An example of how low the population is - Ranked Matchmaking — Guild Wars 2 Forums
Home PVP

An example of how low the population is - Ranked Matchmaking

Broady.2358Broady.2358 Member ✭✭✭

I feel this one picture sums up how low the PvP population is right now ----------> https://imgur.com/cunZI2s

For context I was playing on a Monday at around 3pm. I am currently Gold 3 (as seen in the picture) and in DuoQ with my friend who is in Plat 1.
Matchmaking then pairs up against Zan from Rank 55 Dragons who I believe is P3/Legendary and puts Tenebrae from French Worms on my team. The population is so low that I, as a gold 3 player even with a plat 1 duo partner are in the same game as these top tier players. I understand this is happening because of a low population but surely this should be alarming to developers to see this kind of disparity in skill level in their RANKED game mode.

Its frustrating for both parties. I go into a 1v1 scenario, expecting it to be a fairly even fight as we both should be similar rating but instead I could end up completely destroying them OR they completely destroy me. In a gamemode that has all the intentions of promoting skill level through different ranks, the games themselves discredit this.

Comments

  • Asum.4960Asum.4960 Member ✭✭✭✭

    Were they duo queing?
    Something I noticed a lot of top players do is duo que together, one on their Legendary rated main account and the other on a low rank account to balance out their rating so they can stomp way lower skilled G2-P2 players together, swapping back and forth accounts to keep their top spots/get insane winrates.

    Regardless of that though just bad matchmaking happens a lot too and it is a problem. Besides disabling duo que for P+ again, which I would very much so like to see, to fix that mentioned aspect at least, I'm not sure what Anet can do about general population issues.

    R.I.P. Build Templates, 15.10.2019

  • Asum.4960Asum.4960 Member ✭✭✭✭
    edited June 22, 2020

    @Aktium.9506 said:

    @Broady.2358 said:
    I, as a gold 3 player even with a plat 1 duo partner are in the same game as these top tier players.

    Your rating isn't even taken in consideration by the matchmaker if you duo queue with someone with a higher rating.

    It uses the highest player's rating in the party for matchmaking purposes.

    Are you sure about that? Then the population/matchmaking must be a lot more dire than I thought with frequently seeing literally 2 of the Top 5 players in a team vs entire teams with not a single person on the leaderboard. Seems odd.

    R.I.P. Build Templates, 15.10.2019

  • Aktium.9506Aktium.9506 Member ✭✭✭

    @Asum.4960 said:

    @Aktium.9506 said:

    @Broady.2358 said:
    I, as a gold 3 player even with a plat 1 duo partner are in the same game as these top tier players.

    Your rating isn't even taken in consideration by the matchmaker if you duo queue with someone with a higher rating.

    It uses the highest player's rating in the party for matchmaking purposes.

    Are you sure about that? Then the population/matchmaking must be a lot more dire than I thought with frequently seeing literally 2 of the Top 5 players in a team vs entire teams with not a single person on the leaderboard. Seems odd.

    Yes. Otherwise you could just have a buddy tank an alt as low as he can and then queue with him on your main to farm free wins against golds.

  • Asum.4960Asum.4960 Member ✭✭✭✭
    edited June 22, 2020

    @Aktium.9506 said:

    @Asum.4960 said:

    @Aktium.9506 said:

    @Broady.2358 said:
    I, as a gold 3 player even with a plat 1 duo partner are in the same game as these top tier players.

    Your rating isn't even taken in consideration by the matchmaker if you duo queue with someone with a higher rating.

    It uses the highest player's rating in the party for matchmaking purposes.

    Are you sure about that? Then the population/matchmaking must be a lot more dire than I thought with frequently seeing literally 2 of the Top 5 players in a team vs entire teams with not a single person on the leaderboard. Seems odd.

    Yes. Otherwise you could just have a buddy tank an alt as low as he can and then queue with him on your main to farm free wins against golds.

    Yes, that's pretty much what I thought was happening as only reasonable explanation for why the matchmaking is so bad.
    Most of the time when I face Top 10 ranked players they are duo queuing with some account I've never seen before or can't find on the leaderboard (while playing equally well), which eventually made me suspicious, thinking that's what is happening.

    I expect to face them when I reach P3, but not still after dropping as low as P1, especially since I generally don't seem to have a balancing top player on my side. Oh well.
    Rating differences of ~300 or more between players seems a bit much unless the system is gamed somehow, unless the population/matchmaking is really just that bad.

    R.I.P. Build Templates, 15.10.2019

  • Aktium.9506Aktium.9506 Member ✭✭✭
    edited June 22, 2020

    @Asum.4960 said:
    Yes, that's pretty much what I thought was happening as only reasonable explanation for why the matchmaking is so bad.
    Most of the time when I face Top 10 ranked players they are duo queuing with some account I've never seen before or can't find on the leaderboard (while playing equally well), which eventually made me suspicious, thinking that's what is happening.

