I totally get that some heroes are more self-sufficient than others, and some heroes pretty much need a team to work with them to get value from them.
That said, I suppose my question would be, if you aren’t impacting the games in a fashion that changes the outcome, then why are you convinced you are trapped at the wrong rank? And what prevents you from playing something that would allow you to have more of an impact until you feel you are at a more proper rank?
I’m not so convinced it would be honestly. Take your situation as an example, how much easier do you honestly think it would be to climb if you’re already unable to impact the outcome in what is supposed to be a relatively balanced match? Do you think you would have more impact in a different system, or do you simply think you would be lucky enough to end up higher than you are now due to random chance or pure attrition?
Although this is someone’s take on HotS, the same 50/50 ideas and the problems of a team based game with different classes to choose from largely applies to OW. It was interesting for me, at least, to hear a rather big name give their take on the pros and cons of a 50/50 matchmaking system:
I’ve stated before that I don’t believe that balancing teams will necessarily prevent players from climbing, but that it will make it astronomically more difficult than it should be to climb.
Think on this: I can get any account I touch to Diamond, I have done this with 4-5 different accounts. I can tell you that once I play around 5-6 games on an account that is mid-high gold, the games suddenly become extremely difficult to win (I would hypothesize that the matchmaker has figured out I am playing better so it begins to balance teams accordingly). To elaborate, these mid-high gold matches are just as difficult to win as matches at around 3100. To be even more specific, the players at say, 2400 are not as good at 3100, but the match is just as difficult to win because of the way the teams are balanced. I will destroy people at rank 2400 and yet still barely win, while at rank 3100 I will simply pull my weight and still barely win. I have seen this on countless occasions.
In other words, it seems that wherever you are placed within around 800-1000 SR the matches will be pretty much the same difficulty to win because the matchmaker can offset your skill. Why can GM+ players climb easy? Because the matchmaker can’t reasonably find players to offset their skill.
What this means (to me) goes back to what you and I have discussed some time ago: I believe the matchmakers primary goal is to create “fair” matches, and getting people to their correct and accurate rank takes a back seat. This leads to huge frustration from players who know they should be at a higher rank (and they eventually will get there) yet are having an epic struggle to get there when game after game they are demolishing people yet still barely winning, or barely losing matches. Was it a fair game? Yes. But it shouldn’t have been.
It’s not necessarily rank, as plat-diamond is where I feel I belong, but more the SR level I’m at. I know, going off past seasons, I should be ‘locked’ around 2750-2800 SR, maybe 2850 or so. Currently, though, after starting out the season with a 68% win rate and initially settling onto a rough 50/50 win/loss ratio around 2900, I’ve been going down waaay too far. I’m getting golds and unranked people being placed on either team, some carried by Diamonds (usually 4-stacking). It’s a mess. Has been for the past 20 games or so.
Skill, mostly. I miiiight be able to play Reinhardt, possibly Zen. I don’t know.
You can forget about DPS, though. No way I can play that and climb.
If we were having a simple Elo-ranking, I’d be ranked higher due to my performance being ‘good enough’ to be there. Not sure if I already mentioned it in this thread (could’ve been another), but I feel that I’m being held back by the emphasis that’s being placed on winning, and not my performance. When producing the end of match SR number, the game looks at the outcome first and foremost, and then applies a rather meager ‘performance’ adjustment.
And then when matchmaking, that SR is completely ignored in favor of producing fair matches, based on some hidden MMR and some additional arbitrary criteria, such as being on win/loss streaks and what have you.
To a degree, this makes sense, though. With variances we know that already will exist between MMR, SR, and where a person’s theoretical true skill would be, it could reasonably take a walk in the park and turn it into a slow and, in some cases, painful slog when you hit within 1000 SR of where you should end up.
This is exactly what we’ve talked about before. I think the problem comes into play with tackling the problem of figuring out what someone’s true rank is supposed to be if you aren’t exposing them to 50/50 scenarios. To me, you learn more from players in (what are supposed to be) balanced matches than you do from pure randomness.
I think both ways would result in taking a loooong time to get people to their proper ranks (on average), but one of them involves higher match quality and a more enjoyable experience along the way.
I could be wrong, though. That’s just my take on it.
For active players MMR and SR are basically the same (Overwatch Forums).
“At no point in MMR calculations do we look at your win/loss ratio and win/loss ratio is never used to determine who to match you with or against. We are not trying to drive your win/loss percentage toward a certain number” – Jeff Kaplan (Overwatch Forums)
At present, it’s too skewed towards win/loss, with little actual performance-based adjustment. But you don’t want to go the other extreme, either, where winning or losing doesn’t matter and people pad stats like crazy just to rank up. But a middle ground, 50/50 split between performance and winning would probably work.
If I lose a game, but performed extremely well, I should be ‘rewarded’ by not losing as much SR. Similarly, when I win but perform poorly, I should be ‘punished’ by not gaining as much SR. So instead of throwing the SR system out, make it actually work. Tweak the numbers.
Oh, and get rid of decay, already. It’s literally forcing people to play, instead of naturally pushing them down the order as more and more people start climbing past the idle players.
This is how MMR already works from what we’ve been told. You can lose a game and have your MMR barely move, as an example. The result is that in your next game, you would gain slightly more or lose slightly less SR. Granted, those small variations could potentially translate to like 1 or 2 SR, which is so inconsequential that it would be mostly indistinguishable, but the point remains, that the mechanism you’re describing is already baked in… albeit in a slightly more convoluted manner than you describe.
You would also lose less than normal sub-diamond just as a byproduct of performance based SR under the current model. Performing out of your mind would lose you less SR than doing poorly, as an example. Again… hard to say how dramatic that impact is or what specific metrics govern it… 'cause transparency doesn’t exist.
I think the issue is that some of your players would be high enough that the top 500 leaderboards they shoot for wouldn’t get passed after they met the minimum requirements. The win % it would take to continue to climb to pass them would be an insurmountable hurdle in many cases where people just sit at the top and refuse to play.
Your diamond - masters players are much less of a problem when removing decay than the your leaderboards at the highest portion of GM where rewards are handed out based on where you end the season at.
So, we know that this does happen. Decayed players actually are “500-800SR lower than their true rank and are placed in a match they have a 50% chance of winning”. If they don’t play very many games then they won’t rank up to where they belong.
I don’t think this is what you mean, though I could be wrong.
And I don’t disagree that any general player could be “500-800SR lower than their true rank”. I’ve been there myself.
I just don’t understand how anyone thinks that they could be “placed in a match they have a 50% chance of winning” if the system hasn’t correctly identified their skill by either MMR or SR.
Seriously, please explain on what basis the MM can create these 50% chance matches with an incorrect estimation of someone’s skill (which does happen).
I would honestly think that OW is more complicated from a matchmaking standpoint due to hero switching existing, but I mean yeah, it’s not a perfect translation between the games. Just interesting to hear someone’s take on the principle.
because it’s a 50/50 chance of winning doesn’t mean the winning party won because of random aspects. If you’re good enough you will tilt the balance in such a way that it favours your team and you can win, therefore you can climb. I’ve said it before and I’ll say it again, would you rather have 60/40 chances? What if you were on the receiving end of that 40% how would you feel? Don’t be a special snowflake and ask for matchmaking that only benefits you, 50/50 ensures that the better players can climb and it ensures we have a F A I R match. Stop acting like snowlfakes
There’s nothing fair about matching a GM with a Bronze, just so a bunch of Plats can participate, too. It’s a bit of an extreme example, but that’s kind of what’s happening now.
We’ve been down this road, I think the way MMR is calculated is different than the way you, kaawumba, and friendly fire believe it is calculated. I also think there is more to the matchmaker than what we are being told. I stated in an earlier post :
the matchmaker quickly picks up that I am playing far beyond my current SR and it adjusts. Why would they do this? My guess is to deter/prevent boosting. Sure a Grandmaster can boost people wherever they want, but a plat trying to boost a low gold to plat, or even a high silver to gold wouldn’t be a walk in the park. I have witnessed this myself.
If you’re interested, try some weird experiments like this, very eye-opening experience.
When you’re in a match that you have a 50% chance of winning you cannot “play good enough” to tip the odds in your favor, because the 50% chance you have of winning is already based on how you play when you try hard (assuming you’re not throwing). In other words, the matchmaker has already observed how well you play when you try your hardest and it will expect you to do that just in order to keep the match 50/50.
But let’s just say you did “play good enough” and tilt the balance and win the match; the matchmaker will observe this and arrange your future games based on how well you just played and you will be met with equally difficult opposition in upcoming games and you will be stuck at your rank. So no matter how well you play you are quickly met with equal opposition, just for the sake of creating “fair” matches.
FWIW: I disagree with this at a very fundamental level. 50/50 matchmaking (when everyone is at their proper rank) would be based upon the average of their performances, and not their peak. Someone’s proper rank would not classify as their absolute best, because it is unreasonable to expect their best on every occasion.
That’s basically the point of the entire system, yeah. The 50/50’s aren’t where people really are, though, it’s where the system thinks they should be. So, even before you examine things like how many support mains landed on your team, we have to provide the disclamer that your 50/50 match could be a 45/55 or a 60/40 before it ever begins on some occasions.
(Not trying to argue against you, just piggybacking it to provide my own commentary.)
![:frowning:](https://d74zh8edaijub.cloudfront.net/en/overwatch/images/emoji/twitter/frowning.png?v=5" ":frowning:")
