Page 3 of 7 FirstFirst
1
2
3
4
5
... LastLast
  1. #41
    Scarab Lord Ealyssa's Avatar
    Join Date
    Jun 2009
    Location
    Switzerland, Geneva
    Posts
    4,140
    Gamers, so dumb they can't understand that Overwatch rating system is NOT a progression system. You aren't supposed to "climb" in the ranks. Placement place will give a rough level and it will then tune it dpending on your performance.

    Is the system 100% perfect ? Nope, it can't. As long as you queue solo you will always be affected by the players you are with.
    But is it good enough to give a quite accurate representation of where you belongs ? Yes, obviously. No matter how salty people having difficulty facing failure may be.
    Quote Originally Posted by primalmatter View Post
    nazi is not the abbreviation of national socialism....
    When googling 4 letters is asking too much fact-checking.

  2. #42
    Play alot and play what your team needs.
    Also climbing as heroes like phara, mercy ( very slow though) and roadhog is easier IMO as they are less dependent on the team.

  3. #43
    Bloodsail Admiral
    Join Date
    Jan 2009
    Location
    AR
    Posts
    1,197
    Quote Originally Posted by Endus View Post
    There's literally no way people could have "proven" this, because they straight-up don't have access to the data. Only Blizzard does. Looking at single-issue factors is deliberately ignoring how the system is stated to work.



    Except that Blizzard has explicitly stated that they take each individual hero's performance into account. They don't compare a Zenyatta's average healing to a Mercy's. They compare your healing on Zenyatta to OTHER Zenyattas.

    Originally Posted by Blizzard Entertainment
    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.
    Winning more than you lose should generally see you gain rank, but personal performance is absolutely and deliberately part of the mix. If you're averaging out at a 50/50 win ratio, personal performance WILL be the factor that determines if you're gaining or losing rank.
    They even said is a minor factor compared to wins/losses though, the only issue here is losing 15 SR when you win 5 games and lose three, hell i'd be fine with that if was even, i'd be fine with MAYBE losing 5 if it was 2:1 ratio, but 15 SR lost over a 5:3 is not something that should happen.

    Also, people play characters differently, so I don't know how they could possibly get accurate statistics off that, unless its just, well if you don't play the same way as everyone else you're screwed!

    I mean, as another mercy example, some, very successful mind you, mercy's dmg boost, some pistol (honestly unless you're boosting a soldier or bastion that's hitting things the pistol does more dmg, but that's another story) how do you rank these two? Which does Blizz decide has more value?

    Some people play(ed) very aggressive ana, do you hurt the people who don't get as many kills?

    What about super agressive zen's vs buff/debuff zens?

    A zarya against a team that doesn't shoot shields vs a zarya agains a team who does?


    There's just too much to factor in for it to have anything more than a minor impact on SR. Honestly I think the biggest factor is what the system thinks you're supposed to be it and will fight you if you get any higher. They just refuse to be transparent about the WHOLE system

  4. #44
    Quote Originally Posted by Onikaroshi View Post
    They even said is a minor factor compared to wins/losses though, the only issue here is losing 15 SR when you win 5 games and lose three, hell i'd be fine with that if was even, i'd be fine with MAYBE losing 5 if it was 2:1 ratio, but 15 SR lost over a 5:3 is not something that should happen.

    Also, people play characters differently, so I don't know how they could possibly get accurate statistics off that, unless its just, well if you don't play the same way as everyone else you're screwed!

    I mean, as another mercy example, some, very successful mind you, mercy's dmg boost, some pistol (honestly unless you're boosting a soldier or bastion that's hitting things the pistol does more dmg, but that's another story) how do you rank these two? Which does Blizz decide has more value?

    Some people play(ed) very aggressive ana, do you hurt the people who don't get as many kills?

    What about super agressive zen's vs buff/debuff zens?

    A zarya against a team that doesn't shoot shields vs a zarya agains a team who does?


    There's just too much to factor in for it to have anything more than a minor impact on SR. Honestly I think the biggest factor is what the system thinks you're supposed to be it and will fight you if you get any higher. They just refuse to be transparent about the WHOLE system
    That's the problem. I CANNOT believe that they can build a system that accurately judges your performance. It's just far too difficult, there are way too many things that can't be quantified. As another example, say the enemy team is about to take a point, but you barely get on and extend it for 5 seconds then die, rezzing by the time they get to the next point. The only stats they see are 5 extra seconds obj time and a death, which is probably going to look like a negative, but you've played correctly.

