My Main Acct vs. Alt Acct Ratings Played - What's Going On Here?

Khalisto.5780 · January 20, 2020

@Psycoprophet.8107 said:Imagine solo queue was the only queue lol I think people's post would be far different as would the top ratings. Not belittling anyone I just bet that would be the case.

I've seen suggestion to not let you duo when you reach plat, guess they working on the Swiss tournaments and we won't see this any soon

Khalisto.5780 · January 20, 2020

@Khalisto.5780 said:
@Ghetx.1752 said:
@Laera Myrith.7605 said:
@Ghetx.1752 said:
@Laera Myrith.7605 said:
@bluri.2653 said:
@Don Vega Van Kain.9842 said:
@bluri.2653 said:Ah yes the daily conspiracy theories.
90% winrate is impossible!My alt wins more than my main!50% forced w/r!
Its almost comical how much excuses you guys go through

Your alt account is almost all the time queued with a high ranked player in a duo so if you can't even play solo with your alt account just.....you know.

So is my main account so whats ur point again?

I was r1 and 2 last season but decayed both my accounts on purpose.

Stop feeding the self deluded trolls, i have 2 accounts also. My main sits comfortably within 50 rank points of my alt. The reason people have such bad match ups IMO, is because they queue at stupid hours when the population is low. Just my 2 cents.

Wait sec,doesn't this game has millions of players?Ingame time should not matter with big population ArenaNet and their fans keep promoting.Its not like there's 50 people play during night time and rest of 100 k are asleep.

The game might have "millions of players" but pvp as a mode doesn't. If you queue at 9am in the morning when people are at work/school its only logical that the matchmaker has a smaller pool of players to draw from.

Even if game has 5.000 players only,there's always going to be at least 100 people who play pvp,at all times.100 players is more than enough to keep queue alive at any given time at any Division.

There's no f way literally every single player is at work,shool or asleep.Its simply not possible by every mathematical and universal probability and possibility for everyone to be busy doing something else.

Not everyone has a job or goes to school,most people actually don't.

the pvp player base have in the best case scenario 2k players

when i get home I'll explain better how i come to that conclusion

ok, let's go

Devs made 3 statements about pvp, they're out there on forums

1) the population graphics is pretty much a bell curve2) this bell is centered on gold 2 (most populated division)3) bronze has more player than legendary, but not that much more

now let's go to assumptions and connecting dots

since it's a bell the division's number of player kinda mirror each other, the lowest divisions being a little more crowded (i guess)since gold 2 is the center the mirroring would be like

     gold 2

gold 3 - gold 1plat 1 - silver 3plat 2 - silver 2plat 3 - silver 1legendary - bronze

now we can see most, if not all plat players in the leaderboards

the ranked 250 guy was sitting at 1519, below him but still in plat we may have something around 30-50 players, and something like 50 players that haven't play the minimum required matches to be in the leaderboards <----lots of guessing here

so we have max something around 350 players in plat division, a little bit more in silver maybe 450, reaching 800 players, and that's the whole plat division and the whole silver division

the whole gold must have about the same population of these 2 combined reaching 1k, if that much

and with the legendaries and bronzes i think you have max 100 more players

now we also know the top 50 is more like 20 guys with multiple accounts, top 250 must have 150-170 players max, so the actual number of players is smaller

JusticeRetroHunter.7684 · January 20, 2020

For those that are asking Boyer for a video, I have known him for a long time (pretty sure he’s on my FL) and without I doubt I know he’s an experienced player. I’ve since quit Spvp since last year so I barely remember my spvp friends anymore but I’m sure a lot of the duel arena people are well acquainted with each other.

Anyway, about the matchmaker, there is a fundamental flaw with the matchmaker in that ranking isn’t a representation of actual skill level. This is an objective truth, because the system is designed to reward individual ranking for games won based on a random assortment of 4 other players.

Take this hypothetical scenario as an example

Bob is a noob to the game, Alice is a pro to the game. Both have some arbitrary unknown ranking. And both end up in matchs together:

Alice and Bob are required to play together in this first match. If they win, Bob is rewarded the same as Alice. They both gain the same increase in rank and are both considered by this system to be better players than the other team for winning this match.

Let’s now introduce Charlie and Daphne; players with some arbitrary skill level and ranking. Bob and Charlie end up playing against Alice and Daphne in the 2nd match. In this case, Alice and Daphne lose to Bob and Charlie.

According to the system, this now places Alice as a worse player than Bob, even though Alice is objectively a better player than Bob.

Now extrapolate this trend for thousands of players, and what ends up being the case, is that the ranking number becomes meaningless. Thus using the RANKING NUMBER TO DECIDE MATCHMAKING is fundamentally flawed as it drives towards trends like the example mentioned above.

