Plat 2, 3, Leg Players should not be in lower tier games at all; fun for no one.

[Crab Fear.1624]

When these players make it into average games they can 1 v 3 or 4.

They have a different understanding of the conquest than an average players concept, and the cohesion(lack of) is apparent.

If you are the high level player and you are stuck with the low level players not doing what you expect, you will be frustrated at an almost inevitable loss.

If you are the low level player, you may mistake the high level game play for being a broken spec.

Or, you may believe PvP is just not for you if all you are good for is respawn.

Being new, you might feel inclined to make your way back to the pack, but if the pack is wiped out by one dude....you run to a uncontested point, but 1 v 1 may not be your thing, and of course...wiped.

The experience of pvp is RUINED by allowing high level players to leach into lower tier games.

The lower tiers need to battle amongst themselves, and find new champions though trialing experiences, and not disheartening experiences like the one detailed above.

Higher tier players may get less games, but they need to be pitted against players of equal skill to validate their claims of greatness.

I have a hunch that many players, especially at some higher levels, ahve a good idea of who is a threat, when the least amount of dangerous competition is on, and they play those hours with minimal game to reduce the risk of rating loss (not all players do this, but..)

If you can only play against players close to your rating (especially above 1600), it might make for more fun games.

It is worsened when the pair of duo gods gets thrown in against average players.

I don't think duo q is a problem.

I believe it is the flexibility of the match maker.

The search widens to much.

Deadlock players above 1600 rating so they can only face opponents of 1500+ rating.

It will make for better games, as rating camping will be more difficult and should get more experienced players more active.

Should not see a top 10 player in a gold game.

[mrauls.6519]

If only the population was bigger. Honestly, our best bet is praying for better class balance. I'm surprised we haven't seen another balance patch yet.

[Trevor Boyer.6524]

They need to tighten the margins of match possibility for sure.

I'd rather wait 6-7 minutes for a match where all of my teammates and opponents were within 150 rating of each other, than wait 4 minutes for a match when I'm seeing people ranging from 1300 to 1800, and then I have to wait another 10 minutes for that lopsided garbage to end.

I'd also like to point out that the following:

@"Crab Fear.1624" said:

If you are the low level player, you may mistake the high level game play for being a broken spec.

Or, you may believe PvP is just not for you if all you are good for is respawn.

Might actually be the single most important thing that Arenanet should be considering right now, in terms of drawing in and keeping player base.

In the last couple of seasons I've lost a lot of GW2 pvp buddies, who had grown quite tired of being in matches with 1700+ duos against them, when they were already struggling enough against players of their own division.

People like to feel "as if they played well or contributed adequately" and that feeling is what maintains interest in the game mode. But when players are always being drop-kicked back into the respawn each and every time they poke their head into combat for 2s, and everyone on their team is telling them how bad they are and how little they contributed, people bow out and don't come back to the game mode because it isn't fun.

[Turkeyspit.3965]

And what is to stop them from creating an alt account?

I like that GW2's combat system allows for skill to shine forth, rather than it being mostly about gear/stats like other MMOs, but at the same time, the amount of impact a single player can have, even in as small a setting as 5v5, is an issue. Can it be fixed? Nope. Double edged sword.

For what it's worth, it is one of the reasons I quit PvP after giving it a season. I'm not a great PvPer, though I did get better over the 120+ games I played, and towards the end I found myself more successful in 1v1 and being better aware of when to peel off / rotate, etc, but in the end I realized that my gameplay was average, but consistent, which meant the deciding factor was the randomness of who I was grouped with vs. who I was facing.

I had roughly a 54% win ratio, but most of those games were 70-500 one way, or the other. If I got a lucky queue, we stomped, and if not, we got stomped, all the while I was playing the same as I always was. Out of 10 games, maybe 2 of them were a 400-500 score, win or lose.

Losing 70-500 isn't fun, nor is winning by those same ratios, and I just reached the conclusion the game mode wasn't for me.

If they ever do 5v5 ranked queues where people can form teams of friends, and there are enough teams for decemt queues that also keep the high end teams from facing off against the low end teams, maybe then ill give it another look.

[Trevor Boyer.6524]

@Turkeyspit.3965 said:And what is to stop them from creating an alt account?

And that single statement speaks worlds about what is wrong with the ranked game mode.

[Crab Fear.1624]

Never mind it was mean.

[Stallic.2397]

Yeah you noobs, stop ruining my game play...

