7/9 games decided by matchmaker

[draxynnic.3719]

So, to give some context:

Back when I was playing sPvP regularly, I generally finished a season somewhere in gold (noting that "regularly" meant "average of one game per day", so... not enough to get into leaderboards even if my rating was high enough). However, I quit around the release of SotO out of a combination of having already achieved most of my goals from sPvP, growing increasingly dissatisfied with match quality, and the changed daily system being the final push.

With the rush event this week, I figured I'd record the results (all in soloqueue) and see how things worked out. Here's what I had when I completed the rush and decided that I wasn't close enough to the next ranked chest to be worth continuing:

Day 1:
Game 1: Lost 167-502
Game 2: Won 501-157
Game 3: Won 503-119
Day 2:
Game 4: Won 501-181
Game 5: Lost 111-500
Day 3:
Game 6: Lost 128-502
Game 7: Lost 479-500
Day 4:
Game 8: Won 500-150
Day 5:
Game 9: Won 504-491

Now, it's a small sample size, but you should be able to see the reason for the thread title here. Out of nine games, the losing side only managed to get over two hundred points in two of the games played. One of those, incidentally, nearly turned into "game decided by wall bug" when I got pushed into a wall by a downed revenant and couldn't get out, but the enemy team either didn't realise or was sporting enough to kill me anyway, so we only lost a captured point rather than being 4v5 for the rest of the game.

Other observations that I considered worth noting down is that in game 4, the game actually felt like it was somewhat in balance for most of the first half, and then the enemy team just seemed to collapse and ended up being farmed at spawn (even though I had cat aggro and was playing at a handicat at the time). I suspect, but cannot say for sure, that they had an AFKwitter. Game 5 I definitely had an AFKwitter on my team, but the enemy team also had at least one top-ten player (keep in mind the context that I'm a normally-gold player that's months out of practice) and it was pretty clear from early on that it would take a miracle turnaround to win, so I don't think that AFKwitter substantially altered the outcome.

Now, the win-loss ratio is even enough that I'm not complaining on that front, but in terms of match quality... well, the losses felt like I'd have needed to be a top player myself to be able to substantially affect the outcome, while the wins felt like all I needed to do was not be a leech and it didn't really matter how well I played there either. Games 7 and 9 were the only ones where I felt my skill and ability to make the right decisions, or conversely my mistakes, actually made a meaningful influence on the outcome.

I don't know if there's anything ArenaNet can realistically do at this point (although I do think there are issues with how rank points are assigned that can make it easy for players to end up in a bracket that doesn't match their skill, especially early in a season), but it does feel like it's reinforcing my decision that sPvP just isn't really worth it. For soloqueue at least, the matchmaker just has such an outsized effect on the outcome that it's pretty much a coin flip with extra steps, with a small chance that it comes down edges and you actually get a good match.

[Saiyan.1704]

41 minutes ago, draxynnic.3719 said:

So, to give some context:

Back when I was playing sPvP regularly, I generally finished a season somewhere in gold (noting that "regularly" meant "average of one game per day", so... not enough to get into leaderboards even if my rating was high enough). However, I quit around the release of SotO out of a combination of having already achieved most of my goals from sPvP, growing increasingly dissatisfied with match quality, and the changed daily system being the final push.

With the rush event this week, I figured I'd record the results (all in soloqueue) and see how things worked out. Here's what I had when I completed the rush and decided that I wasn't close enough to the next ranked chest to be worth continuing:

Day 1:
Game 1: Lost 167-502
Game 2: Won 501-157
Game 3: Won 503-119
Day 2:
Game 4: Won 501-181
Game 5: Lost 111-500
Day 3:
Game 6: Lost 128-502
Game 7: Lost 479-500
Day 4:
Game 8: Won 500-150
Day 5:
Game 9: Won 504-491

Now, it's a small sample size, but you should be able to see the reason for the thread title here. Out of nine games, the losing side only managed to get over two hundred points in two of the games played. One of those, incidentally, nearly turned into "game decided by wall bug" when I got pushed into a wall by a downed revenant and couldn't get out, but the enemy team either didn't realise or was sporting enough to kill me anyway, so we only lost a captured point rather than being 4v5 for the rest of the game.

