1. #6841
    Quote Originally Posted by Endus View Post
    Given that plenty of people, like myself, do not see the same kinds of results is why I keep suggesting that the variable that's changing (the player) is what makes the difference, here. If it were a problem with the system as a whole, we'd all be seeing that kind of behaviour, and we're not.
    But so far, the majority of the players that post here, are. Anyone who posts about odd losing/winning streaks, as well as people that have widely swinging win/loss ratios are seeing this happen.
    You also keep reenforcing the idea that it's based on what character you are on in a combination of what map and how other people play that character on that map. This is obviously not the case, or else there would be no way to track an SR gain/loss if I swap characters every time I die and I happen to die 6 times each round. If I do that, and we win, and I gain 45 points (which this did happen roughly 3 weeks ago), then where is the computation? Is this to suggest that if I play a hero for 45 seconds to a minute, it's going to factor how I did on that character against everyone that's ever played that character for roughly the same amount of time? Imagine how much more rewarded you are for swapping characters nonstop at that point.
    You have also not been able to point out how I can gain 30+ SR playing as Mercy 1 game with significantly worse play and gain 10-20 the next with much higher stats and playing like I was a diamond player. You haven't been able to explain how I gained 37 points in a game that went 1 round with us on attack and the entire team left before they started their attack round. There's more going on to suggest it's not just a simple equation like you are making it out to be.

  2. #6842
    Moderator Endus's Avatar
    Join Date
    Feb 2010
    Location
    Ottawa, ON
    Posts
    47,526
    Quote Originally Posted by Eapoe View Post
    But so far, the majority of the players that post here, are. Anyone who posts about odd losing/winning streaks, as well as people that have widely swinging win/loss ratios are seeing this happen.
    Streaks and outliers are expected in any system like this. That alone doesn't demonstrate anything.

    And it's hardly surprising that only people who feel the system is unfair make complaints. Nobody makes posts about how fair the system seems to them out of nowhere.

    So no; it isn't remotely "the majority of the player base". Not even a little. If it were as chaotic as you folks claim, people wouldn't be able to climb by performing well. Boosting friends couldn't work. And so forth. The system does work, for most people, which is why there's still an expected bell curve of players in terms of SR distribution.

    You also keep reenforcing the idea that it's based on what character you are on in a combination of what map and how other people play that character on that map. This is obviously not the case, or else there would be no way to track an SR gain/loss if I swap characters every time I die and I happen to die 6 times each round. If I do that, and we win, and I gain 45 points (which this did happen roughly 3 weeks ago), then where is the computation? Is this to suggest that if I play a hero for 45 seconds to a minute, it's going to factor how I did on that character against everyone that's ever played that character for roughly the same amount of time? Imagine how much more rewarded you are for swapping characters nonstop at that point.
    It would assess the time played per character, per map. That's not that hard to figure out. As for you disputing that this is how it works;

    Originally Posted by Blizzard Entertainment
    Fortunately, when we do put you in a match that we know isn't a 50/50, we adjust your SR gain or loss based on your calculated change of winning. So if you did get placed into a match with only a 20% chance to win and then you lose, you shouldn't lose much SR.
    Originally Posted by Blizzard Entertainment
    So then why do points for losses and wins seem so random? Well, the amount of MMR (and SR) you go up or down isn't simply a matter of whether you won or lost, and what was your predicted chance of winning. There's a couple of other things at work. One is the matchmaker's confidence in what your MMR should be. Play a lot of games, it gets more certain. Don't play Overwatch for a while, it gets less certain. You go on a large win or loss streak, it gets less certain. The more certain the matchmaker is about your MMR, the less your MMR will change in either direction based on a win or loss.