Lol, in all seriousness, there should be greater consideration when combining players from different leagues. Gold and plat is a big difference in skill

Its not unkind either to think so. How else can you gain skill if the game doesn't match you with players that are at least the same level as you?

If the game focuses on creating quick matches rather than quality matches, then ranked becomes a question of how well you can carry rather than how well you can play your role.

[megilandil.7506]

@Crab Fear.1624 said:When these players make it into average games they can 1 v 3 or 4.

They have a different understanding of the conquest than an average players concept, and the cohesion(lack of) is apparent.

If you are the high level player and you are stuck with the low level players not doing what you expect, you will be frustrated at an almost inevitable loss.

If you are the low level player, you may mistake the high level game play for being a broken spec.

Or, you may believe PvP is just not for you if all you are good for is respawn.

Being new, you might feel inclined to make your way back to the pack, but if the pack is wiped out by one dude....you run to a uncontested point, but 1 v 1 may not be your thing, and of course...wiped.

The experience of pvp is RUINED by allowing high level players to leach into lower tier games.

The lower tiers need to battle amongst themselves, and find new champions though trialing experiences, and not disheartening experiences like the one detailed above.

Higher tier players may get less games, but they need to be pitted against players of equal skill to validate their claims of greatness.

I have a hunch that many players, especially at some higher levels, ahve a good idea of who is a threat, when the least amount of dangerous competition is on, and they play those hours with minimal game to reduce the risk of rating loss (not all players do this, but..)

If you can only play against players close to your rating (especially above 1600), it might make for more fun games.

It is worsened when the pair of duo gods gets thrown in against average players.

I don't think duo q is a problem.

I believe it is the flexibility of the match maker.

The search widens to much.

Deadlock players above 1600 rating so they can only face opponents of 1500+ rating.

It will make for better games, as rating camping will be more difficult and should get more experienced players more active.

Should not see a top 10 player in a gold game.

as a poor silver-gold pvX player i think you cant not be more correct in your statements some tier barriers have to be hard barriers to cross by matchmaker but thats not entirely anet faults its also players fault aka alting to carry some friends or just trolling in low tiers, only imagine for a moment what kind of sport will be football/soccer if messi or cr7 or whatever player of firts line divisions could pop in a regional team wearing a fake moustache

[rng.1024]

At this point, looking at the rating spread of the leaderboard, we might be better off readjusting the volatility of rating gains/losses to adjust for dwindling population.

What this does is effectively making it alot harder to reach todays top ratings, and easier to reach the average rating than it currently is.

You can think of it as removing bronze and legendary division. Why is this helpful?

The algorithm works by starting out at your rating, and expands from there. We already know the player allocation by division isn't linear (hence filtering by skill rather than luck) and we're looking at a curve with least players at the top and most at the bottom. Right now the ratings range from 0 - 2100, so with high rating comes a smaller pool of teammates and opponents to draw from.

By reducing rating volatility, we would be looking at f.ex a 901 - 1799. All players currently below 901 would go up, and all players above 1799 would go down. That sounds bad, right? Thing is it would only affect the minority of the population - those below 901 would have to play with and against better players, and those above 1799 would have to play more games not against eachother - but everyone would now have a bigger pool of teammates/opponents to draw from within the search range of the algorithm.

In the short term it would put low-skilled players in better matches, and high-skilled players in worse ones - but how else are we going to improve as a community? Not only would this lower queue times for everyone, but it would also create more balanced matches without drastic measures other than changing 1 value in the algorithm.

[RisenHowl.2419]

@rng.1024 said:At this point, looking at the rating spread of the leaderboard, we might be better off readjusting the volatility of rating gains/losses to adjust for dwindling population.

What this does is effectively making it alot harder to reach todays top ratings, and easier to reach the average rating than it currently is.

You can think of it as removing bronze and legendary division. Why is this helpful?

The algorithm works by starting out at your rating, and expands from there. We already know the player allocation by division isn't linear (hence filtering by skill rather than luck) and we're looking at a curve with least players at the top and most at the bottom. Right now the ratings range from 0 - 2100, so with high rating comes a smaller pool of teammates and opponents to draw from.

By reducing rating volatility, we would be looking at f.ex a 901 - 1799. All players currently below 901 would go up, and all players above 1799 would go down. That sounds bad, right? Thing is it would only affect the minority of the population - those below 901 would have to play with and against better players, and those above 1799 would have to play more games not against eachother - but everyone would now have a bigger pool of teammates/opponents to draw from within the search range of the algorithm.