    I expect to face them when I reach P3, but not still after dropping as low as P1, especially since I generally don't seem to have a balancing top player on my side. Oh well.

    It's not what's happening though.

    The matchmaker does all it can to find a match for legend players too so they don't have to wait 40min - 1h like they had to once.

    This does mean you're gonna run into high P3 and Legs when you're low plat sometimes.

    @Exedore.6320 said:
    Please don't spread misinformation. The average rating of the party is used for matchmaking. Numerous people have tested and confirmed with math.

    I'd like to see source for that. I know it uses the average rating of the roster for matchmaking as seen in the matchmaking algorithm. But I'm pretty sure it excludes the lowest member of a duo when it does this.

  • Eugchriss.2046Eugchriss.2046 Member ✭✭✭

    @Aktium.9506 said:
    But I'm pretty sure it excludes the lowest member of a duo when it does this.

    I'd like to see source for that.

  • Exedore.6320Exedore.6320 Member ✭✭✭✭
    edited June 22, 2020

    @Aktium.9506 said:

    @Exedore.6320 said:
    Please don't spread misinformation. The average rating of the party is used for matchmaking. Numerous people have tested and confirmed with math.

    I'd like to see source for that. I know it uses the average rating of the roster for matchmaking as seen in the matchmaking algorithm. But I'm pretty sure it excludes the lowest member of a duo when it does this.

    You can do it yourself, as I and many others have done in the past:

    • Play a few solo queued games and note your rating adjustment. If you have a stable rating, this should be around 12-15 points.
    • Find a friend who is several hundred rating different from you. Have them note their rating adjustments. They should also be 12-15.
    • Queue together and each note your rating adjustments.

    If average is used for matchmaking, the higher rated player gets few points on a win and loses many more in a loss compared to solo. The lower rated player gets more points on a win and loses fewer points on a loss compared to their solo queue adjustment.

    If highest rated player's rating was used for matchmaking, then the higher player's adjustment would be the same as solo, but the lower rated player would see enormous gains on a win and almost nothing on a loss.

    Everyone who has done this quick test saw the first outcome.

  • Khalisto.5780Khalisto.5780 Member ✭✭✭✭

    I sarted this season 1180 and something won my first 3 matches and in the 4th one i was playing against anya and mur that att were top 1 and 2. I think things are even worse in NA, you start facing this top 50 duoa when you are about 1350, so you have to dodge this duos and the lower duos also dodging the highest ones if you play solo like me.

  • Aktium.9506Aktium.9506 Member ✭✭✭

    @Exedore.6320 said:
    You can do it yourself, as I and many others have done in the past:

    • Play a few solo queued games and note your rating adjustment. If you have a stable rating, this should be around 12-15 points.
    • Find a friend who is several hundred rating different from you. Have them note their rating adjustments. They should also be 12-15.
    • Queue together and each note your rating adjustments.

    If average is used for matchmaking, the higher rated player gets few points on a win and loses many more in a loss compared to solo. The lower rated player gets more points on a win and loses fewer points on a loss compared to their solo queue adjustment.

    If highest rated player's rating was used for matchmaking, then the higher player's adjustment would be the same as solo, but the lower rated player would see enormous gains on a win and almost nothing on a loss.

    Everyone who has done this quick test saw the first outcome.

    That's just rating gain/loss. It has nothing to do with the actual matchmaking. I'm also aware it differs depending on the your own rating.

    I've done this but for a different reason in the past. I tanked myself really hard once down to like low gold 3 trying to make a really bad build work. Asked a friend of mine in P3 for duo to get myself back up at my usual rating knowing we'd face harder opponents but also getting teammates that know how to play the game. Worked more or less the way I wanted it to. We got nothing but teammates and opponents within his rating range for all our matches. Even though he lost 17-20 rating per loss when I lost 7-12 tops the few losses we had until I got up near a similar rating to his.

  • silklash.6154silklash.6154 Member ✭✭
    edited June 25, 2020

    lol matchmaking in gw2 is so kitten. As a gold 2 player, I somehow was placed in 3 games against 3 dukes/duchesses of the arena, other p1 players like Cuilan, and some p2 players.

    They complained that the rest of us sucked. Well no kitten, I'm not a plat level player.

    I'm never touching GW2 PvP again, there's just no point. It's not fun for me, it's not fun for my teammates, and it's probably not fun for the people I'm playing against, so what's the point. I get maybe queue times are an issue, but if the solution from the developers are to have p2 players interacting/matched up against g1-2 players, it's just gonna drive players away because there's no purpose to playing the game.