Other observations that I considered worth noting down is that in game 4, the game actually felt like it was somewhat in balance for most of the first half, and then the enemy team just seemed to collapse and ended up being farmed at spawn (even though I had cat aggro and was playing at a handicat at the time). I suspect, but cannot say for sure, that they had an AFKwitter. Game 5 I definitely had an AFKwitter on my team, but the enemy team also had at least one top-ten player (keep in mind the context that I'm a normally-gold player that's months out of practice) and it was pretty clear from early on that it would take a miracle turnaround to win, so I don't think that AFKwitter substantially altered the outcome.

Now, the win-loss ratio is even enough that I'm not complaining on that front, but in terms of match quality... well, the losses felt like I'd have needed to be a top player myself to be able to substantially affect the outcome, while the wins felt like all I needed to do was not be a leech and it didn't really matter how well I played there either. Games 7 and 9 were the only ones where I felt my skill and ability to make the right decisions, or conversely my mistakes, actually made a meaningful influence on the outcome.

I don't know if there's anything ArenaNet can realistically do at this point (although I do think there are issues with how rank points are assigned that can make it easy for players to end up in a bracket that doesn't match their skill, especially early in a season), but it does feel like it's reinforcing my decision that sPvP just isn't really worth it. For soloqueue at least, the matchmaker just has such an outsized effect on the outcome that it's pretty much a coin flip with extra steps, with a small chance that it comes down edges and you actually get a good match.

What is your question....? Just kidding...maybe. I'll attempt to answer but it's not going to be satisfying to you in the slightest.

First off, matchmaking ATTEMPTS to make things 50/50 as close as possible but this is exploitable.... that said... you are not in this exploitable margin by any sense of the word. Possible? Absolutely, but you're not causing the odds in your favor as a PvP solo player below Gold2.. Others... a few... are causing the odds... in THEIR favor.... You're just participating in their mini gamemode. You don't notice is happening. That said, nothing is stopping you from hitting Gold3/Plat1 as a solo player. I'm dead serious.

Secondly and excuse my insult, you are not a higher rated player to bend the odds towards you're favor because you don't know how to do that. You don't need to show me game footage or Twitch vods. I can see that through your OP. Because you're a casual (like 99% of others) who don't feel the need to grind games just to climb from Silver2 to Gold3.... which i have done. Many other have done that... I have unfortunately done that... but for what gain? Exactly. Nothing to gain from the 155+ frustrated Ranked games in a season. It's not worth the sweat whether you CAN or CAN'T make the grind.

[Gotejjeken.1267]

Here's a question...what do you expect from sub-gold games? Same applies to unranked really.

I've done most of my rush even in unranked because of this exact reason--newer players will inevitably queue ranked and make things more difficult than they usually are.  

The point I'm getting at is, in unranked (as a general)/sub-gold everyone is a 'less than stellar' player.  You may get a 'top X' player on your team because of the matchmaker, but that is literally the only outlier here.  When everyone is subpar, it can't be blamed on the matchmaker.  

I can see on the flip side as a p+ player if you are stuck on carry duty all the time, that gets frustrating.  But if you are on the other end of the spectrum, there's unfortunately not a ton you can do but learn to outskill your opponents at least 2 to 1.  Then you will rise in rank and get to where most of us are complaining about, and that's g3+ where you are stuck in 'elo hell' between being on carry duty and being on some top X players stream with their duo carry chasing you all game.  

Whether all that makes pvp 'worth it' or not is subjective--I would tend to agree with you at the lower skill level end it most likely isn't, because it is a huge gap between 'low skill' and 'high skill'.  As said above, if you are in low skill your teammates will most likely either be new to the game or no skill at all, and if you are 'high skill' your opponents are probably the same 10 streamers on mains or alts.  

The conclusion for me is, sPvP has no middle ground now.  There is no 'average' rank.  So, if you are even average you will end up in 'low skill' and most likely justifiably ragequit the mode.  

[draxynnic.3719]

Just to clarify: When I say I normally finish a season (when I do finish a season) around gold, to be precise, I normally finish around gold 2. I have climbed as high as Platinum 1, although in full self-awareness, that probably involved being carried higher than I probably should have been (see comment about there being problems with how ranking is awarded). So while I acknowledge I'm not the most dedicated sPvP player in the game, I think you're both assuming I'm lower than I actually am: this isn't a sub-gold issue. I'm pretty sure I was Gold 2 when I quit last year: I didn't quit because I was losing, I quit because of the "might as well be flipping a coin" effect I described, where even the wins weren't satisfying. I could play one more game to see where the game thinks I am now, but on the other hand, from my perspective that would essentially be qualifying on 2 or 3 games where I actually made a difference for good or ill.