In the short term it would put low-skilled players in better matches, and high-skilled players in worse ones - but how else are we going to improve as a community? Not only would this lower queue times for everyone, but it would also create more balanced matches without drastic measures other than changing 1 value in the algorithm.

This is a great idea

[Bazsi.2734]

@rng.1024 said:At this point, looking at the rating spread of the leaderboard, we might be better off readjusting the volatility of rating gains/losses to adjust for dwindling population.

What this does is effectively making it alot harder to reach todays top ratings, and easier to reach the average rating than it currently is.

You can think of it as removing bronze and legendary division. Why is this helpful?

The algorithm works by starting out at your rating, and expands from there. We already know the player allocation by division isn't linear (hence filtering by skill rather than luck) and we're looking at a curve with least players at the top and most at the bottom. Right now the ratings range from 0 - 2100, so with high rating comes a smaller pool of teammates and opponents to draw from.

By reducing rating volatility, we would be looking at f.ex a 901 - 1799. All players currently below 901 would go up, and all players above 1799 would go down. That sounds bad, right? Thing is it would only affect the minority of the population - those below 901 would have to play with and against better players, and those above 1799 would have to play more games not against eachother - but everyone would now have a bigger pool of teammates/opponents to draw from within the search range of the algorithm.

In the short term it would put low-skilled players in better matches, and high-skilled players in worse ones - but how else are we going to improve as a community? Not only would this lower queue times for everyone, but it would also create more balanced matches without drastic measures other than changing 1 value in the algorithm.

Maybe I'm dumb, but how exactly would this result in better matchmaking? Sometimes I get matches where one team has a legendary(or high plat) premade, but with the population dwindling, matchmaking had to resort to bringing gold2/3 players into the match to make teams. It doesn't matter how you change personal ratings, the difference in skill remains. Those high end premades will roll over the average players on the enemy team even if you make them only 50 rating apart on the leaderboard.

[Dantheman.3589]

Most ppl don’t seem to realize or maybe have never done it, but a duo of 2 plat3-legend players, unless a bunch on, is like a 7-8 min q time ime some times up to like 13 mins. Nothing can we do to fix it. Ppl say lower margin it’s ole to have 7 mins qs- well quess we always have those already and should we make that 30 mins instead? Think about it ppl will say game is dead

[DoritoGod.5817]

id wait up to 15 minutes if it meant every q i did wasnt a rage inducing garbage game with silvers doing beasts and standing still on node

[Leonidrex.5649]

@DoritoGod.5817 said:id wait up to 15 minutes if it meant every q i did wasnt a rage inducing garbage game with silvers doing beasts and standing still on node

its not how it works sunshine, if you are legend/p3 and wanna ply with legend/p3 only you would NEVER find a game, making matchmaking look for lower ranks is lesser of 2 evils, unless you wanna sit 5hours in queue waiting for one of only 50 people to log so you can get in the game, if they not online unlucky no pvp for you

[Stand The Wall.6987]

I think this is an effect of population, balance, and how anets glicko system rates ppl. it doesn't account for cheating and it wasn't made for solo calculations based on team effort.

[rng.1024]

@Bazsi.2734 said:

@"rng.1024" said:At this point, looking at the rating spread of the leaderboard, we might be better off readjusting the volatility of rating gains/losses to adjust for dwindling population.

What this does is effectively making it alot harder to reach todays top ratings, and easier to reach the average rating than it currently is.

You can think of it as removing bronze and legendary division. Why is this helpful?

The algorithm works by starting out at your rating, and expands from there. We already know the player allocation by division isn't linear (hence filtering by skill rather than luck) and we're looking at a curve with least players at the top and most at the bottom. Right now the ratings range from 0 - 2100, so with high rating comes a smaller pool of teammates and opponents to draw from.

By reducing rating volatility, we would be looking at f.ex a 901 - 1799. All players currently below 901 would go up, and all players above 1799 would go down. That sounds bad, right? Thing is it would only affect the minority of the population - those below 901 would have to play with and against better players, and those above 1799 would have to play more games not against eachother - but everyone would now have a bigger pool of teammates/opponents to draw from within the search range of the algorithm.

In the short term it would put low-skilled players in better matches, and high-skilled players in worse ones - but how else are we going to improve as a community? Not only would this lower queue times for everyone, but it would also create more balanced matches without drastic measures other than changing 1 value in the algorithm.

Maybe I'm dumb, but how exactly would this result in better matchmaking? Sometimes I get matches where one team has a legendary(or high plat) premade, but with the population dwindling, matchmaking had to resort to bringing gold2/3 players into the match to make teams. It doesn't matter how you change personal ratings, the difference in skill remains. Those high end premades will roll over the average players on the enemy team even if you make them only 50 rating apart on the leaderboard.

By premade do you mean a duo? The change I was suggesting isn't going to affect that in the least, that's a whole other discussion to tackle.

And you are completely right, the skill gap would still be there - only now the matchmaker wouldn't have to dip as low (like you mentioned f.ex gold 2) in order to find players for both teams. It would find enough already f.ex in plat 1.

It seems counterintuitive, but since both teams are affected (getting lower and higher rated players mixed in) it will create more possible combination within team average ratings because everyone (from bronze to legend) are now closer in rating.

It's all about the interval the matchmaker is willing to use before the 5 minute mark of the queue. Right now, with too few players close to your rank online, this interval is reached often before enough players are found. Essentially Id like to squeeze more players into that interval so that there's a higher chance of you getting a match before it has to look beyond that limit and create really unbalanced matches for the sake of just getting you in a game.

My point is you would be less likely to have 1 legend and 4 golds on your team, and more likely to get a plat 3, a plat 2 and 3 plat 1's. This would also go for the opposite team, so neither get the huge outliers we see today.

As for "high end premades", the chances of both teams getting them instead of only one side, would be significantly higher than it is right now as an effect of this.

[bethekey.8314]

I suggested something along these lines about a year ago:

https://en-forum.guildwars2.com/discussion/66261/opt-in-matchmaking-to-improve-the-quality-of-pvp

I don't know what the best implementation of it would be, but I do feel something could be done to help this problem.

[JTGuevara.9018]

I'm baaccck! Y'all miss me?

Yeah...despite the "absence", my mind has not changed whatsoever about this matchmaking system. It is easily exploitable.

Now, on to the topic at hand. Plat 2, 3 and legends in lower tiers? Well that is called smurfing! High-ranked players do it quite often! It is done in this game as it's done in other genres like FPS, MOBA, etc. Is it really a surprise to anyone that has had experience with this matchmaking? ELO-based systems are easily exploitable when done in team settings or anything else other than what was intended for them, 1v1s.

[Luthan.5236]

I would not mind bigger differences ... as long as the overall team average rating was balanced. (1000 and 2000 player vs 1500 an 1500 would be okay for example.) The big problem is that - afaik - rating change is calculaged by own rating vs. enemy team average. In these cases people above their average would always lose a lot in case of loss. So it is better to have teams where within the team the ratings are close to each other.

What is the biggest confirmed rating difference within teams that occurs regularly nowadays? If platin can't be matched up with platin the matchmaker looks for gold 2-3 I guesss? But surely there is enough population in gold so he doesn't need to go any lower? (Edit: With how calculating the rating gain/loss works now it seems to be a problem for Australian players ... EU it is afternoon at their prime time I think and US still early morning. The example from Noodica posted in the post below me seems troublesome and I guess it happens regularly for Australian players. With the rating system fixed so it uses own team average vs. enemy team average to calculate the score change ... it might be less of a problem. Nothing can change the low population at certain times though - which might be a bigger problem for Australian players.)

I'd prefer if ratings for each player where shown at the end (when it reveals who duo queued). Less speculating. Whe had 1 guy complaining about a bad balance match lost 500 vs 50 or so when he posted in the forums about bad matchmakter and a dev said according to ratings it was balanced. (They must just have played badly.)

[Noodica.5428]

As somebody who plays plat 2 average, but also plays at plat 3 I couldn't agree with this post more.

Early this season, whilst I was in mid plat 3 I was still being matched with like gold 1 players. It doesn't help that the (australian) time zone I play in makes matchmaking bad as it is.. Now that I'm in Plat 2 (from losing like 50 rating a loss yet only gaining like +5 for a win) I still get matched with and against silver players and it is beyond understandable. The ranking system should exist for a reason and it is so frustrating losing insane amounts of rating simply for being matched with players who are new to the game mode and have no clue of what to do or how rotation works.

Fix this.

[Trevor Boyer.6524]

@"bethekey.8314" said:I suggested something along these lines about a year ago:

https://en-forum.guildwars2.com/discussion/66261/opt-in-matchmaking-to-improve-the-quality-of-pvp

I don't know what the best implementation of it would be, but I do feel something could be done to help this problem.

I just noticed & read that thread. I have no idea why that thread did not get more attention than it did. That might actually be the #1 suggestion that has been dropped in this pvp subforum, ever.

I dropped my 2 cents into it for debate. I said that:

My idea for an opt in would be along the lines of allowing players to choose their own queue wait times.