    As a minor factor, we also do evaluate how well you played the heroes you used in a match. The comparison is largely based on historical data of people playing a specific hero (not medals, not pure damage done), and we've done a lot of work to this system based on the community's feedback. In fact, I've seen some people indicate that they don't think we're doing this anymore. We still are. While it's a minor factor compared to wins/losses (The best way to increase your SR is still to play together and win as a team!), doing so does help us determine your skill more accurately and faster.
    https://us.battle.net/forums/en/over...9737390#post-3

    Originally Posted by Blizzard Entertainment
    The calculation of your SR adjustment after a match doesn’t look at your teammates, but instead compares you to the performance of other similarly skilled players with that hero across an enormous pool of competitive matches. So, we compare your Genji play to the play of other Genjis, Ana vs. Anas, etc.
    https://us.battle.net/forums/en/over...ic/20754415323

    Originally Posted by Blizzard Entertainment
    In Overwatch, whether your MMR goes up or down is contingent on winning or losing. But there are a number of factors that determine how much that rating goes up or down. For example, what map you’re playing on and whether you were attacking or defending is factored in. We know the win rates on attack/defend on all of the maps and we normalize accordingly. Not all wins and losses are equal. We also look at your individual performance on each of the heroes you played during the match. Everyone has better and worse heroes and we have tons of data showing us what performance levels should be like on those heroes. We also look at your opponents and whether or not their matchmaking rating is higher or lower than yours. These are just a few of the things that are considered when determining how your skill should go up or down.
    https://us.battle.net/forums/en/over...5504371#post-3

    I'm not making any of this stuff up. It's all stuff Blizzard themselves have stated in describing how the SR system functions.

  3. #6843
    Quote Originally Posted by Endus View Post
    Streaks and outliers are expected in any system like this. That alone doesn't demonstrate anything.

    And it's hardly surprising that only people who feel the system is unfair make complaints. Nobody makes posts about how fair the system seems to them out of nowhere.

    So no; it isn't remotely "the majority of the player base". Not even a little. If it were as chaotic as you folks claim, people wouldn't be able to climb by performing well. Boosting friends couldn't work. And so forth. The system does work, for most people, which is why there's still an expected bell curve of players in terms of SR distribution.



    It would assess the time played per character, per map. That's not that hard to figure out. As for you disputing that this is how it works;

    Originally Posted by Blizzard Entertainment
    Fortunately, when we do put you in a match that we know isn't a 50/50, we adjust your SR gain or loss based on your calculated change of winning. So if you did get placed into a match with only a 20% chance to win and then you lose, you shouldn't lose much SR.
    Originally Posted by Blizzard Entertainment
    So then why do points for losses and wins seem so random? Well, the amount of MMR (and SR) you go up or down isn't simply a matter of whether you won or lost, and what was your predicted chance of winning. There's a couple of other things at work. One is the matchmaker's confidence in what your MMR should be. Play a lot of games, it gets more certain. Don't play Overwatch for a while, it gets less certain. You go on a large win or loss streak, it gets less certain. The more certain the matchmaker is about your MMR, the less your MMR will change in either direction based on a win or loss.

    As a minor factor, we also do evaluate how well you played the heroes you used in a match. The comparison is largely based on historical data of people playing a specific hero (not medals, not pure damage done), and we've done a lot of work to this system based on the community's feedback. In fact, I've seen some people indicate that they don't think we're doing this anymore. We still are. While it's a minor factor compared to wins/losses (The best way to increase your SR is still to play together and win as a team!), doing so does help us determine your skill more accurately and faster.
    https://us.battle.net/forums/en/over...9737390#post-3

    Originally Posted by Blizzard Entertainment
    The calculation of your SR adjustment after a match doesn’t look at your teammates, but instead compares you to the performance of other similarly skilled players with that hero across an enormous pool of competitive matches. So, we compare your Genji play to the play of other Genjis, Ana vs. Anas, etc.
    https://us.battle.net/forums/en/over...ic/20754415323