I'm not really posing a question per se, mostly presenting my experience to demonstrate (and quantify, albeit with a low ) an issue - one which I suspect is a bit of a vicious cycle when it comes to population, since while I'm sure there is some level at which your skill always matters, most people aren't exactly going to start off there. (I don't think it's a problem of mid skill not existing, more that low skill and high skill both tend to get mixed into mid such that games are often decided by which team had more (possibly under-ranked) high-skill players and which had more (possibly over-ranked) low-skill players.

[arazoth.7290]

Well I had recently some sweaty matches because I had to farm the other team for many snowball effects. Because the moment I left for a decap because teamates didn't decap 2 capped nodes, they started dying like flies and couldn't be rallied...

So that I wouldn't get in a outmanned situation of more then 3 (2 were killable but 3 or more nope), I had to keep that going. So that in the hope teamates hopefull enough cap other node. 

In the end I got +5 rating 😅 for sweating my kitten off 😂😂.

Ps. It was solo queue which is what I mostly do

[SoftFootpaws.9134]

i just wish it would stop putting us in teams of the same players over and over, i'm pretty sure this is responsible for alot of losses (and wins for the other team). i bet if you record the data of the players that end up on your team, over half of them will be the same through the entire run, especially at night!

 

i've found that all you can do when you keep getting similar teams is to log off, because it won't change.

Edited January 28 by SoftFootpaws.9134

[ccccc.4963]

4 hours ago, SoftFootpaws.9134 said:

i've found that all you can do when you keep getting similar teams is to log off,

Stop logging in

[Yaki.9563]

This isn't a GW2 issue. It's an issue in all PvP, including WvW, and including other games. I just got done playing Warhammer: Return of Reckoning and suffering through 8 lopsided matches in a row in their solo/duo queues. The same happened many times in WoW when I used to play it. The match is generally determined in the first 2 minutes either because one side is dominant enough that the other team has no chance, or the other team loses cohesion and morale.

[Flowki.7194]

The game is fking horendous.. its dead. Every game is warrior/necro/gaurdian 95%. Low skill support, low skill damage, low skill face tank. Other 5% are the very few players actually playing something remotely resembling skill/risk.. along with the sweaty duo plats who all seem to have gravitated to Hollo/untamed/SB

 

MMR is least of the concerns.

Edited January 28 by Flowki.7194

[Gotejjeken.1267]

10 hours ago, draxynnic.3719 said:

Just to clarify: When I say I normally finish a season (when I do finish a season) around gold, to be precise, I normally finish around gold 2. I have climbed as high as Platinum 1, although in full self-awareness, that probably involved being carried higher than I probably should have been (see comment about there being problems with how ranking is awarded). So while I acknowledge I'm not the most dedicated sPvP player in the game, I think you're both assuming I'm lower than I actually am: this isn't a sub-gold issue.

Okay, when you say you finish around gold 2, and an entire season (120-150 games) it seems as if most of those games are g1 and below? 

As last season I started at top end of g3, fell down to high end of g1, and then finished in mid-g3 again.  That's with ~160 games and I think a 51% win rate (won 1-2 more games than I lost).  That's on full dice roll solo q, a build that was good like 3 years ago, no decay playing, etc. etc.

It's the point about extremes earlier--'gold 2' is average now and then top g3+ is Top 250+.  Gold 1 and especially below that might as well be the 'clueless' category.  So even if you are in gold 1 most of the time, you are close enough to the 'clueless' players to get bad games 70+% of the time.  

This is for NA btw, but I imagine the same would hold on EU as I thought someone said Top 250 there was 1550 instead of the 1470 it was on NA last season.  Which would mean your 'clueless' category is probably even gold 2 over there.  

It's not a matchmaker thing unless you are higher up.  The tldr; about being average or lower is as again I said above, there is no average.  There are top players, clueless players, then the outliers stuck in a vicious cycle between that.  