    It also seems like a pointless exercise because over the long haul your win/loss ratio at a given level absolutely should reflect your performance (assuming random partners). And if you play with the same group, who cares then? Your worse players borrow some rating from the best ones. Whatever.

    They're trying really hard to come up with something that proxies win/loss when just win/loss works better.
    Quote Originally Posted by Rogalicus View Post
    WoD was like ordering a meal, eating it and leaving. Legion is like ordering a meal, instantly puking it back in the platter due to atrocious taste and watching everyone around doing the same, eating their own puke again and again while telling each other 'Wow, there's so much to eat, I can't believe it'.

  5. #45
    Quote Originally Posted by Ealyssa View Post
    Gamers, so dumb they can't understand that Overwatch rating system is NOT a progression system. You aren't supposed to "climb" in the ranks. Placement place will give a rough level and it will then tune it dpending on your performance.

    Is the system 100% perfect ? Nope, it can't. As long as you queue solo you will always be affected by the players you are with.
    But is it good enough to give a quite accurate representation of where you belongs ? Yes, obviously. No matter how salty people having difficulty facing failure may be.
    Yes and no, if your goal is to keep getting better then it's not unfair to expect to keep going up in rank (provided you actually improve).

    Isn't the entire purpose of being competitive (regardless of game) to strive towards learning and improving?

  6. #46
    I play to get a couple hundred rating to 2500 so I can get my 800 competitive points as solo q in the beginning of the season. Doesn't take long.
    Quote Originally Posted by True Anarch View Post
    Never claimed I was a genuis.
    Quote Originally Posted by Furitrix View Post
    I don't give a fuck if cops act shitty towards people, never have.

  7. #47
    Quote Originally Posted by By the Emperor View Post
    Yes and no, if your goal is to keep getting better then it's not unfair to expect to keep going up in rank (provided you actually improve).

    Isn't the entire purpose of being competitive (regardless of game) to strive towards learning and improving?
    True, but a lot of people, especially in games like this, just close their ears and pretend they're always playing well and it's always their team, also selective memory etc.

  8. #48
    Moderator Endus's Avatar
    Join Date
    Feb 2010
    Location
    Ottawa, ON
    Posts
    44,336
    Quote Originally Posted by Onikaroshi View Post
    They even said is a minor factor compared to wins/losses though, the only issue here is losing 15 SR when you win 5 games and lose three, hell i'd be fine with that if was even, i'd be fine with MAYBE losing 5 if it was 2:1 ratio, but 15 SR lost over a 5:3 is not something that should happen.
    You've given no reason why that should be true. Not one. You declare it as if it's some truism, and all it's rooted in is "you don't like it".

    It's a "minor factor" because winning a lot with poor performance will net you more SR than losing a lot with great performance. If you're winning a lot more than you're losing, if your performance is reasonable, you'll win. Personal performance is a lesser factor, but that doesn't make it a non-factor, which seems to be your claim.

    Also, people play characters differently, so I don't know how they could possibly get accurate statistics off that, unless its just, well if you don't play the same way as everyone else you're screwed!
    Not at all. If your way is "better", you'll obviously have better performance, and this scoring will push you higher as a result. If your way is "worse", then it'll hurt you. If it's the same either way, it won't have any effect.

    It isn't a single-factor system.

    I mean, as another mercy example, some, very successful mind you, mercy's dmg boost, some pistol (honestly unless you're boosting a soldier or bastion that's hitting things the pistol does more dmg, but that's another story) how do you rank these two? Which does Blizz decide has more value?

    Some people play(ed) very aggressive ana, do you hurt the people who don't get as many kills?

    What about super agressive zen's vs buff/debuff zens?

    A zarya against a team that doesn't shoot shields vs a zarya agains a team who does?
    The answer to all of those is pretty easy to figure out.

    A Mercy that's able to use her damage boost effectively and her sidearm is getting more kills/assists than Mercies who don't. Which will show up in the stats.
    An aggressive Ana that gets/contributes to more kills is more of a boon to her team, if her healing doesn't take a hit (and if it does, you've got two factors with opposing effects).
    A Zen or Zarya using literally half their kit effectively is, of course, objectively worse than one who uses all of it.

    Again; it isn't a single-factor system, and we have no idea what the weighting might be, but it's an active ongoing process.

    There's just too much to factor in for it to have anything more than a minor impact on SR. Honestly I think the biggest factor is what the system thinks you're supposed to be it and will fight you if you get any higher. They just refuse to be transparent about the WHOLE system
    There is no assumption by the system about what you're "supposed to be", in the way you're talking. There's a system that's set up regarding their confidence in your MMR, but that's reliant on how many games you've played; as you play more, it gets more and more confident as to where your MMR "should be". If you take a break, or have a big win/loss streak, it gets less sure. And this is a factor that applies in BOTH directions; if you're where the system is confident you should be, you won't lose or gain much, either way.

    It isn't randomly picking a number based off year-old data and punishing you for it. Straight-up not happening.

    Quote Originally Posted by Tonus View Post
    That's the problem. I CANNOT believe that they can build a system that accurately judges your performance. It's just far too difficult, there are way too many things that can't be quantified. As another example, say the enemy team is about to take a point, but you barely get on and extend it for 5 seconds then die, rezzing by the time they get to the next point. The only stats they see are 5 extra seconds obj time and a death, which is probably going to look like a negative, but you've played correctly.

    It also seems like a pointless exercise because over the long haul your win/loss ratio at a given level absolutely should reflect your performance (assuming random partners). And if you play with the same group, who cares then? Your worse players borrow some rating from the best ones. Whatever.

    They're trying really hard to come up with something that proxies win/loss when just win/loss works better.
    They're not trying to proxy win/loss, they're trying to improve on win/loss. Win/loss is still a core part of the system; if they wanted to rely solely on that, it would be easy to do so, that's not their goal.

  9. #49
    Epic! Queen Gremlin's Avatar
    Join Date
    May 2011
    Location
    In a Dark place!
    Posts
    1,543
    Solo q is hell on earth between 2300 - 2500 :|

  10. #50
    Bloodsail Admiral
    Join Date
    Jan 2009
    Location
    AR
    Posts
    1,197
    Quote Originally Posted by Endus View Post
    You've given no reason why that should be true. Not one. You declare it as if it's some truism, and all it's rooted in is "you don't like it".

    It's a "minor factor" because winning a lot with poor performance will net you more SR than losing a lot with great performance. If you're winning a lot more than you're losing, if your performance is reasonable, you'll win. Personal performance is a lesser factor, but that doesn't make it a non-factor, which seems to be your claim.



    Not at all. If your way is "better", you'll obviously have better performance, and this scoring will push you higher as a result. If your way is "worse", then it'll hurt you. If it's the same either way, it won't have any effect.

    It isn't a single-factor system.



    The answer to all of those is pretty easy to figure out.

    A Mercy that's able to use her damage boost effectively and her sidearm is getting more kills/assists than Mercies who don't. Which will show up in the stats.
    An aggressive Ana that gets/contributes to more kills is more of a boon to her team, if her healing doesn't take a hit (and if it does, you've got two factors with opposing effects).
    A Zen or Zarya using literally half their kit effectively is, of course, objectively worse than one who uses all of it.

    Again; it isn't a single-factor system, and we have no idea what the weighting might be, but it's an active ongoing process.



    There is no assumption by the system about what you're "supposed to be", in the way you're talking. There's a system that's set up regarding their confidence in your MMR, but that's reliant on how many games you've played; as you play more, it gets more and more confident as to where your MMR "should be". If you take a break, or have a big win/loss streak, it gets less sure. And this is a factor that applies in BOTH directions; if you're where the system is confident you should be, you won't lose or gain much, either way.

    It isn't randomly picking a number based off year-old data and punishing you for it. Straight-up not happening.



    They're not trying to proxy win/loss, they're trying to improve on win/loss. Win/loss is still a core part of the system; if they wanted to rely solely on that, it would be easy to do so, that's not their goal.
    You keep saying the same thing over and over, but there's no proof the system works your way either. Its not transparent, they don't want us to know.

    The competitive forums are FULL of people ranging from low SR to high SR complaining how the SR system is broken. Hell after the Orisa patch supports were gaining less SR on a win and losing more on a loss than before the patch and I'm not entirely sure that is even fixed.

    Oh btw, on the MMR thing, Jeff said himself that the System has an MMR it thinks you belong at and will give you higher SR gains to reach that point from below and higher SR losses to reach it from above. Right from the horses mouth, believe it was at blizzcon.

  11. #51
    Yes it is.

  12. #52
    Quote Originally Posted by Endus View Post
    If you won 5 games, lost 3, and ended up lower than where you started, it's because your performance over those 8 games was well below what they'd expect for someone of your rank. You're ranked too highly, and SHOULD lose rank as a result.

    That's not the system being "broken", that's exactly how it's meant to work.
    Not really. Its always supposed to match you with roughly the same skill level both on your team and the enemy team. I shouldn't gain 10 points on average for a win and lose an average of 20 for a loss. Here and there sure, but not constantly. I played last season and it was a tough grind. I got my golden gun, got to Diamond and said cya. I only log in for events now. Its really not worth it to grind until they balance out the drastic swings in gain / loss ranking wise. If I start at 2450 and go 7-3 and en up 2462, there are horrible issues with the system.

    And please don't sit here and try and say this is an outlier. Its the norm.

  13. #53
    Bloodsail Admiral
    Join Date
    Jan 2009
    Location
    AR
    Posts
    1,197
    Quote Originally Posted by Tonkaden View Post
    Not really. Its always supposed to match you with roughly the same skill level both on your team and the enemy team. I shouldn't gain 10 points on average for a win and lose an average of 20 for a loss. Here and there sure, but not constantly. I played last season and it was a tough grind. I got my golden gun, got to Diamond and said cya. I only log in for events now. Its really not worth it to grind until they balance out the drastic swings in gain / loss ranking wise. If I start at 2450 and go 7-3 and en up 2462, there are horrible issues with the system.

    And please don't sit here and try and say this is an outlier. Its the norm.
    Not to mention "streaks" win a bunch, YAY LOTS OF SR (that you didn't earn) but lose a bunch in a row and you'll find yourself down shit creek without a paddle, or a boat, and your mouth wide open.

  14. #54
    i went from 2200 to 3200 with just solo q. it was terrible and toxic, albeit once you get to high 2800-2900 it gets alittle better but still. Once the season ends and I can get my golden guns, aside from doing the placements and getting the free end-of-season CP from your tier, i dont plan on doing competitive anymore.
    Intel i7 4790k @ 4.4GHz | Asetek 550LC 120MM Watercooler | MSI Z97 PC MATE | XFX Radeon R9 390X 8GB | ADATA 16GB DDR3-2133MHz XPG V3 |
    1TB WD Caviar Blue | 128GB Sandisk Z400S SSD | NZXT Phantom 410 | EVGA 750W SuperNova G1 80+ GOLD | External 2TB WD Elements

  15. #55
    Bloodsail Admiral
    Join Date
    Jan 2009
    Location
    AR
    Posts
    1,197
    Quote Originally Posted by kaamila View Post
    i went from 2200 to 3200 with just solo q. it was terrible and toxic, albeit once you get to high 2800-2900 it gets alittle better but still. Once the season ends and I can get my golden guns, aside from doing the placements and getting the free end-of-season CP from your tier, i dont plan on doing competitive anymore.
    Heard diamond can be pretty bad though, play on meta or die type thing.

  16. #56
    Quote Originally Posted by Endus View Post
    They're not trying to proxy win/loss, they're trying to improve on win/loss. Win/loss is still a core part of the system; if they wanted to rely solely on that, it would be easy to do so, that's not their goal.
    I don't think you can improve on win/loss is my point - win/loss is a perfect representation of your level in my opinion.
    Quote Originally Posted by Rogalicus View Post
    WoD was like ordering a meal, eating it and leaving. Legion is like ordering a meal, instantly puking it back in the platter due to atrocious taste and watching everyone around doing the same, eating their own puke again and again while telling each other 'Wow, there's so much to eat, I can't believe it'.

  17. #57
    Moderator Endus's Avatar
    Join Date
    Feb 2010
    Location
    Ottawa, ON
    Posts
    44,336
    Quote Originally Posted by Onikaroshi View Post
    You keep saying the same thing over and over, but there's no proof the system works your way either. Its not transparent, they don't want us to know.
    Other than, y'know, posts by the development team telling us how it works. Which you're still ignoring, despite me quoting them.

    The competitive forums are FULL of people ranging from low SR to high SR complaining how the SR system is broken. Hell after the Orisa patch supports were gaining less SR on a win and losing more on a loss than before the patch and I'm not entirely sure that is even fixed.
    Those forums are ALWAYS full of people using insufficient data and wild speculation to "prove" anything they want to.

    Oh btw, on the MMR thing, Jeff said himself that the System has an MMR it thinks you belong at and will give you higher SR gains to reach that point from below and higher SR losses to reach it from above. Right from the horses mouth, believe it was at blizzcon.
    And how does it "think you belong at" that given MMR?