    Originally Posted by Blizzard Entertainment
    In Overwatch, whether your MMR goes up or down is contingent on winning or losing. But there are a number of factors that determine how much that rating goes up or down. For example, what map you’re playing on and whether you were attacking or defending is factored in. We know the win rates on attack/defend on all of the maps and we normalize accordingly. Not all wins and losses are equal. We also look at your individual performance on each of the heroes you played during the match. Everyone has better and worse heroes and we have tons of data showing us what performance levels should be like on those heroes. We also look at your opponents and whether or not their matchmaking rating is higher or lower than yours. These are just a few of the things that are considered when determining how your skill should go up or down.
    https://us.battle.net/forums/en/over...5504371#post-3

    I'm not making any of this stuff up. It's all stuff Blizzard themselves have stated in describing how the SR system functions.
    I never said it was the majority of the playerbase, I stated anyone here who has posted about it has encountered it, and there are quite a few. I've also stated it's not as black and white as you make it out to be, which is clearly evidenced in the last blue post you linked. There's also the post where it states it gets "less certain" based on win/loss streaks, which shows the issue of SR spikes when it comes to gain/loss. Their (Blizzard's) own posts show the contradiction in the rank system based off the info given.
    You also stated 1 factor as to how gain/loss is determined to which I stated there is more to it, and according to those posts you linked, there clearly is. If a system is "uncertain" where you should be placed due to a win/loss streak, there is clearly an issue if said system then gives varying gains/losses based on how you play, clearly showing a big swing in SR per match. This is something I've pointed out (as well as others) to be an issue with how the ranking system works.

  4. #6844
    Wouhou ! Finally got the Survivor achievement with 3 randoms (Mccree, S76, Zen, Torb) ! Now to work on the legendary one. >-<
    If you have to attack a person instead of what they're saying, then you've already lost the argument.

  5. #6845
    Moderator Endus's Avatar
    Join Date
    Feb 2010
    Location
    Ottawa, ON
    Posts
    47,526
    Quote Originally Posted by Eapoe View Post
    I never said it was the majority of the playerbase, I stated anyone here who has posted about it has encountered it, and there are quite a few. I've also stated it's not as black and white as you make it out to be, which is clearly evidenced in the last blue post you linked. There's also the post where it states it gets "less certain" based on win/loss streaks, which shows the issue of SR spikes when it comes to gain/loss. Their (Blizzard's) own posts show the contradiction in the rank system based off the info given.
    The uncertainty in question is what you should reasonably expect in any system with semi-randomized results. Streaks and outliers will happen, that does not demonstrate that there is any flaw.

    You also stated 1 factor as to how gain/loss is determined to which I stated there is more to it, and according to those posts you linked, there clearly is.
    Let's go to the recent post history;

    Quote Originally Posted by Endus View Post
    As for how it determines this, we know the basics. It figures out from your performance overall where it thinks you "should be". The performance in any individual match also factors in. As does the combined SR difference between teams. Factor all those in, and that's what adjusts the win or loss value up or down from the 25 I assume is the baseline (that's what my data tends to balance out around, at least).
    I mention three factors in that paragraph. And didn't even claim they were the only ones, just the basics.

    If a system is "uncertain" where you should be placed due to a win/loss streak, there is clearly an issue if said system then gives varying gains/losses based on how you play, clearly showing a big swing in SR per match. This is something I've pointed out (as well as others) to be an issue with how the ranking system works.
    You're confusing the uncertainty in the internal MMR it has for you, based on your performance and match history, which the per-match adjustments made for performance in any given match, which don't have any stated uncertainty.

    Two separate things.

  6. #6846
    Quote Originally Posted by Endus View Post
    The uncertainty in question is what you should reasonably expect in any system with semi-randomized results. Streaks and outliers will happen, that does not demonstrate that there is any flaw.



    Let's go to the recent post history;



    I mention three factors in that paragraph. And didn't even claim they were the only ones, just the basics.



    You're confusing the uncertainty in the internal MMR it has for you, based on your performance and match history, which the per-match adjustments made for performance in any given match, which don't have any stated uncertainty.