Edited January 28 by Gotejjeken.1267

[Saiyan.1704]

I had two seasons in a row where I got placed in Gold 1 and had to grind back up to Plat. It was enduring. Probably took me a week worth of grinding.

Placements only matters if you get placed relatively high. The higher you're placed, the less work you have to do.

[Saiyan.1704]

No ccccc stop typing

[ccccc.4963]

3 hours ago, Yaki.9563 said:

The match is generally determined in the first 2 minutes either because one side is dominant enough that the other team has no chance, or the other team loses cohesion and morale.

Games healthy and fun when everyones on power dps.  lopsided games happen when one team is too imbal with sidenode/support/decap players (useless/bad/wvw sentry cappers)

Edited January 28 by ccccc.4963

[Goldilock.2584]

Pvp comes down to 2 players 6/10 games can those two carry the map better than the 2 players on the other team. Those two players first look at their comp they should always assume the 3 others are just playing conquest and not 5v5 for map control. They let those 3 players play the roles the class they picked put them in and just farm hard the 3 worse players on the other team aswell as +ing into the better 2 on the other team. The higher the iq of players that are a duo will win most games doing this. This is why players like naru can win most games no matter who they have on their team and stomp the bad 3 and out play the better 2. If they loose its because the 3 they have just are achiveing nothing around them. Gw2 pvp is about who can farm 3 players better than the other team.

So if u get a top player just play around them be the 1 than caps be the one that watches mid and close and let them roam agressivly and if where u are is the play needed to carry map control they will come back to peel u out.

[Flowki.7194]

3 hours ago, Goldilock.2584 said:

Pvp comes down to 2 players 6/10 games can those two carry the map better than the 2 players on the other team. Those two players first look at their comp they should always assume the 3 others are just playing conquest and not 5v5 for map control. They let those 3 players play the roles the class they picked put them in and just farm hard the 3 worse players on the other team aswell as +ing into the better 2 on the other team. The higher the iq of players that are a duo will win most games doing this. This is why players like naru can win most games no matter who they have on their team and stomp the bad 3 and out play the better 2. If they loose its because the 3 they have just are achiveing nothing around them. Gw2 pvp is about who can farm 3 players better than the other team.

So if u get a top player just play around them be the 1 than caps be the one that watches mid and close and let them roam agressivly and if where u are is the play needed to carry map control they will come back to peel u out.

You make it sound so easy. Meanwhile, there are 2 core gaurds + a DH/scourge camping on mid, and an SPB+druid camping your home (multiple ofther specs for same effect). The mid fight needs a 4v3 to win and not just stale mate (bad if they have cap), and the home fight needs a 3v2 to break a 5 min stalemate. Do the math, and then factor in feeders.

 

There are too many easy specs in the game, that are unpunishable within a decent time frame, outside of +1ing them.. or with a vast skill gap. That is why gold is absolute hell, and any decent player on a full support, or anything not able to skillfully roam and overpower such specs, gets caught up in the MMR range, the lottery of what team has the least newbs, or what team has the most pro roamer... who typically offpeak duo. A typical average skill roamer in this game is not doing anything.. they are going up against bunker/duelists like druids, and other pro hollo roamers etc. In-fact the average skill level roamer makes up 90% of my losses.. all they do is go feed on far all game and call it "rotating". I got top offense as support yersterday, from being forced to cap back our home all game, while 3 of our players "rotated" to far all game, and never capped it once.

Edited January 29 by Flowki.7194

[Bunbury.8472]

tfw you hit a level where you can no longer carry games, but you refuse to accept the reason you can no longer carry is because you are correctly ranked

😇

@Goldilock.2584 if only playing like that didn't get boring as hell... it does work tho

😇

Edited January 29 by Bunbury.8472

[Flowki.7194]

53 minutes ago, Bunbury.8472 said:

tfw you hit a level where you can no longer carry games, but you refuse to accept the reason you can no longer carry is because you are correctly ranked

😇

@Goldilock.2584 if only playing like that didn't get boring as hell... it does work tho

😇

Define correctly ranked in game where plats actively go out of their way to avoid each other, and farm silvers/golds on the current FOTM roamer/jump spec. I question the true skill level of any current plat player, who are absolutely untested in equal ranked compitition. I mean, not bragging at all.. but ive killed multiple  "plat" players in duels with a fking core rev support build. As far as I can see, 80% of "plats" are just que dodging FOTM duo golds, at best. Ive played against Boyce and similar caliber players more than enough times now to see the obvious skill gap, which the average "plat" players are bridging with que dodging and hard sweating fotm builds. Or in other words, if MMR was capped to elo, like it should be, 80% of those so called plats would nose dive into gold 2, and plenty of decent support players would rise to p3 or more. The game is absolutely tilted to favour a very select and toxic playstyle.. plats farming golds/silver.. and people just brush that off like its nothing "it gets boring". Yeah, kills the game aswell.

 

No offense to Boyce and the likes, but it fking baffles me why people watch those live streams, litteral top level players in games with players with half their skill, stomping the kitten out of them. People actually watch this?... which is 100% the problem. The games best players should litterally be banding together and demanding some fair compitition, MMR locks that matter.. and stop this onsluaght of newb farming and que dodging. Obviously the top players don't give a kitten, they are making money from it.

Edited January 29 by Flowki.7194

[arazoth.7290]

1 hour ago, Bunbury.8472 said:

tfw you hit a level where you can no longer carry games, but you refuse to accept the reason you can no longer carry is because you are correctly ranked

😇

@Goldilock.2584 if only playing like that didn't get boring as hell... it does work tho

😇

Now you have to ascend to god lvl player skill wise to carry/farm even high plats duo occasionally 

[arazoth.7290]

1 hour ago, Flowki.7194 said:

Define correctly ranked in game where plats actively go out of their way to avoid each other, and farm silvers/golds on the current FOTM roamer/jump spec. I question the true skill level of any current plat player, who are absolutely untested in equal ranked compitition. I mean, not bragging at all.. but ive killed multiple  "plat" players in duels with a fking core rev support build. As far as I can see, 80% of "plats" are just que dodging FOTM duo golds, at best. Ive played against Boyce and similar caliber players more than enough times now to see the obvious skill gap, which the average "plat" players are bridging with que dodging and hard sweating fotm builds. Or in other words, if MMR was capped to elo, like it should be, 80% of those so called plats would nose dive into gold 2, and plenty of decent support players would rise to p3 or more. The game is absolutely tilted to favour a very select and toxic playstyle.. plats farming golds/silver.. and people just brush that off like its nothing "it gets boring". Yeah, kills the game aswell.

 

No offense to Boyce and the likes, but it fking baffles me why people watch those live streams, litteral top level players in games with players with half their skill, stomping the kitten out of them. People actually watch this?... which is 100% the problem. The games best players should litterally be banding together and demanding some fair compitition, MMR locks that matter.. and stop this onsluaght of newb farming and que dodging. Obviously the top players don't give a kitten, they are making money from it.

On average people prefere the ez way then having it whole the time giving their sweating best

[draxynnic.3719]

On 1/28/2024 at 9:23 PM, arazoth.7290 said:

Well I had recently some sweaty matches because I had to farm the other team for many snowball effects. Because the moment I left for a decap because teamates didn't decap 2 capped nodes, they started dying like flies and couldn't be rallied...

So that I wouldn't get in a outmanned situation of more then 3 (2 were killable but 3 or more nope), I had to keep that going. So that in the hope teamates hopefull enough cap other node. 

In the end I got +5 rating 😅 for sweating my kitten off 😂😂.

Ps. It was solo queue which is what I mostly do

Yeah, this looks like one of the issues with how points are assigned. The game gives or takes away points based on your MMR versus the average of the enemy team - but last I heard, it doesn't take into account the average of YOUR team. So you can have situations where you're being asked to carry a team that has a lower average MMR than the other team, but your own MMR is above the average of the enemy team. This often makes for the hardest games to win, but you get very few points if you do, but stand to lose a lot if you fail.

Meanwhile, the reverse is true - someone who's basically being carried by their team can get a lot of points for fairly minimal effort.

While I think this eventually stabilises if you play enough games, it does mean that at any given time there are a bunch of players who are significantly above or below their proper rating due mostly to whether they've had games where they're expected to carry versus being carried. Which just makes it harder for the matchmaker to make fair matches.