    Your performance and history. That's sort of the point.

    Quote Originally Posted by Tonkaden View Post
    Not really. Its always supposed to match you with roughly the same skill level both on your team and the enemy team. I shouldn't gain 10 points on average for a win and lose an average of 20 for a loss. Here and there sure, but not constantly. I played last season and it was a tough grind. I got my golden gun, got to Diamond and said cya. I only log in for events now. Its really not worth it to grind until they balance out the drastic swings in gain / loss ranking wise. If I start at 2450 and go 7-3 and en up 2462, there are horrible issues with the system.

    And please don't sit here and try and say this is an outlier. Its the norm.
    If the system A> matched you with equally-skilled opponents, and B> didn't adjust your SR changes by your personal performance, then gaining SR would literally be a random occurrence, not reflective of any actual improvement in your gameplay.

    The system is meant to not work like that, for pretty clear reasons.

    Quote Originally Posted by Tonus View Post
    I don't think you can improve on win/loss is my point - win/loss is a perfect representation of your level in my opinion.
    If you solely go by win/loss, then it's going to be effectively impossible for anyone to change their rank by playing better. Random chance is going to outweigh the small influence you'll make on your team.

  18. #58
    Quote Originally Posted by Endus View Post
    If you solely go by win/loss, then it's going to be effectively impossible for anyone to change their rank by playing better. Random chance is going to outweigh the small influence you'll make on your team.
    That's just it - random chance will even out over time. So if you're better than everyone else you're playing with, you should win more than you lose.
    Quote Originally Posted by Rogalicus View Post
    WoD was like ordering a meal, eating it and leaving. Legion is like ordering a meal, instantly puking it back in the platter due to atrocious taste and watching everyone around doing the same, eating their own puke again and again while telling each other 'Wow, there's so much to eat, I can't believe it'.

  19. #59
    Quote Originally Posted by Sky High View Post
    solo que left me 60 points above bronze league thanks to my placement matches being shit shows. it's at the point where I dare not do comp in solo from fear of being knocked down there.
    you didnt just afk those like everyone else since they mean nothing?

  20. #60
    Bloodsail Admiral
    Join Date
    Jan 2009
    Location
    AR
    Posts
    1,197
    Quote Originally Posted by Endus View Post
    Other than, y'know, posts by the development team telling us how it works. Which you're still ignoring, despite me quoting them.



    Those forums are ALWAYS full of people using insufficient data and wild speculation to "prove" anything they want to.



    And how does it "think you belong at" that given MMR?

    Your performance and history. That's sort of the point.



    If the system A> matched you with equally-skilled opponents, and B> didn't adjust your SR changes by your personal performance, then gaining SR would literally be a random occurrence, not reflective of any actual improvement in your gameplay.

    The system is meant to not work like that, for pretty clear reasons.



    If you solely go by win/loss, then it's going to be effectively impossible for anyone to change their rank by playing better. Random chance is going to outweigh the small influence you'll make on your team.
    Good lord, the one thing you quoted stated that while it does do some performance checking SR gain/loss was more based off wins, the performance is minor.

    And more on the MMR thing, the system should not actively adjust SR gain/loss based on where it "thinks" you should be. Outside of when it places you lower out of placements and gives you extra to catch back up. It shouldn't be you get above where it thinks and give you less SR on win JUST because your current MMR is above your average MMR or whatever it looks it, nor should it take more SR on loss for the same reason. That's just retarded but basically what Jeff said it does during that panel.

    On top of that streaks are just silly, why do they not want you to play the game if you lose twice? Why do they reward you for getting good teams (or the enemy having crap teams) with SR that you did not earn just because you won a few games in a row?

    There's so many things wrong with the current system it just needs to be scrapped, wiped and based more on a win/loss with personal performance sprinkled in and NO looking at previous matches to see if you get extra/less SR.

    Right now its like.... 1800 team vs 1850 team, 1800 team wins, sweet you're the underdog so you get extra! but your performance was meh so we take some away, oh you're actually a 1720 player? well lets take some more away from that. And suddenly a 30 SR win turns into 12, hell I've seen people get TWO sr on a win, there's even a video floating around where someone lost SR from a win

    https://www.youtube.com/watch?v=-4ijyNZHQTY

    Even though the other team left that is NO reason to LOSE 2 sr, break even maybe, but not lose.

Posting Permissions

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