    Two separate things.
    Streaks and outliers in a matchmaking system do, in fact, show flaws with the system.
    Your 3 factors, in my opinion, count as 1 as they all deal with which hero you pick. Slight difference of opinion there and I'll adjust any future responses around it.
    I'm not confusing anything, as read with complete understanding states that the MMR develops uncertainty if you "stay logged off for extended periods of time," and/or "go on consistent win/losing streaks." That shows an inconsistency with MMR, also showing a flaw with it. As a for example, 3 games tonight since I got home, all losses and I went down almost 90 SR. The only differences between my personal performance from this morning was...damn near nothing, yet every game net me about 25-40 SR losses. Going from yesterday when losses were averaging 10-20 points total to today, personal performance hasn't changed. So, somewhere over the last 24 hours the game the MMR decided I was too high and now needed to be dropped down a bit, even though I've been winning 3 out of 5 games, and only awarding me 18 points on my win for the night before I called it quits for the day? The game I won was on Mercy, the character I've played almost nonstop since the changes (yes, people still have me solo heal most games), the games I lost were Mercy (except 1 where I swapped Zen for Ana), and yet the game MMR felt I suddenly deserved 30+ SR losses over those games when I was losing 10-20 for the last week. Oh, and also felt a need to decrease my average win gain. Yeah, seems like a flawless system to me.
    Just a last minute edit, let's address the elephant in the room. The system decides where it thinks I should be and adjusts gains/losses accordingly. That means, for the sake of arguement (like 1 poster just recently), I could technically lose rating based on how the system determines where I should be. In a season where I fell from 2200 at the start to about 2k, and climbed my way up to 2667, just to play 4 games and drop a total of 80 points (3 losses/1 win), where tomorrow I could win 3/lose 1 and be at less than what I was today because the system "feels" I should be where I'm at IS a flawed system. Slowly clawing my way up, just to be violently thrown back down because the game decides I should be there during a 9 hour break in playing the game is messed up. Especially when considering the sudden change in SR difference between wins/losses.
    Last edited by Eapoe; 2017-10-13 at 05:37 AM.

  7. #6847
    Moderator Endus's Avatar
    Join Date
    Feb 2010
    Location
    Ottawa, ON
    Posts
    47,526
    Quote Originally Posted by Eapoe View Post
    Streaks and outliers in a matchmaking system do, in fact, show flaws with the system.
    That's a statement you could only make if you've never studied any kind of statistical theory, ever.

    Because if I'm flipping a coin, that pattern will develop streaks and outliers. It's a factor of randomized outcomes, nothing more.

    I'm not confusing anything, as read with complete understanding states that the MMR develops uncertainty if you "stay logged off for extended periods of time," and/or "go on consistent win/losing streaks." That shows an inconsistency with MMR, also showing a flaw with it.
    That uncertainty is desired, not a flaw. It allows for the system to adapt to a player's change in performance, to more-rapidly push them to the rank they should be playing at. If you log off after pushing hard and practicing every day, and come back three months later with a hand injury, you should probably be dropping in rank really quickly, and the uncertainty component to your MMR allows for that, to more-quickly get you to a rank where you've got a decent shot at winning, rather than your prior rank that you can't keep up at.

    As a for example, 3 games tonight since I got home, all losses and I went down almost 90 SR. The only differences between my personal performance from this morning was...damn near nothing, yet every game net me about 25-40 SR losses. Going from yesterday when losses were averaging 10-20 points total to today, personal performance hasn't changed. So, somewhere over the last 24 hours the game the MMR decided I was too high and now needed to be dropped down a bit, even though I've been winning 3 out of 5 games, and only awarding me 18 points on my win for the night before I called it quits for the day? The game I won was on Mercy, the character I've played almost nonstop since the changes (yes, people still have me solo heal most games), the games I lost were Mercy (except 1 where I swapped Zen for Ana), and yet the game MMR felt I suddenly deserved 30+ SR losses over those games when I was losing 10-20 for the last week. Oh, and also felt a need to decrease my average win gain. Yeah, seems like a flawless system to me.
    I'm not seeing anything at all in what you describe that would be "wrong", to begin with. And given that the only variable in play was your personal performance, I'm gonna call into question your claim that you were playing just as well on both days.