On 1/28/2024 at 9:42 PM, SoftFootpaws.9134 said:

i just wish it would stop putting us in teams of the same players over and over, i'm pretty sure this is responsible for alot of losses (and wins for the other team). i bet if you record the data of the players that end up on your team, over half of them will be the same through the entire run, especially at night!

 

i've found that all you can do when you keep getting similar teams is to log off, because it won't change.

When I do play multiple matches a day, I usually try to space them a few hours apart for this reason.

[Flowki.7194]

43 minutes ago, draxynnic.3719 said:

Yeah, this looks like one of the issues with how points are assigned. The game gives or takes away points based on your MMR versus the average of the enemy team - but last I heard, it doesn't take into account the average of YOUR team. So you can have situations where you're being asked to carry a team that has a lower average MMR than the other team, but your own MMR is above the average of the enemy team. This often makes for the hardest games to win, but you get very few points if you do, but stand to lose a lot if you fail.

Meanwhile, the reverse is true - someone who's basically being carried by their team can get a lot of points for fairly minimal effort.

While I think this eventually stabilises if you play enough games, it does mean that at any given time there are a bunch of players who are significantly above or below their proper rating due mostly to whether they've had games where they're expected to carry versus being carried. Which just makes it harder for the matchmaker to make fair matches.

When I do play multiple matches a day, I usually try to space them a few hours apart for this reason.

Thats the problem with full support, or even hybrid support. I can get to g3, but then im the "carry" player, and get silver dps vs plat hollos etc. Then, 10-15 games of that, back down to g1.. and side noding with a fking hybrid support and getting 25-0. They geneuinely need to fix the way supports get points. If you have over 600 healing, you should have reduced point loss or increased point gain.. if they intend on keeping this god aweful MMR.

[Cuks.8241]

Snowball effect is something that is present in most pvp games. Now I'm not saying that all the games author listed are this way due to snowball because we're talking about a solo queue system in a team game. A team needs to play as a team to snowball efficiently. But in any case an early advantage can often lead to snowball.

It is not unusual in competitive games that even when teams are very equal the final scoreboards have huge gaps.

For example, this happens a lot in my experience. You will have a very close initial team fight and you only lose it because of minor details. Maybe you went for a bad stomp and didn't make it and lost a fight. The fight might turn in your favour if it would be repeated and things look very balanced in that fight. But after that fight all goes downhill. Enemy gets advantage, your team doesn't regroup, they trickle back alone, they get frustrated, they try to force with bad and risky rotations... 

Keep in mind also that the better the players, the better they will know how to keep an advantage and snowball it. They will bleed you out slowly, they will take a point and not just relinquish it, it will not be easy to get it back. They will take the advantage when you're trying to stabilise to take the objective. And some objectives are really strong and snowbally in this game.

[DoomNexus.5324]

12 hours ago, Flowki.7194 said:

They geneuinely need to fix the way supports get points. If you have over 600 healing, you should have reduced point loss or increased point gain.. if they intend on keeping this god aweful MMR.

Anet should in general realize that different roles should be rewarded for different tasks differently. This also requires recognizing which role (or at least in which category of the holy trinity) a build falls. A burn firebrand is not comparable to a support firebrand for example. Probably easiest thing to take into account is the amulet but there are other more sophisticated ways for sure. Just matching amulets with the elite spec tho, would result in a matrix of still manageable size while probably already giving a pretty good estimate what to expect from that character. Imo that's also a mandatory feature for the matchmaker to ensure proper team compositions and not match a team with support, team fighters and a roamer against only roamer builds with no support or kitten like that and I think it's kinda mind-boggling that GW2 still doesn't have that after 12 years. Well.. not so mind-boggling when you consider that PvP has been abandoned since at least 9 years ago but still.

I do realize that a really good system for that is hard to execute in a complex game like gw2 but there certainly are ways (other games can do it too, even just third-party rating tools exist for some games) and even just a semi-decent system will be magnitudes more fair than just looking at wins and losses.

[GuriGashi.5617]

Is is what it is lil bro 

[DanAlcedo.3281]

ALL match outcomes are determined by the matchmaker. That's literally it's job. 

Forcing a 50/50 winrate. 

Players are just a pawn the matchmaker moves around. 

The outcome of matches is predictable. 

Won a match? Statistically, given enough matches played, you are supposed to lose the next one. 

You know, matches being a coin flip and all.