Guild Wars 2

kdaddy.5431:

Ok so heres the thing.

In S1-3 i was placed on certain players teams. For instance i had a duo with OE and his partner and we had Five gauge. We lost but the game was insanely close. This resulted with a pip loss. Nothing to big, later that day OE and his duo partner were on my team again and 2 ESL players were on the other side. We won this game. Once again i only gained a pip.

The problem now for everyone is when im losing a match i should of won im getting -20. If i lose a match i had no chance to win in i lose -15.

When i win im getting +10- +18 and it only gets worse for players higher up and some have even posted in the low golds of winning and getting a +4 and losing a -20.

I dont know what they set the algorithm to. Anet will not show us open ratings in a match or even our team mates badge rating after a match.
…

The issue is as i stated in many threads is that your rank gain/loss is based on your mmr vs avg mmr of the enemy. It does not take into account mmr of your teammates. The system assumes it was a fair match of two teams of same rating – which obviously absolute nonsense given how every single match nowadays is extremely unbalanced due to various reasons (MM, low pop etc.). Preaching about it for 3 seasons now – getting effectively ignored anyway.

I (rating 1569) just got matched against A DUO OF #1 ranked player and the #6 ranked player (rating 1949 and 1826). Obviously my team of soloq nobodies around my rating got absolutely destroyed.

Just why?

Cynz.9437:

kdaddy.5431:

Ok so heres the thing.

In S1-3 i was placed on certain players teams. For instance i had a duo with OE and his partner and we had Five gauge. We lost but the game was insanely close. This resulted with a pip loss. Nothing to big, later that day OE and his duo partner were on my team again and 2 ESL players were on the other side. We won this game. Once again i only gained a pip.

The problem now for everyone is when im losing a match i should of won im getting -20. If i lose a match i had no chance to win in i lose -15.

When i win im getting +10- +18 and it only gets worse for players higher up and some have even posted in the low golds of winning and getting a +4 and losing a -20.

I dont know what they set the algorithm to. Anet will not show us open ratings in a match or even our team mates badge rating after a match.
…

The issue is as i stated in many threads is that your rank gain/loss is based on your mmr vs avg mmr of the enemy. It does not take into account mmr of your teammates. The system assumes it was a fair match of two teams of same rating – which obviously absolute nonsense given how every single match nowadays is extremely unbalanced due to various reasons (MM, low pop etc.). Preaching about it for 3 seasons now – getting effectively ignored anyway.

That is exactly why the match maker needs to put similar ranked players on the same team.

Cynz.9437:

kdaddy.5431:

Ok so heres the thing.

In S1-3 i was placed on certain players teams. For instance i had a duo with OE and his partner and we had Five gauge. We lost but the game was insanely close. This resulted with a pip loss. Nothing to big, later that day OE and his duo partner were on my team again and 2 ESL players were on the other side. We won this game. Once again i only gained a pip.

The problem now for everyone is when im losing a match i should of won im getting -20. If i lose a match i had no chance to win in i lose -15.

When i win im getting +10- +18 and it only gets worse for players higher up and some have even posted in the low golds of winning and getting a +4 and losing a -20.

I dont know what they set the algorithm to. Anet will not show us open ratings in a match or even our team mates badge rating after a match.
…

The issue is as i stated in many threads is that your rank gain/loss is based on your mmr vs avg mmr of the enemy. It does not take into account mmr of your teammates. The system assumes it was a fair match of two teams of same rating – which obviously absolute nonsense given how every single match nowadays is extremely unbalanced due to various reasons (MM, low pop etc.). Preaching about it for 3 seasons now – getting effectively ignored anyway.

Yes i was kind of against that thought process at first when S5 hit because i thought it would work itself out.

But i didnt take into account alot of things and i didnt look at it from the mid level players perspective getting stuck on the lower MMR team. Or high level players not playing each other and Anet putting the high level players on the same team.

So by the end of S5 i have turned my views on it and i agree with a majority of the things you have brought up on that topic.

Faux Play.6104:

Cynz.9437:

kdaddy.5431:

Ok so heres the thing.

In S1-3 i was placed on certain players teams. For instance i had a duo with OE and his partner and we had Five gauge. We lost but the game was insanely close. This resulted with a pip loss. Nothing to big, later that day OE and his duo partner were on my team again and 2 ESL players were on the other side. We won this game. Once again i only gained a pip.

The problem now for everyone is when im losing a match i should of won im getting -20. If i lose a match i had no chance to win in i lose -15.

When i win im getting +10- +18 and it only gets worse for players higher up and some have even posted in the low golds of winning and getting a +4 and losing a -20.

I dont know what they set the algorithm to. Anet will not show us open ratings in a match or even our team mates badge rating after a match.
…

The issue is as i stated in many threads is that your rank gain/loss is based on your mmr vs avg mmr of the enemy. It does not take into account mmr of your teammates. The system assumes it was a fair match of two teams of same rating – which obviously absolute nonsense given how every single match nowadays is extremely unbalanced due to various reasons (MM, low pop etc.). Preaching about it for 3 seasons now – getting effectively ignored anyway.

That is exactly why the match maker needs to put similar ranked players on the same team.

I think its just impossible since the player base if up and down. For instance i played 20 games in about 2 days, i went on a 3 game losing streak and every game we should of won but the people i was playing with did not understand basic rotations.

I lost my mind and havent logged in for about a week. This game for me is impossible to play at times even if the game play is fun.

I would like to know how are people disputing Acandis’ findings?

Stop, look at what he posted.
What the matchmaking should have done is split the two higher rated duos to face each other. Then use the 3rd duo/remaining players to round out to a relatively even rating.

In this case it failed and I think this is an issue with the code that needs to be reviewed.

This isn’t about what’s fair and to whom it’s fair. What rating win/loss they get. It’s about providing the most even matchup. In this case, one team was stacked with 4 nearly highest rated players, vs pugs and lower rated duo.

I do hope that this example is taken and the code is reviewed. All these people can be in the same matchup, this is ok, but the matchmaker should have after finding 10 people re-organized them to correct teams.

What I am worried is that the matchmaker found 10 people (which seems fine plats mostly), and then assigned them teams in order of queue entry: blue/red/blue/red etc.
That broke the team set up. Instead of taking 2 highest combined duo ratings, splitting to red/blue, add 3rd duo to lower rated, fill out remaining teams according to rating

in this case que is formed of:
3540 ; 2p
3440 ; 2p
3130 ; 2p
1710 ; 1p
1680 ; 1p
1630 ; 1p
1510 ; 1p

Then it’s easy from there to go blue red blue red by taking the above groups into teams based on rating, the lowest team gets the next player until they are even or above the opposing team
The ending result would be:
blue : 8440 (3540+1710+168+1510) 1 duo but higher average
red: 8200 (3440+3130+1630) 2 duos but lower average

The average rating would be less than 50pts difference.

That didn’t happen above, and the result was a roflstomp, a match not fun for any player. Winners would get bored easily, losers would be disheartened.

Poobah.6254:

I (rating 1569) just got matched against A DUO OF #1 ranked player and the #6 ranked player (rating 1949 and 1826). Obviously my team of soloq nobodies around my rating got absolutely destroyed.

Just why?

Sounds like just another day in PvP, eh?

It’s almost not even worth trying any more. I’m either getting games where my team mates are so bad that I mechanically can’t support their decision to run in circles, or the enemy is so bad that all I have to do is spawn camp the guys running in circles. Ranked is pointless without tight matching.

Guess it’s time to ditch it and run with tourneys.

Matches like this happen all the time. Its a joke really. You are only as strong as the weakest link on your squad. How are two 1500 MMR people going to win fights against multiple people 250 MMR + higher then them. Its impossible.

steelheart.7386:

Matches like this happen all the time. Its a joke really. You are only as strong as the weakest link on your squad. How are two 1500 MMR people going to win fights against multiple people 250 MMR + higher then them. Its impossible.

Keep in mind 1500 is just outside the top 250. When you need 10 players to make a match there aren’t always going to be enough high level people available.

Perhaps daily/monthly tourneys should be used for rating once your MMR gets high enough. I would revamp them to be more like the would cup format where you have an initial round robin with groups of 4 or 5 teams before the bracket starts. That way you are guaranteed at least 3 or 4 matches. the advantage of this is the high rated players would get a lot more for wins, and it would eliminate match manipulation.

steelheart.7386:

Matches like this happen all the time. Its a joke really. You are only as strong as the weakest link on your squad. How are two 1500 MMR people going to win fights against multiple people 250 MMR + higher then them. Its impossible.

Based on rank gain/loss system it is possible otherwise you wouldn’t lose ranks worth of 2-3 wins for such matches. I really would love explanation for devs how they even came to such idea.

How many games all of them played ?
I am trying to find Track atm on the leaderboards (need 60 games played to be shown) and i cannot find him

Any chance in Season 8 to have more games needed per week ?
Lets say 30 per week (instead of 15) , so MMR works faster and better ?

Because of what happens currently, all the titles and leaderboard lost any meaning to me.
Those players that are on top are skilled, there is no discussion on this subject. But as long as they don’t face a fair challange to keep on testing them, the community won’t be easily breathtaken by their rank. As I see it on my own example, I am nowhere near to be amazed by any of the top players, if they use tricks to keep their status or, as some people already mentioned, they get stacked with each other to create a wall between high plat and high legend. I’ve thought they are super amazing and I have to git gud. I still have lots of things to learn and I don’t think that with my current internet connection and weak rig, I’ll be able to reach legendary. But still, with all I got I can safely stay on mid platinum level, which is good enough for now.

I’ll give you this. As much as I am not fond of power creep, at this point it might be the only thing to save the MM system. In theory, if new elite specs turn out to be strong (high chance) they will bring back a lot of old players as well as some freshies as well. The point is, with two strong elite specs that cannot be mixed together, we should have a forced meta of at least 2 good builds per class. Thanks to that, people will have more to choose from and if they like it, there will be more of them to play it, thus give a chance for PvP to restore itself.
It’s like a bowl (PvP mode) and cereals (current pvp playerbase). With new expansion milk will pour (new or returning players) and it all should be a quite good breakfast (beginning of pvp in the new expansion) for our matchmaking system. Hope there won’t be too much sugar in it (over power creep of the new elite specs)

Sounds like the green leaf strikes again.

I really would like an explanation for this match up

Imho either Anet lies about its matchmaking: many – including myself – have the opinion that certain games are meant to end in a certain way. A.k.a. – they are fixed or heavily biased so that the lower rated part of the community gets more wins via getting thrown into teams that win the game for them.

Or the matchmaking still cant measure player skill. The potential of individual players varies wildly, to the extent that I cant even comprehend how they get into the same match.

This season I’ve won all the close games (within 100 score difference) except for 2. All my other losses are from total blowouts. Its still 75% luck and maybe 25% skill or even less. Whether these are forced wins/losses or just bad luck I dunno.

Either way those games where everything is decided before the teams leave their bases should be much lower. 27 of my 30 losses this season were all games we had no chance winning.

Not fun, not competitive… and each season there are multiple topics about it on the forum, yet nothing changes.

I’d also like to add this screenshot which shows that losing a total blowout match and a close game results in the same loss in rating. So according to Anet those teams were in both cases close to eachother in skill level however the actual score of the match says its the other way around. (Losing a game 172-503 and 424-503 both resulted in a 13 rating loss – which tells me that MMR says NOTHING about skill level.)

consider this:

if matchmaking is so bad, would adding the ability to form 5 man premades back into ranked actually be detrimental?

like, it gives you the option of having a good team that you can reliably win matches with. sure it’s at the expense of ‘fair matchmaking’, but we don’t even have this anyway. so do we lose anything at all?

choovanski.5462:

consider this:

if matchmaking is so bad, would adding the ability to form 5 man premades back into ranked actually be detrimental?

like, it gives you the option of having a good team that you can reliably win matches with. sure it’s at the expense of ‘fair matchmaking’, but we don’t even have this anyway. so do we lose anything at all?

So allow people to get carried to ratings where they don’t belong and they ruin matches for other people like in previous seasons?

Or having 5 man premade farming full solo teams all day thus making people quit?

What an idea….

Cynz.9437:

choovanski.5462:

consider this:

if matchmaking is so bad, would adding the ability to form 5 man premades back into ranked actually be detrimental?

like, it gives you the option of having a good team that you can reliably win matches with. sure it’s at the expense of ‘fair matchmaking’, but we don’t even have this anyway. so do we lose anything at all?

So allow people to get carried to ratings where they don’t belong and they ruin matches for other people like in previous seasons?

Or having 5 man premade farming full solo teams all day thus making people quit?

What an idea….

the idea is in jest my dude

but honestly, how much more ruined can matchmaking get?

There’s only one thing left to they to do: solo queue only.

Its not broken, the player pool is atrocious that’s the problem. According to what I’ve seen in gold I could imagine the pool of available player for people in legend is never and ever above a 100 and am being generous here. The longer your queues goes the worst the match will be, since the restriction goes down, after 3min in queue the restriction for multiple of the same profession in the same team is removed. I would guess a same thing could apply for skill rating, but I can’t get proof of that.

yanniell.1236:

There’s only one thing left to they to do: solo queue only.

While i love to pair with someone (fried or new player that might become a friend …) and that social aspect should not be ignored, this will solve a bunch of problems ….

So i think they need multiple Q´s and game modes. While 2:2 might be to much change from balance focues, 3:3, 4:4 and 5:5 modes might be possible.

Wolfric.9380:

yanniell.1236:

There’s only one thing left to they to do: solo queue only.

While i love to pair with someone (fried or new player that might become a friend …) and that social aspect should not be ignored, this will solve a bunch of problems ….

So i think they need multiple Q´s and game modes. While 2:2 might be to much change from balance focues, 3:3, 4:4 and 5:5 modes might be possible.