    It's not always a big deal. If the enemy Widow is headshotting you a bunch, your performance is gonna suffer, even if you're doing great otherwise. And it's particularly difficult to claim you were equally capable with all the heroes you play; I'd suggest it's far more likely you're better with some than others, like most people. I play a better Zen than I do a Mercy, and my SR scores reflect that.

  8. #6848
    Did they change Mercy's Guardian Angel(or whatever it's called) or is it bugged? It's like I can't control Mercy the way I used to before her change. Like flying longer than I intended too, which wasn't a problem before.
    Also, how fucking close do I need to be to ress someone? Feel like I have to run up in melee range almost to ress someone.

  9. #6849
    The Insane Jessicka's Avatar
    Join Date
    Feb 2010
    Location
    Manchester
    Posts
    17,428
    Quote Originally Posted by Explicit Teemo Nudes View Post
    Did they change Mercy's Guardian Angel(or whatever it's called) or is it bugged? It's like I can't control Mercy the way I used to before her change. Like flying longer than I intended too, which wasn't a problem before.
    Also, how fucking close do I need to be to ress someone? Feel like I have to run up in melee range almost to ress someone.
    Yes, they ‘fixed some bugs’ with her momentum and flying with GA.

    And yes, you actually have to put yourself in danger to res.

  10. #6850
    Blizzard really needs to add some feature to block players from ending up on your team. 3 games in a row even though I waited 10 minutes between the first and second game and 20 minutes between the second and third game, I get a reinhardt that's just charging in, dies while charging or he charges in and starts swinging his hammer and screams at the team to learn to play.
    Quote Originally Posted by ChampionChains View Post
    I apologize for my mistake, you're a kpop fan which is borderline worse than pedophilia.

  11. #6851
    Quote Originally Posted by Explicit Teemo Nudes View Post
    Did they change Mercy's Guardian Angel(or whatever it's called) or is it bugged? It's like I can't control Mercy the way I used to before her change. Like flying longer than I intended too, which wasn't a problem before.
    Also, how fucking close do I need to be to ress someone? Feel like I have to run up in melee range almost to ress someone.
    The new flight mechanic is you select a target and start your flight and then you can go where you want for the distance of the flight, you are not required to go from point A to B. This way Junk rat can’t expect you to land at B and bomb there.
    If you push a button that finds you a 'random group' and it gives you a random group of people with random skill and random knowledge then you have no right to complain that a 'random group' button did what it was designed to do. The fault lies in your inability to make friends to play with instead of relying on a button designed to be random. It is a 'random group' button, not a 'best of the best' button.

  12. #6852
    The Insane Jessicka's Avatar
    Join Date
    Feb 2010
    Location
    Manchester
    Posts
    17,428
    New season high last night after 5 straight win, so I've climbed over 400SR in a week, and still 260SR off my Career high. Wish I wasn't working today so I could have stayed up to see if the win streak would keep going. 2200 seems my 'average' anyway so I'm also semi-tempted to stop here for the season and just mess about in JR and practice/relearn some of my lesser played heroes.

  13. #6853
    Quote Originally Posted by DeadmanWalking View Post
    The new flight mechanic is you select a target and start your flight and then you can go where you want for the distance of the flight, you are not required to go from point A to B. This way Junk rat can’t expect you to land at B and bomb there.
    No, it's the opposit. Could already do that before, now I don't and just fly a longer distance than I wanted to.

  14. #6854
    Haven't played in some time, decided to give it a shot and get some of the new halloween stuff.

    Didn't take long. Doomfist is retarded OP. Not playing anymore.

  15. #6855
    Quote Originally Posted by Pendulous View Post
    Haven't played in some time, decided to give it a shot and get some of the new halloween stuff.

    Didn't take long. Doomfist is retarded OP. Not playing anymore.
    As someone who plays quite a bit of Junkrat it's my favorite thing to bait Dumbfist into a trap and then blow him up while he sits there completely helpless. Alternatively, Sombra hack makes him a harmless sitting duck.