  • Eddbopkins.2630Eddbopkins.2630 Member ✭✭✭✭
    edited June 25, 2020

    I havnt played this game in weeks and im scared that my g2 as-s will be put up against p2 and higher people. Knowing that i will just stay away from gw2 altogether. I dont need to be made a fool of against people who are hundreds of rank points above me.

  • @Asum.4960 said:

    @Aktium.9506 said:

    @Broady.2358 said:
    I, as a gold 3 player even with a plat 1 duo partner are in the same game as these top tier players.

    Your rating isn't even taken in consideration by the matchmaker if you duo queue with someone with a higher rating.

    It uses the highest player's rating in the party for matchmaking purposes.

    Are you sure about that? Then the population/matchmaking must be a lot more dire than I thought with frequently seeing literally 2 of the Top 5 players in a team vs entire teams with not a single person on the leaderboard. Seems odd.

    Actually you're probably right @Asum.4960 . Everyone sort of just believes and says that it uses the highest players rating for MMR, though there's no indication that this is the case.

    The matchmaking algorithm is available on the wiki, and DuoQ/teams aren't even mentioned at all there. Really there's nothing that would suggest that the match matchmaker doesn't use each individual's rating when matchmaking.

    Apparently a dev confirmed one time that DuoQing does use the highest rated player's rating for matchmaking, but nobody seems to be able to find the post/stream where this was said. I've done digging before and there's not one dev comment on it on these forums, or the old forums though there's always the chance I missed something.
    Really the closest thing to a proper discussion on it is this from the old forums: https://forum-en.gw2archive.eu/forum/game/pvp/Match-rating-loss-based-on-your-team-s-rating

    And even then, most of the replies appears to be contrived bull come up with on the spot by a top player and advocate for Solo/DuoQ. It's one of the first mentions of it, and that same top player is one of the only people still talking about it now.

    There's no solid evidence that the matchmaker uses the highest rated player's rating when matchmaking for a DuoQ, and even if it did; that would still be a very flawed system, because it would help higher-rated players boost lower rated accounts way faster either way. Bringing a silver-rated account with the initial rating boost into high-plat/legendary games is an amazing idea. 10/10
    Probably fell victim to this yourself, OP. 🙏

    Remove or Split DuoQ, revolution of our time! 😷

  • Shiyo.3578Shiyo.3578 Member ✭✭✭✭
    edited June 26, 2020

    Once you get to mid gold3 you start getting top20 players against you 75% of the time. It's honestly unfun and silly going from playing people I can 1vs2 to top20 immediately. Meanwhile I get teammates that just feed all game...

    There's too few medium skilled players so the entire matchmaking system doesn't function.

    @Khalisto.5780 said:
    I sarted this season 1180 and something won my first 3 matches and in the 4th one i was playing against anya and mur that att were top 1 and 2. I think things are even worse in NA, you start facing this top 50 duoa when you are about 1350, so you have to dodge this duos and the lower duos also dodging the highest ones if you play solo like me.

    Yeah this is EXACTLY what happens. It's awful

  • Swagg.9236Swagg.9236 Member ✭✭✭✭

    @Eugchriss.2046 said:

    @Broady.2358 said:
    For context I was playing on a Monday at around 3pm.

    Most if not all eu countries have ended their quarantine and people are back to work/school.

    This is the most pathetic defense I've seen for GW2's PvP community population.

  • phokus.8934phokus.8934 Member ✭✭✭✭

    @Aktium.9506 said:

    @Asum.4960 said:
    Yes, that's pretty much what I thought was happening as only reasonable explanation for why the matchmaking is so bad.
    Most of the time when I face Top 10 ranked players they are duo queuing with some account I've never seen before or can't find on the leaderboard (while playing equally well), which eventually made me suspicious, thinking that's what is happening.

    I expect to face them when I reach P3, but not still after dropping as low as P1, especially since I generally don't seem to have a balancing top player on my side. Oh well.

    It's not what's happening though.

    The matchmaker does all it can to find a match for legend players too so they don't have to wait 40min - 1h like they had to once.

    This does mean you're gonna run into high P3 and Legs when you're low plat sometimes.

    @Exedore.6320 said:
    Please don't spread misinformation. The average rating of the party is used for matchmaking. Numerous people have tested and confirmed with math.

    I'd like to see source for that. I know it uses the average rating of the roster for matchmaking as seen in the matchmaking algorithm. But I'm pretty sure it excludes the lowest member of a duo when it does this.

    How can you say you're pretty sure when there's no information in the wiki matchmaker regarding this?
    A more logical thought is that duo queuing doesn't matter and the matchmaker will attempt to build a game based on the professions and average Glicko-2 rating for both teams.

    So even if you have a legendary player and a gold player queued up together, you could still play against a legendary player, plat players, and gold players as long as the average rating is within an acceptable deviation.

  • jsp.6912jsp.6912 Member ✭✭✭

    and imagine NA is worse