1. So if a player wants quick queue but more volatile match making "As in 1300s in games with 1700s" select 3-5 minute queue time.
2. If a player wants more reasonable match making select 5-7 minute queue time.
3. If a player is willing to wait for quality matches, they can select a 7-10 minute queue time, or maybe even a 15 minute setting.

I mean this could very seriously work and it would be easy to incorporate into the UI. A player pushes "ranked queue" and then a drop down box opens up with 5 or 6 selectable time formats, asking them how long they are willing to wait.

And I think Arenanet would be truly surprised at how many people would be willing to wait upwards of 10+ minutes for better quality match making.

[Crab Fear.1624]

I know some believe averages in mmr rating for the team "make up" for it.

But here is my thought:

Team A (2000, 1000, 1000, 1000, 1000)

Team B (1200, 1200, 1200, 1200, 1200)

Team A will likely win especially if the 2000 rated player is on an s tier class.

The 2000 rated player can take just about any match up in that game.

The difference in player skill for 1200 and 2000 is galactic.

1800 rated is still lightyears ahead of silver.

1700 would require cryogenic preservation.

I have been in games against BotBs and they stall 3-4 of my team somewhere off node.

Games against a god and demi, they kept everyone in respawn.

Played a game with them and we 500 - 0 the other team

That may be fun for winners, idk.

Is it fun for the losers?

[Exedore.6320]

@Crab Fear.1624 said:But here is my thought:

Team A (2000, 1000, 1000, 1000, 1000)

Team B (1200, 1200, 1200, 1200, 1200)

This team setup can't happen unless these are the only people playing PvP in the last 15min.

[bethekey.8314]

@Trevor Boyer.6524 said:

@"bethekey.8314" said:I suggested something along these lines about a year ago:

https://en-forum.guildwars2.com/discussion/66261/opt-in-matchmaking-to-improve-the-quality-of-pvp

I don't know what the best implementation of it would be, but I do feel something could be done to help this problem.

I just noticed & read that thread. I have no idea why that thread did not get more attention than it did. That might actually be the #1 suggestion that has been dropped in this pvp subforum, ever.

I dropped my 2 cents into it for debate. I said that:

My idea for an opt in would be along the lines of allowing players to choose their own queue wait times.

1. So if a player wants quick queue but more volatile match making "As in 1300s in games with 1700s" select 3-5 minute queue time.
2. If a player wants more reasonable match making select 5-7 minute queue time.
3. If a player is willing to wait for quality matches, they can select a 7-10 minute queue time, or maybe even a 15 minute setting.

I mean this could very seriously work and it would be easy to incorporate into the UI. A player pushes "ranked queue" and then a drop down box opens up with 5 or 6 selectable time formats, asking them how long they are willing to wait.

And I think Arenanet would be truly surprised at how many people would be willing to wait upwards of 10+ minutes for better quality match making.

Thanks for the kind words. In my opinion, poor match quality is the root of many problems in Guild Wars 2 PVP (e.g. toxicity, build diversity, and vastly different views on balance) and I hope it gets addressed somehow.

[DoritoGod.5817]

@"Crab Fear.1624" said:I know some believe averages in mmr rating for the team "make up" for it.

But here is my thought:

Team A (2000, 1000, 1000, 1000, 1000)

Team B (1200, 1200, 1200, 1200, 1200)

Team A will likely win especially if the 2000 rated player is on an s tier class.

The 2000 rated player can take just about any match up in that game.

The difference in player skill for 1200 and 2000 is galactic.

1800 rated is still lightyears ahead of silver.

1700 would require cryogenic preservation.

I have been in games against BotBs and they stall 3-4 of my team somewhere off node.

Games against a god and demi, they kept everyone in respawn.

Played a game with them and we 500 - 0 the other team

That may be fun for winners, idk.

Is it fun for the losers?

this isnt true, in fact its the opposite in this meta. here's what happens: 2k rating player decides to just dominate a side node. he pushes far and wins his 1v1 in around 20 seconds, his team loses mid bc teammates are worse. 2k player now gets 1v3'd and easily lives even downing a player but cant finish bc cleave is too low and theres no safe stomps. eventually one leaves and he's left 1v2, which he holds node decap while easily living. his teammates fed off spawn and maybe got home capped but then one monkey went for beast and instantly their team lost 3v3. repeat the above steps until the game is lost 400-500. 2k player never made a single mistake, never died, won every single fight he was in, and still lost bc his teammates god farmed by slightly better players.