  16. #6856
    The Insane Jessicka's Avatar
    Join Date
    Feb 2010
    Location
    Manchester
    Posts
    17,428
    Quote Originally Posted by Pendulous View Post
    Haven't played in some time, decided to give it a shot and get some of the new halloween stuff.

    Didn't take long. Doomfist is retarded OP. Not playing anymore.
    DF is considered worse than trash tier. The fact I’ve never lost a game with him on my side seems irrelevant to the actual stats.

  17. #6857
    Quote Originally Posted by Pendulous View Post
    Haven't played in some time, decided to give it a shot and get some of the new halloween stuff.

    Didn't take long. Doomfist is retarded OP. Not playing anymore.
    He was pretty OP for a while but he got nerfed and doesn't really fit in the meta right now at all.

    I wouldn't say he's below trash tier but he's certainly not OP anymore.

  18. #6858
    Can't get out of gold because there are so. many. goddamn. trolls.

    I'm astonished there are this many in the world let alone the game.

    - - - Updated - - -

    Quote Originally Posted by Pendulous View Post
    Haven't played in some time, decided to give it a shot and get some of the new halloween stuff.

    Didn't take long. Doomfist is retarded OP. Not playing anymore.
    I have literally never seen him picked in competitive. Not even by trolls.
    Quote Originally Posted by Tojara View Post
    Look Batman really isn't an accurate source by any means
    Quote Originally Posted by Hooked View Post
    It is a fact, not just something I made up.

  19. #6859
    Quote Originally Posted by Woobels View Post
    He was pretty OP for a while but he got nerfed and doesn't really fit in the meta right now at all.

    I wouldn't say he's below trash tier but he's certainly not OP anymore.
    Of all the characters in the game DF is trash tier just below Mei. I had 2 in a comp game last night and no surprise we lost as they refused to switch. Now mind you in the hands of the right person, yes they can still be good but the general population can’t bring that level of skill.
    If Mei has the ability to freeze in a cone on multiple targets to punish players who deathball she would be useful, but she only freezes one at a time and if there are other enemies near they wipe her out.
    She gives up lethality for survival and since she isn’t a tank she doesn’t really survive past her abilities.

    - - - Updated - - -

    Quote Originally Posted by Mormolyce View Post
    Can't get out of gold because there are so. many. goddamn. trolls.

    I'm astonished there are this many in the world let alone the game.

    - - - Updated - - -



    I have literally never seen him picked in competitive. Not even by trolls.
    There are trolls at every tier, you would be kidding yourself to believe otherwise.
    If you push a button that finds you a 'random group' and it gives you a random group of people with random skill and random knowledge then you have no right to complain that a 'random group' button did what it was designed to do. The fault lies in your inability to make friends to play with instead of relying on a button designed to be random. It is a 'random group' button, not a 'best of the best' button.

  20. #6860
    Quote Originally Posted by DeadmanWalking View Post
    Of all the characters in the game DF is trash tier just below Mei. I had 2 in a comp game last night and no surprise we lost as they refused to switch. Now mind you in the hands of the right person, yes they can still be good but the general population can’t bring that level of skill.
    If Mei has the ability to freeze in a cone on multiple targets to punish players who deathball she would be useful, but she only freezes one at a time and if there are other enemies near they wipe her out.
    She gives up lethality for survival and since she isn’t a tank she doesn’t really survive past her abilities.
    Mei is a great character for holding objectives, but the meta still seems to favor the unfun dive comps atm. Her lethality comes from her ultra accurate right click, left click is more utility. You can easily counter snipe a wm or hanzo with her, or pick off a mercy in the backlines if you have any aim at all.

    Doomfist is just garbage. Amusingly, he's even more garbage if the enemy team has a Mei :P
    Quote Originally Posted by Kaleredar View Post
    Nah nah, see... I live by one simple creed: You might catch more flies with honey, but to catch honeys you gotta be fly.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •