn’t enough” you can always try for that glorious hero armor set.

Also if you don’t enjoy playing pvp for the pvp then why the hell are you doing it. The truly competitive players don’t give two puppies about aesthetics, they know they are good and prove it in tourneys. You should try the same.

@ Teutos.8620

What you’re saying is how the system should be working. In practice there are problems and I suspect it’s not just a population problem.

I’ve partial queued with friends and we get matched with full teams around our skill level. But then we get people with no clue what they’re doing to round out the team. That happens more often than you’d think and it’s incredibly frustrating. Something is definitely broken.

I’ve looked at the matchmaking pseudo code on the wiki and there are a few issues with it which I’ve pointed out on the forums. We still don’t know how ANet applied Glicko2 to individuals on a team.

Exedore.6320:

What you’re saying is how the system should be working. In practice there are problems and I suspect it’s not just a population problem.
[…]
That happens more often than you’d think and it’s incredibly frustrating. Something is definitely broken.

Maybe we should generate a thread again where we report matches, we think, that should not have happened?

How about we only post that kind of matches, that fall into the following criteria:

• all 10 players participated during the whole match, so no 4v5s. 4v5s are a pain, but they are not a match making problem.
• The loosing team did not reach 150 points. The end score has to represent how bad a team got farmed.

And the post must include:

• Your account name you played the match with, unless it is the same as the one posting.
• The time of the match including your time zone.
• Name of the PvP-Map where the match took place
  This will help the devs to find the problematic match, and take a look at the matchmaking.

Nice to also include:

• How many players of both teams were not invite-able to your party.
• IP-Address of the pvp server (/ip in chat), not sure if that would help the devs to find the match or is useless information.

Teutos.8620:

Exedore.6320:

What you’re saying is how the system should be working. In practice there are problems and I suspect it’s not just a population problem.
[…]
That happens more often than you’d think and it’s incredibly frustrating. Something is definitely broken.

Maybe we should generate a thread again where we report matches, we think, that should not have happened?

How about we only post that kind of matches, that fall into the following criteria:

• all 10 players participated during the whole match, so no 4v5s. 4v5s are a pain, but they are not a match making problem.
• The loosing team did not reach 150 points. The end score has to represent how bad a team got farmed.

And the post must include:

• Your account name you played the match with, unless it is the same as the one posting.
• The time of the match including your time zone.
• Name of the PvP-Map where the match took place
  This will help the devs to find the problematic match, and take a look at the matchmaking.

Nice to also include:

• How many players of both teams were not invite-able to your party.
• IP-Address of the pvp server (/ip in chat), not sure if that would help the devs to find the match or is useless information.

That’s all been done before and that area is in regards to how good matchmaking really is.

My main concern is whether or not individual player rating is accurate for matchmaking to do its job. Sure the numbers are saying it is but, is it really?

If I end the majority of my 200 games with a point spread of 400-450 my individual rating should be higher than a person who ends his games at 300-399

I’ve had pseudo code questions (and posted code) but the devs never answered back. I believe a 5 point difference (rather than 100) should be taken into account.

NeXeD.3042:

bhagwad.4281:

I think GW2 must be the only game which doesn’t increase rewards for better play. I mean…the system is designed to keep you at a 50% win rate ideally. Fine, so far so good. But the astounding part is that the progression rewards are the same regardless of your MMR!

Inexplicably, it means that being a more skilled player earns you the same rewards per unit of time than a lesser skilled one, given the ideal 50% win rate! It would make sense if progress in your track was affected by your MMR – but no. It doesn’t work that way.

Inb4 people start saying “Rewards in PvP don’t matter” or “Skilled players don’t care for rewards”. If rewards don’t matter, then Anet shouldn’t give them at all. No reward track progression, no gold…nothing. And while you’re at it, take out rewards for PvE as well. Just make everything free. Because…we’re all supposed to play just for the pure joy of it right? Right?

Bottom line: The reward structure for PvP is deeply messed up. If you’re going to have rewards, then they must be rationalized. Their very presence is enough of an indication that Anet feels “just the play” isn’t enough.

Now why don’t they do an actual good job of it and increase the rewards for those with higher MMRs?

until anet makes you weapons that all look like signs that say “I need this weapon to show other people I’m good because my skill alone isn’t enough” you can always try for that glorious hero armor set.

Also if you don’t enjoy playing pvp for the pvp then why the hell are you doing it. The truly competitive players don’t give two puppies about aesthetics, they know they are good and prove it in tourneys. You should try the same.

What is this “should” you’re talking about? Since when did the manner of playing a video game become a question of ethics? We all play for different reasons and Anet already knows this which is why we have rewards in the first place!

All I’m saying is that if they’re going to give rewards, those rewards must be rationalized and not the same for players of different skill levels.

Also, skins/titles/minis whatever are for showing outside a PvP arena as well. PvE players have their badges. We should too.

And stop trying to impose your value systems on others. I play for rewards. You are free not to. Both of us can exist side by side without interfering with each other, can we not?