I explained this in more depth on another thread a while ago, where I go into more detail about the extrapolation, but essentially matchmaking turns into random noise. A player with 1200 rating (like your main) is actually playing with a skill level of a platinum 3; with and against other players of 1200 rating (who may also not even be playing at a gold 1 skill level, could be higher or lower, nobody really knows)

On a personal note, I’ve seen video evidence of skill clickers in plat 1 games. How hard do veteran players have to carry players like this in such high elo? And if you are able to carry, you are just making the skill clicker get a higher rank.

The system is broken from the bottom up man, that’s the sad truth and it’s always been like that since S1.Edit; also yes the matchmaker does in fact take personal play history as a factor for match making, as well as time spent online. You can read all the things it takes into account on the wiki...dishonor, guild tag, play time, class history, games played...they all factor into the matchmaking. Don’t believe what anet reps might tell you otherwise, they say “it’s not that much” but it really is, based on that code it’s a deciding factor for what team you will be on in a match.

Zexanima.7851 · January 20, 2020

@Sigmoid.7082 said:
@Zexanima.7851 said:
@"bluri.2653" said:Ah yes the daily conspiracy theories.
90% winrate is impossible!

It shouldn’t be possible in a balanced match making system.

This isn't exactly true. Balanced doesn't always mean 50/50. You can have a 50% chance to win every match you play and still lose every single one.

Lose streaks can happen but over time it'll average out. Lose streaks should also not be the norm.

50% forced w/r!

This is how it SHOULD work. The matchmaker should pair you with similarly rated people people and you’ll average close to 50%, a littler higher if your good. Go find me a main stream competitive ranked game where the best players average a 90% win rate. I’ll wait.

And it does try to do exactly this with who is available. Its just the amount of people available isn't that great. It also means the game is less competitive , meaning less need to play games to remain at the top, resulting in fewer games played which in turn means you don't play enough games to average out your win-rate. Those at the top have such a high win-rate because they are very far above the skill curve and there is a lack of competition.

Also Anet has also said that they don't want it to be so punishing that the better you get the less you play because queue times are so high. They have tightened the rating variance a few times in the algorithm but there is still a range is searches in and even small changes in rating can be pretty big differences in skill and game knowledge. If they enforced a very strict skill range I can see the same people saying they are getting bad matches saying they are not getting enough matches since the queue times will become so much longer. There is also no guarantee the match quality would get significantly better.

It's almost like they have neglected the mode so long and the player base has thinned out so much that it's started to break down the match making system. Even if it's "working the best it can" that's still not how things should be. Legendary players should never get paired with golds but I've seen it happen. Specially for new players and players of other PvP games this looks really bad. Just, like, put work into the mode and promote it ffs anet so your damn match maker can do it's job right.

Zoricus.2439 · January 20, 2020

@JusticeRetroHunter.7684 said:
The system is broken from the bottom up man, that’s the sad truth and it’s always been like that since S1.

Very True. Trev's experiences are NOT an outlier. old time vets have the same issue on their main accounts for a long time now. Ranked punishes you mercilessly for things out of your control for no transparent reason.

Trevor Boyer.6524 · January 22, 2020

My alt is 1611 and rank 19 today.

Both game played on the alt felt very balanced. The games were EXTREMELY close, like 500 to 475 wins. It actually felt like everyone on my team and against me were of relatively equal skill level. Of course I played hard, but it didn't feel like I had to carry.

30 minutes prior to this, I had been playing on my main. Must have played about 6 to 8 games, working to grind my way back up out of that lose streak. Most of the games felt super sweaty and overly difficult. The matches in gold 3 on the main were several times more difficult than the plat 2 matches on the alt.

I just... I don't know what else to say or how else to explain it. ^ There is nothing right or normal about this.

Trevor Boyer.6524 · February 22, 2020

Ok, I'm fed up with this. I have no idea where to even start to sort out this problem. I'm sure submit ticket will yield little to no results, so I may as well make this on alert in the forums, in hopes that @Ben Phongluangtham.1065 @"Cal Cohen.2358" may pay attention, take it seriously, and look into it.

So firstly, take a look at this screen shot that was taken today on 2/22/2020, during the last 48 hours of this season. This is me on my alt that I've been talking about:

Here is the meat of it:

This ALT during season 20 has played as high as 1620 and as low as 1560. It generally played 1570 -1600 rating over 44 games played. Each seasons when I play this ALT, its match making has no funky rhythm with strange enormous win or lose streaks. It stays very solidly within this margin of rating and match quality. The matches make sense and it feels like the algorithm is doing its job. In fact, it feels so clean when I am on this ALT, that it makes realize it's no wonder that people "call you out on a conspiracy theory" if you're complaining about super strange bad match making, if they've only ever experienced the type of match quality that I get while playing this ALT.
The main during season 20, had bounced back & forth between 1320'ish rating and bottom 1500 margins several several times. Maintaining even a 1500 bottom plat level on this main account is ridiculously overly difficult to do. Each time I even TOUCH 1500+, I can expect a strong lose streak to occur almost immediately as the rating pushes 1500+. This generally always pushes me down into 1350 margins or lower, and the matches down in gold 2 are much much more difficult than the plat 2 rated matches that I get on the ALT, to the point that it doesn't make sense. And these beyond impossible to carry matches will keep occurring until the main account's win rate is 50% or bellow. Once it goes 50% or bellow, suddenly there will be a sudden enormous shift in the difficulty of matches, and I'll go on a tyrannical win streak back to bottom plat, and then this process just keeps rinsing & repeating over and over, with this strange volatility in match making. Lose streaks for 10+ games that are impossible to carry, and then win streaks 10+ that are impossible to lose. Yet this type of algorithm behavior DOES NOT HAPPEN while on the ALT. What is this?

It often feels as if: "I am not allowed to go beyond 1500 on the main", yet every season when I log in and play on the ALT, I mysteriously become top 50 material and never fall out of plat 2? Hopefully Ben & Cal are paying attention to this, and I'd like them to go review the match history & rhythm of these two accounts. If they were to do this, they could come in here and post to verify that I am being 100% truthful in saying that the ALT has played between 1560-1620, but the main somehow plays between 1320-1520'ish. There is only so long that we can toss copout types of responses at the information I'm posting in this thread, such as: "Well you play better on your ALT" or "You got lucky" or "The main is unlucky" because after viewing the ever growing sample size of these two account's very different rhythms within the algorithm, it begins to become obvious that something is going on here that is creating a very large difference in the match quality between the ALT and the Main. Why would the ALT always play tightly within a 40 to 60 rating margin up and down, whereas the Main would be getting shot up and down the leaderboards in a 200 rating variation constantly? I play the same builds on each account, do the same things, have same settings, even the character used is a copy/pasted toon who looks exactly the same and has the same name.

From a long time consumer's standpoint, this is beyond frustrating. I've invested a lot of money into the main account over the course of 8 years. I don't want that main account to be getting bogus rigged match making.

So what is it that's really going on here? Well this is pretty much what I've narrowed it down to:

Very old accounts with seriously inflated glory rank and total matches played are given way more difficult matches than newer accounts. If this is the case, this needs to be changed ASAP. Players reach a peak after awhile, but as the years go on that glory rank & total games played keeps going up. It isn't fair at all that this should force impossible odds in matches, when player peak skill is achieved much earlier on in an account's history.
Something that would have to do with the original form of dishonor, which could pigeonhole an account into bad matches over and over again. If an account was tagged with an inordinate amount of 1st type dishonor, it would keep being placed into teams with AFKers, Cheaters, Rage Quits, ect ect, and substantially increasing the rate of losing matches for stupid reasons. I understand the idea & purpose behind 1st type dishonor, but if this bad match quality is due to this, that's too much discrimination. I have no idea how this system works internally or how powerful it is at herding certain accounts into or away from each other, but if this is the case & source of the Main's bad match quality, there needs to be some kind of reset & erase factor this 1st type dishonor. I often wonder if I had pissed off the wrong person in-game at some point, and my main was slapped with some kind of permanent level of 1st type dishonor.
The Main could be being targeted by certain players within the community who manipulate matches because they do not like me. Why do some of these people hate me so much? Well, there are a lot of reasons for that, but the biggest reason is that I am loud and outspoken, especially when it comes to exposing illegitimate activity within the leaderboards. However, I honestly don't think that this is the case. The bad match quality doesn't feel like throws or manipulations, which normally I can spot rather easily. No, this bad match quality feels internal, as if the algorithm itself or some kind of administrative function were against my Main account.
And that leads me to my biggest question recently: Are there people who are allowed special administrative functions during match making? Let me start by explaining something odd that I noticed. So when I queue on the ALT, 9 out of 10 queues will have the box pop up and everyone hits accept and the game begins. But when I am on the Main, just about every game I queue has the box pop up, someone declines, box pops up again, someone declines, box pops up again, decline, and then finally when the game starts for the Main, the match is almost always an incredibly lopsided no-win situation. This almost always happens after heavy queue decline activity. After awhile, I began to wonder: "Are there people who can see who is in the queue and what team they are being put on? Are they declining over and over so they can try to land on a favorable side? Is it possible that some people have some way, like in custom arenas, to swap players from red to blue and from blue to red during ranked matches? Why would my Main account get so much of this fishy queue decline activity, but the ALT rarely ever sees it happen at all?" <- This is no embellishment and it is very suspicious at best.

So yeah anyway, something is wrong with this. I sincerely wish that the devs would look into this. If they care to do so, the ALT is: Asuran.5469

~ PS: And no, I am not account sharing or throwing or win trading. I own this ALT in my name, just the same as the Main.

phokus.8934 · February 22, 2020

It’s literally not any of those things.

There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.

The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.

I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

Trevor Boyer.6524 · February 22, 2020

@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

ArlAlt.1630 · February 22, 2020

@Trevor Boyer.6524 I took the challenge, started doing the obligatory rank 20 games on this ACC so I can queue ranked. Will post once I'm done.

Sindrener.1592 · February 24, 2020

@Trevor Boyer.6524 said:
@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

phokus.8934 · February 24, 2020

@Trevor Boyer.6524 said:
@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

A few things:1) 44 games in a season is not a lot to settle the rating volatility/deviation; it's just mathematically not accurate to say otherwise. Previous seasons games played is irrelevant when a new season starts. Soft resets occur using this algorithm: (Rating last season played + 1200)/2. Take into consideration the pvp population, their relative rating and the matchmaker building around a player. But consider this: Glicko-2 is used for 1v1s so we can safely assume that each team is treated as one player so the matchmaker builds as close a match as possible based on average rating, classes, duo queue, and time in queue.2) Your proof means jack if you don't have any of it controlled and corroborated by others. Hell, Sind just proved you wrong by indicating his alt accounts are well within his mains rating. Even when I watch Naru stream on his main and alt he's within his mains rating.3) No one's calling you a liar but you're in the deep end with legitimate conspiracy shenanigans. The 4 lists you mentioned in a previous comment have no bearing on anything regarding match quality. You just assume it has to be one of those because you're choosing to ignore the facts and any sense of reason. So yeah, you're the one with the crazy conspiracy theories as to why something is off when the simplest answer is the sad state of pvp population.

Bazooka.3590 · February 24, 2020

Rows exist whether Anet recognizes or not. If you stuck in a row you can fall endlessly. Once I fell from plat to near silver just to see if there is a bottom. Turned out it is if you break your keyboard:PAccording to this binary generator rows can be very long. I hope they are just the matter of luck and not some built in manipulation (as far as I know Anet denies this theory).Whether it is true or not rows are the most annoying things in this game and they should not exist. Maybe MM should do something with it and/or some changes in the rating system could help to prevent the player to lose hundreds of points at once. The game could be much more tolerable toward personal efforts.

yanniell.1236 · February 24, 2020

@bluri.2653 said:
@Trevor Boyer.6524 said:
@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu

130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

You do not even know how to solo queue. Duo queue proves absolutely nothing on the matter in hand.

Eddbopkins.2630 · February 24, 2020

Coding new code for a balanced PvP match maker in a spaghetti bowl full of spaghetti is really tough apperently.

Sindrener.1592 · February 24, 2020

@yanniell.1236 said:
@bluri.2653 said:
@Trevor Boyer.6524 said:
@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu

130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

You do not even know how to solo queue. Duo queue proves absolutely nothing on the matter in hand.

one of the alts have 40% soloq

Dantheman.3589 · February 24, 2020

If you have 2 accounts you play seriously on whichever one sees less common play is very likely to be higher ranked because u play less games and are usually more serious during, unless ofc you are being carried through duo q than which ever one you duo on is more likely to be higher rated

Trevor Boyer.6524 · February 24, 2020

@bluri.2653 said:
@Trevor Boyer.6524 said:
@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu

130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

Eh?

All that really proves is that you in particular are not experiencing what I am talking about. However, other players have stated that they have.

@"phokus.8934" said:
A few things:1) 44 games in a season is not a lot to settle the rating volatility/deviation; it's just mathematically not accurate to say otherwise. Previous seasons games played is irrelevant when a new season starts. Soft resets occur using this algorithm: (Rating last season played + 1200)/2. Take into consideration the pvp population, their relative rating and the matchmaker building around a player. But consider this: Glicko-2 is used for 1v1s so we can safely assume that each team is treated as one player so the matchmaker builds as close a match as possible based on average rating, classes, duo queue, and time in queue.2) Your proof means jack if you don't have any of it controlled and corroborated by others. Hell, Sind just proved you wrong by indicating his alt accounts are well within his mains rating. Even when I watch Naru stream on his main and alt he's within his mains rating.3) No one's calling you a liar but you're in the deep end with legitimate conspiracy shenanigans. The 4 lists you mentioned in a previous comment have no bearing on anything regarding match quality. You just assume it has to be one of those because you're choosing to ignore the facts and any sense of reason. So yeah, you're the one with the crazy conspiracy theories as to why something is off when the simplest answer is the sad state of pvp population.

44 games in a single season is plenty large enough of a sample size to view. At the end of the first day after just the first 15 games, each season after season, we see the same exact players taking the same exact positions and maintaining them for the duration of the season. You know it and I know and everyone knows that is true. Besides that, we aren't discussing just season 20 here. We are also discussing why this same thing has happened for about the past 4 or 5 seasons, which is: The ALT playing 100 to 200 rating higher than the main on average, consistently. In previous seasons I had ran closer to 60 to 100 games on that ALT, yielding the same exact results as what I've posted in here about season 20.
Sind didn't prove anything right or wrong. All he did was prove that he wasn't experiencing some kind of either bug, or massive permanent dishonor, or whatever it is that's clearly going on here, as indicated by the sheer numbers I am providing you. Some people I've talked to say their accounts all share similar match quality, whilst others report the same thing that I am, that they have experienced one particular account which seems to be clearly "marked" and never under any circumstances, plays nearly as high as the other accounts they own.
It's not a conspiracy when number patterns are quite blatantly indicating differences in patterns for very very long durations of time, during several different test samples over several different seasons, and staying consistent in their patterns at that. What we're talking about here is called "results" or "evidence" or the "review of cause & effect" lol, not a conspiracy theory. When you have the kinds of results that I am posting here in this thread, which I will continue to do, that's plausible evidence that's needs to be considered & reviewed.

I think it will be interesting to see how long some of you will continue to be driven by general forum & personal subjective bias, and keep attempting to tell me that none of the numbers I am posting are real, all the while looking for strange excuses to throw at these posted results, as if they were any more plausible or sensical than my OP claim that, "Something is wrong with my main's match making."

@Bazooka.3590 I don't understand exactly what you meant by rows, but I am trying to. Thanks for the input. If you read this and wish to elaborate, that'd be great.

Poisonwell.1209 · February 24, 2020

@Ghetx.1752 said:
@bluri.2653 said:Ah yes the daily conspiracy theories.
90% winrate is impossible!My alt wins more than my main!50% forced w/r!
Its almost comical how much excuses you guys go through

90% winrate is possible because those players never face other 90% winrate players.

MMR system is made that way to put high rated players on same team,and lowest rated people on another team.

Instead of Legendary players fighting each other,they are fighting low gold.

I always have the same feeling about against who leg players fight. I wish there no me more duo in ranking. Also no way to change class in a match. Im very sure this top 25 will never be leg again.

apharma.3741 · February 24, 2020

I get the feeling we're missing the very obvious here.

phokus.8934 · February 25, 2020

@Trevor Boyer.6524 said:
@bluri.2653 said:
@Trevor Boyer.6524 said:
@"phokus.8934" said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu

130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

Eh?

All that really proves is that you in particular are not experiencing what I am talking about. However, other players have stated that they have.

@"phokus.8934" said:
A few things:1) 44 games in a season is not a lot to settle the rating volatility/deviation; it's just mathematically not accurate to say otherwise. Previous seasons games played is irrelevant when a new season starts. Soft resets occur using this algorithm: (Rating last season played + 1200)/2. Take into consideration the pvp population, their relative rating and the matchmaker building around a player. But consider this: Glicko-2 is used for 1v1s so we can safely assume that each team is treated as one player so the matchmaker builds as close a match as possible based on average rating, classes, duo queue, and time in queue.2) Your proof means jack if you don't have any of it controlled and corroborated by others. Hell, Sind just proved you wrong by indicating his alt accounts are well within his mains rating. Even when I watch Naru stream on his main and alt he's within his mains rating.3) No one's calling you a liar but you're in the deep end with legitimate conspiracy shenanigans. The 4 lists you mentioned in a previous comment have no bearing on anything regarding match quality. You just assume it has to be one of those because you're choosing to ignore the facts and any sense of reason. So yeah, you're the one with the crazy conspiracy theories as to why something is off when the simplest answer is the sad state of pvp population.

44 games in a single season is plenty large enough of a sample size to view. At the end of the first day after just the first 15 games, each season after season, we see the same exact players taking the same exact positions and maintaining them for the duration of the season. You know it and I know and everyone knows that is true. Besides that, we aren't discussing just season 20 here. We are also discussing why this same thing has happened for about the past 4 or 5 seasons, which is: The ALT playing 100 to 200 rating higher than the main on average, consistently. In previous seasons I had ran closer to 60 to 100 games on that ALT, yielding the same exact results as what I've posted in here about season 20.
Sind didn't prove anything right or wrong. All he did was prove that he wasn't experiencing some kind of either bug, or massive permanent dishonor, or whatever it is that's clearly going on here, as indicated by the sheer numbers I am providing you. Some people I've talked to say their accounts all share similar match quality, whilst others report the same thing that I am, that they have experienced one particular account which seems to be clearly "marked" and never under any circumstances, plays nearly as high as the other accounts they own.
It's not a conspiracy when number patterns are quite blatantly indicating differences in patterns for very very long durations of time, during several different test samples over several different seasons, and staying consistent in their patterns at that. What we're talking about here is called
"results" or "evidence" or the "review of cause & effect"
lol, not a conspiracy theory. When you have the kinds of results that I am posting here in this thread, which I will continue to do, that's plausible evidence that's needs to be considered & reviewed.

I think it will be interesting to see how long some of you will continue to be driven by general forum & personal subjective bias, and keep attempting to tell me that none of the numbers I am posting are real, all the while looking for strange excuses to throw at these posted results, as if they were any more plausible or sensical than my OP claim that, "Something is wrong with my main's match making."

@Bazooka.3590 I don't understand exactly what you meant by rows, but I am trying to. Thanks for the input. If you read this and wish to elaborate, that'd be great.

Whatever you say, broski. Start playing with a full deck of cards and you'll see that the world is not out to get you or singling you out.

Crab Fear.1624 · February 25, 2020

@phokus.8934 said:
@Trevor Boyer.6524 said:
@bluri.2653 said:
@Trevor Boyer.6524 said:
@phokus.8934 said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu

130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

Eh?

All that really proves is that you in particular are not experiencing what I am talking about. However, other players have stated that they have.

@phokus.8934 said:
A few things:1) 44 games in a season is not a lot to settle the rating volatility/deviation; it's just mathematically not accurate to say otherwise. Previous seasons games played is irrelevant when a new season starts. Soft resets occur using this algorithm: (Rating last season played + 1200)/2. Take into consideration the pvp population, their relative rating and the matchmaker building around a player. But consider this: Glicko-2 is used for 1v1s so we can safely assume that each team is treated as one player so the matchmaker builds as close a match as possible based on average rating, classes, duo queue, and time in queue.2) Your proof means jack if you don't have any of it controlled and corroborated by others. Hell, Sind just proved you wrong by indicating his alt accounts are well within his mains rating. Even when I watch Naru stream on his main and alt he's within his mains rating.3) No one's calling you a liar but you're in the deep end with legitimate conspiracy shenanigans. The 4 lists you mentioned in a previous comment have no bearing on anything regarding match quality. You just assume it has to be one of those because you're choosing to ignore the facts and any sense of reason. So yeah, you're the one with the crazy conspiracy theories as to why something is off when the simplest answer is the sad state of pvp population.

44 games in a single season is plenty large enough of a sample size to view. At the end of the first day after just the first 15 games, each season after season, we see the same exact players taking the same exact positions and maintaining them for the duration of the season. You know it and I know and everyone knows that is true. Besides that, we aren't discussing just season 20 here. We are also discussing why this same thing has happened for about the past 4 or 5 seasons, which is: The ALT playing 100 to 200 rating higher than the main on average, consistently. In previous seasons I had ran closer to 60 to 100 games on that ALT, yielding the same exact results as what I've posted in here about season 20.
Sind didn't prove anything right or wrong. All he did was prove that he wasn't experiencing some kind of either bug, or massive permanent dishonor, or whatever it is that's clearly going on here, as indicated by the sheer numbers I am providing you. Some people I've talked to say their accounts all share similar match quality, whilst others report the same thing that I am, that they have experienced one particular account which seems to be clearly "marked" and never under any circumstances, plays nearly as high as the other accounts they own.
It's not a conspiracy when number patterns are quite blatantly indicating differences in patterns for very very long durations of time, during several different test samples over several different seasons, and staying consistent in their patterns at that. What we're talking about here is called
"results" or "evidence" or the "review of cause & effect"
lol, not a conspiracy theory. When you have the kinds of results that I am posting here in this thread, which I will continue to do, that's plausible evidence that's needs to be considered & reviewed.

I think it will be interesting to see how long some of you will continue to be driven by general forum & personal subjective bias, and keep attempting to tell me that none of the numbers I am posting are real, all the while looking for strange excuses to throw at these posted results, as if they were any more plausible or sensical than my OP claim that, "Something is wrong with my main's match making."

@Bazooka.3590 I don't understand exactly what you meant by rows, but I am trying to. Thanks for the input. If you read this and wish to elaborate, that'd be great.

Whatever you say, broski. Start playing with a full deck of cards and you'll see that the world is not out to get you or singling you out.

I normally get into plat, but I forgot to wear my tin foil hat.

A family member strangely enough made too many frozen burritos during this time.

Trevor Boyer.6524 · February 25, 2020

~ Edit

Weird double post.

Trevor Boyer.6524 · February 25, 2020

@Crab Fear.1624 said:
@phokus.8934 said:
@Trevor Boyer.6524 said:
@bluri.2653 said:
@Trevor Boyer.6524 said:
@phokus.8934 said:It’s literally not any of those things.
There are quite a few variables you’re choosing to ignore so once you stop thinking conspiracy and lunacy then you can have a better understanding on how the matchmaker and Glicko-2 works.
The biggest thing that I see is the variance of games played between your alt and main account. With soft resets occurring at the start of every season, the volatile rating deviation is extremely high - that’s why you get +-30 on wins/losses early on.
I can’t say for certain because I don’t know, only ANet, but their matchmaker appears to have its flaws. This is only based off of the pseudo code and what forum user Ryan found. But nothing that has been described here.

It's easy to toss this kind of a response when you've never experienced a screw job like this. Honestly man, if your account/accounts are feeling like my ALT's match making, I totally understand how what I am saying would sound ridiculous. If our positions were swapped, I'd think you were crazy as well, because I'd assume that "All accounts are being treated equally in match making so this guys just must be bad." But I am telling you that this is not the case and that some accounts are "marked" in some way and are not being treated equally for fair match making. If you were to experience what I am experiencing on the main account, just feel it for about 10 games or so, you would quickly change your opinion of what I am posting here.

And on a side note, this is why I didn't even embark on recording a bunch of videos and editing them on youtube, or trying to stream all of this on Twitch. It would have been a waste of time. This is something that people can't see until they've felt it themselves. I have talked with other people in-game who experience these forced rating resets so to say, like I am. These players, you can visually see in each season, that they are being bounced back & forth across the leaderboards over and over and over, whilst other users I've spoken to only ever get patterns like my ALT, which plays tightly within a 40-60 rating margin plat 2 value. And this is all main accounts I'm talking about here. Something is up with this. What I'm talking about are two strongly identified types of behaviors that the algorithm is tossing accounts:

Type A - A normal pattern like my ALT gets, or what most users typically report their match making to be like, where the difficulty of matches & general match making actually seems to function exactly how it says it should in the patch notes. Players experiencing this will play tightly within a small rating margin and rarely ever go above or bellow it, indicating that volatility has settled and they've found their place in the leaderboards.
Type B - An alternative pattern where, there is some certain integer that the player is not allowed to go beyond or even TOUCH. If the player plays to this integer, say 1500 rating, it's as if an automated Eye Of Sauron switch is flipped within the algorithm and it "Sees You" and then makes sure the player is met with some ridiculous 10+ game lose streak to force some kind of a rating reset. And this pattern will occur over and over and over, every season, never breaking the pattern.

Again, I am not lying to you. And I am looking for an actual reasonable plausible answer as to why some people's main accounts, including my own, are always experiencing Type B behavior, never ever ever breaking Type B pattern, but then other accounts are always experiencing Type A behavior, never dipping into Type B at all. In the case of my ALT account, we're talking 400 some ranked games played, and never experiencing Type B behavior not once. Something is going on that either isn't being talked about or has not yet been identified as a problem.

A couple things to keep in mind again, and no I am not lying to you about this:

You mention variance of games played, but those 44 games played are only from season 20. The ALT has 400 some ranked games played total from various seasons. The ALT always plays plat 2 level and never drops bellow 1550 during 400 some games played. The main in the past 5 or 6 seasons, struggles to maintain 1500 bottom plat and often bounces between 1320-1500. Considering that I am not lying to you, that is a very large sample size of games played, that shows the same pattern between each account, happening over and over.
You mention volatility? What? So it sounds to me like you are saying that "Volatility = Randomness" which doesn't make sense considering the aforementioned 1, which clearly described a pattern between the two accounts which can be seen functioning the same way in every season. Again, I am not lying to you. This has nothing to do with "Volatility in resets tossing random match making" when hard patterns of algorithm behavior completely dictate otherwise.

You say "It's not any of those things" so what it is? And don't act like playing between 1560 and 1620, averaging 1580-1600 over the course of 44 games isn't a substantial sample size to view. We all know that in 44 games played, a player can get a plenty realistic idea of their skill level's rating margin. This is why the same players every season, end up in the same places after day one first 15 games. I know that's true, you know that's true, everyone knows that's true. So let's stop with the "OMGERD CONSPIRACY" attitude, which is kind of ridiculous to throw at the evidence I am posting in here.

I have 3 accounts and pretty much all of them have equal winratio(close to). You are proven wrong.

of the alts has never done ranked before this season
Second alt has done ranked from HoT but never finished a season
my main acc

https://imgur.com/a/lWQodbu

130-30 = main96-21 = HoT alt115-26 = fresh acc from this season

Eh?

All that really proves is that you in particular are not experiencing what I am talking about. However, other players have stated that they have.

@phokus.8934 said:
A few things:1) 44 games in a season is not a lot to settle the rating volatility/deviation; it's just mathematically not accurate to say otherwise. Previous seasons games played is irrelevant when a new season starts. Soft resets occur using this algorithm: (Rating last season played + 1200)/2. Take into consideration the pvp population, their relative rating and the matchmaker building around a player. But consider this: Glicko-2 is used for 1v1s so we can safely assume that each team is treated as one player so the matchmaker builds as close a match as possible based on average rating, classes, duo queue, and time in queue.2) Your proof means jack if you don't have any of it controlled and corroborated by others. Hell, Sind just proved you wrong by indicating his alt accounts are well within his mains rating. Even when I watch Naru stream on his main and alt he's within his mains rating.3) No one's calling you a liar but you're in the deep end with legitimate conspiracy shenanigans. The 4 lists you mentioned in a previous comment have no bearing on anything regarding match quality. You just assume it has to be one of those because you're choosing to ignore the facts and any sense of reason. So yeah, you're the one with the crazy conspiracy theories as to why something is off when the simplest answer is the sad state of pvp population.

44 games in a single season is plenty large enough of a sample size to view. At the end of the first day after just the first 15 games, each season after season, we see the same exact players taking the same exact positions and maintaining them for the duration of the season. You know it and I know and everyone knows that is true. Besides that, we aren't discussing just season 20 here. We are also discussing why this same thing has happened for about the past 4 or 5 seasons, which is: The ALT playing 100 to 200 rating higher than the main on average, consistently. In previous seasons I had ran closer to 60 to 100 games on that ALT, yielding the same exact results as what I've posted in here about season 20.
Sind didn't prove anything right or wrong. All he did was prove that he wasn't experiencing some kind of either bug, or massive permanent dishonor, or whatever it is that's clearly going on here, as indicated by the sheer numbers I am providing you. Some people I've talked to say their accounts all share similar match quality, whilst others report the same thing that I am, that they have experienced one particular account which seems to be clearly "marked" and never under any circumstances, plays nearly as high as the other accounts they own.
It's not a conspiracy when number patterns are quite blatantly indicating differences in patterns for very very long durations of time, during several different test samples over several different seasons, and staying consistent in their patterns at that. What we're talking about here is called
"results" or "evidence" or the "review of cause & effect"
lol, not a conspiracy theory. When you have the kinds of results that I am posting here in this thread, which I will continue to do, that's plausible evidence that's needs to be considered & reviewed.

I think it will be interesting to see how long some of you will continue to be driven by general forum & personal subjective bias, and keep attempting to tell me that none of the numbers I am posting are real, all the while looking for strange excuses to throw at these posted results, as if they were any more plausible or sensical than my OP claim that, "Something is wrong with my main's match making."

@"Bazooka.3590" I don't understand exactly what you meant by rows, but I am trying to. Thanks for the input. If you read this and wish to elaborate, that'd be great.

Whatever you say, broski. Start playing with a full deck of cards and you'll see that the world is not out to get you or singling you out.

I normally get into plat, but I forgot to wear my tin foil hat.

A family member strangely enough made too many frozen burritos during this time.

Oh please

As if "No one in the world were mature or adult enough to give reasonable feedback to assist a gaming company in figuring out problems."

I'm going to go ahead and remind everyone about that other time that I kept harping in the forums that something was off with the match making. Keep in mind that this was not about win trading or match manipulation. No, I had noticed that the algorithm was off, exactly like I am pointing out in this thread now. People kept telling me I was crazy, and that I was a conspiracy theorist, exactly like you guys are doing in this thread here. Finally Arenanet started paying attention. Justin Odell got involved, the guy who worked on the algorithm at the time. Eventually Evan Lesh posted in the thread and let us know that there was indeed a bug in the algorithm, and that it was functioning exactly like I had noticed.

Why don't you go ahead and read entirely through this thread, and click on Evan's link at the end that confirmed everything I was harping on -> https://forum-en.gw2archive.eu/forum/game/pvp/High-MMR-is-punished-for-solo-que/page/1

I was the only person in the forum who had noticed it.

The moral of the story is: "Just because you or a dozen or a hundred people point and call something a conspiracy, doesn't mean it isn't happening."

Crab Fear.1624 · February 25, 2020

@Evan Lesh said:
@Saiyan.1704:That’s typically the problem with the current MMR. Matchmaking isn’t giving enough credit to Player A because MMR is mainly based on the Glicko system – a system based on wins and losses. Player A is actually reducing the score gap but Glicko doesn’t look at that.

Glicko does support score-ratios, but in our tests we found it causes MMR to move so glacially slow that you end up having more bad matches because your MMR can’t keep up with your true skill level.

This is due to the fact that most losing teams still get around 300 points.

We could have a configurable range to rescale the win ratio to help mitigate this (per game mode, unfortunately). We would have to run simulations first, though, because making any changes that affect glicko can be very dangerous.

Do losing teams get 300 points often nowadays.

2015....2020

But to satiate the need, maybe total games is a larger factor in mmr for match making then they admit if your alt gets easier matches.

My Main Acct vs. Alt Acct Ratings Played - What's Going On Here?

Recommended Posts

Link to comment

Share on other sites

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Create an account or sign in to comment

Create an account

Sign in