GW2 doesn’t have the player base to support so many queues. Best to have 5 man be automated tournaments, 2v2 and solo ranked only or solo only once you reach platinum/legend as others said 2 seasons ago.

Can we stop saying the matchmaker is wrong. It picked the people very well! ~250 difference at most is not that bad because of player population.

The issue is how it placed the 7 players (3 2p groups and 4 1p players) into groups. It could have placed them with an average MMR of less than 50, but it did it in such a way that they are much higher than that.

Therefore creating lopsided matches. What we need to understand is whether there is any logic to balance out a match after MM pulls in players, or is it first come first serve basis.

Trevor Boyer.6524:

This is wonderful thread. I noticed these same sort of things a long time ago but that was before we had a relevant rating system to compare/contrast different matches. Everyone told me to “l2p” and ignored what I was pointing out to them. Glad to see others are finally noticing this.

Long story short ~ Over the years I’ve also noticed that some accounts just get better matches much more often than others do. I’ll refrain from writing a text wall of theories but I will say that after 5 years and 10,000 matches of paying attention to this, it isn’t just good luck or a coincidence for these players. No, it’s rigged patterns that break the normal patterns of the algorithm’s match making. I’ve been certain of this for almost three years now. Is it simply meta smurfing tactics or is there something going on that arenanet isn’t telling us about? At this point, I believe it’s a little bit of both.

A few vague examples as to leave out names of actual players:

• Couple guys I know personally, repeatedly get the easiest matches I’ve ever seen. Even if I am not duo with them but land on their team in a ranked que, we get such easy matches that it’s boring. And let me tell you, they are good players! but they aren’t doing anything special to carry matches. They just get stacked with other good players, almost every single time that they que.
• Some other guys that most people know, I notice that whether I am with or against them, their team always wins. They are great players! but I notice the majority of these wins are not from these guys carrying so hard, they are because their team is simply stacked with veteran player base that most people know, almost every time they que.
• New players that arrive fresh out of pve, hoping to grind those wings. I see a lot of these somehow tag easy match after easy match and ride straight into plat with some ridiculously high win rate. This is very odd considering their visible lack of experience and ineptitude in combat. I’ve often wondered if the algorithm has implemented a function to give new guys on the scene easy matches for awhile so they maintain interest. Regardless of if that’s true, whatever is going on here is not only unfair for veteran participants but it’s insulting.
• Other players of various experience levels that I refuse to que with in ranked. I refuse to que ranked with them because I swear to god, I get the worst matches I have ever seen, every time I que with them. I’m not talking situations where you play with someone for the first time, get a 4 game lose streak and then dodge them forever. No, I’m talking people that I’ve played with for years. Some of which are strong players but they almost always, land on the bad side of a team in a ridiculously impossible to win match. I’m not sure if they were infracted too many times and kittened off the wrong person in the anet office or if too many people within the match manipulation community do not like them. Either way, they almost always get scum of the earth complete garbage matches, and I’ve watched it happening to them for years.

I remember sometime ago me and my duo partner carried 3 players who full wiped at mid in 10 seconds by taking sides me 1v2 my friend 1v3 i killed both players then proceded to decap mid etc. The other 3 players r just kittened and we still carried them 500-126, this is what mass carrying means.

I can carry bad teams on a necro even though I’m not very good. If you learn how to kite on a necro and make the other team chase you around your team is free to kill them.

After reading the matchmaking process Anet uses, my idea above is clearly never in play.

The problem may be the “rating padding”. In a high pop environment its impact is minimal, but in GW2 this is creating a large number of unbalanced matches. To the point that most of your games are blowouts because some players were padded too much just to get them into a match.

That being said padding has to be there because of their “if no one around your mmr, go back to q”. That has to be abandoned in a sense. Then padding may not be necessary?

I’m hoping people make a legit conversation with ideas rather than the constant QQ.

I’ve went through a little frustrating streak. It’s cool, happens, not the first and not the last time. But what I want to know, is there really some sort of a "luck" factor to account? I can’t quite tell but I’ll tell you the facts right now.
So in the morning I went for ranked to get to 250 (3-4 matches and done) but after two won matches I got a troll who made us lost two games. Well, okay... happens, I took a break then to do stuff. I returned in the evening to do some games too, and I’ve ended up with 6-7 lost matches in about more than 300 point gap. In the last 2 or 3 matches I was so unpleased that my fighting skills were kind of affected but that doesn’t matter in the long run. Meanwhile, there is a nice dude in my PvP guild who barely has 1k achievs, got a guardian build from someone who was 13 in rank and is killing it all over the place. However, we both have similar percentage to win. Between lost and won matches he has about 10-11 matches after a win streak, I also tend to have it. Yet he is somewhere in 120 rank while I just fell out of plat T2. How to understand this phenomenon?