I feel like every so often you get at least 1 but sometimes 3-4 games in a row that are just HARD AF that you barely win. I think your MMR may get largely impacted by whether you win these games because it seems like you get a series of either crushing defeats or games where you just ROLL the enemy team afterwards.
The MMR system was previously said to have 15 (iirc) MMR values. I wouldn’t be surprised if they correspond roughly to <500, mid bronze, high bronze, low silver, mid silver, high silver…low GM, mid GM, etc. Then, when you establish this MMR value, you get games designed to push you towards it. You’re currently low diamond but your MMR corresponds to a mid plat player? Down to mid plat you go with hard unwinnable games. You’re currently high diamond but have high masters MMR? Easy win streak headed your way.
It does fit in to the 40-60% win chance range. I mean Jeff himself said that even evenly matched games can feel like stomps because of the snowbally nature of overwatch. That being said, it makes sense that there are certain games in the pool of all games where your performance is the deciding factor as to whether you win or not.
I do have to say recently I’ve noticed something interesting. If I pop off and do extremely well in a game, I more often than not lose. However, if I’m a burger and don’t do anything, I’ll win. I imagine this fits into MMR balancing and the win chance of games but I would expect the opposite scenario too - where I don’t do anything and lose and do everything and win. Either I’m finding a low number of games where my impact matters or maybe I’m just not good enough yet to climb lol.
That’s how they get rid of smurfs yes, but the issue is that it takes forever.
I dont think blizzard would not even need system you talk about in the theory. What is affecting your MMR? Your performance right? If your performance is worse than you are supposed to have based on your SR, why would system should give you harder games? It doesnt need to. If your performance is worse than it should be on your rank, you will obviously lose more games just because of that. Having system to push you down even faster by manipulating games would be kind of overkill dont you think? Imagine system designing these games for all 12 players. If that was reality we would see many more hard stomps.
I am going to try to help you here. This is from my climbing run to plat on support role with mercy. All games, sr gain/loss with game results, do you see any patterns which would support your theory?
I have all games recorded so we can check some if you would like to see them to see how teams dominated the match.
What I tend to see is wins and losses come on batches. Do you happen to have time of day/timeframe for each game? How long total did this take you to climb in real time?
You also tended to stagnate in places, especially at 2200, for quite a long time. when you first started how would you classify your skill rank and what would you classify it now (in terms of what SR you felt you were at the time)
My working theory is that after 3 games you get an “advancement” game, meaning that the system tests your skill once you win or lose 3 games in a row.
[edit:] also, when did the Flex queue start during your data collection?
Zax is a fairly talented person from what I recall so it was probably smurfs-ville but that’s not the point. the data is extremely interesting regardless of if it was smurf or natural progression. >_>
24 leavers in 237 games, total. the leavers impacted 10% of games but less than half of them resulted in an sr impacted game (4.6%). That correlates with my personal experience of leavers as well.
No I was just curious, I didn’t imply that him smurfing would have an impact on the data in general.
And yes you are right, the data collected is very interesting and it really shows how much people over-exaggerate the amount of leavers, smurfs etc.
Btw one thing, i played first like 61 games with 1000 sr bronze DPS player. Who was bronze at time when we started, stuck there i believe.
Usualy I play in EU at 18:00-23:30 , somewhere between that. Or at weekend pretty much at any time after 12:00 (lunch time)
I think that stagnation is caused by difference in players at certain places. Meaning that I require some time to adjust and adapt on higher elo, correct my decisions and gameplay for better enemies and better team mates.
I played mercy in mid-high plat before. I can play her in high plat without issues and keep my rank, but to climb there from 2500 is something totaly different as you have to carry games and climbing in low plat as mercy is hard because of few specific reasons. As you can see in my data, even climbing in gold was hard for me.
If you will tell me when flex queue started I can tell you when i will be home but i believe I didnt play that many support and tank games since it started. I climbed most of this before. I was focusing dps role when flex queue started, so i did just some game for tickets. I have this stats for all roles + open q. Check my profile and my topics, you can get to my 4 climbing topics for all roles, there is even my youtube channel. Every single match in my stats is recorded or streamed. Do you can find there dates too.
You tell me if I was it looks like it only in bronze, my sr gains even in silver werent really impressive. But no, its not natural Progression as I played mercy in platinum in past. But it was still progression in a way as i was trying to adapt on my rank as i climbed. And i need time to get shape with hero i didnt play competively for some time, i am not the youngest
I dont think SR and MMR can be so far from each other unless its decay or leaver effect on the account
Btw sorry for making this topic a lot about me when i posted data, but i though it could be usefull for discussion, i hope you dont mind.
another interesting stat is the (identifiable) smurf statistic. 13 out of 237 games. 5.5%. That sounds pretty accurate to me as well.
The (identifiable) troll stat seems really low at just 4 (1.7%)… I feel like this is limited to toxic chat? Based on my personal experience that seems quite low. I’d be interested by what criteria those were ranked.
I still have to recheck like first 70 games to be sure if i didnt miss abything because i was reckless with collecting more details info from my games.
But i have collection like this for others roles, you can find it in my topics which starts “Zax climbing journey”
But yeah, leavers who will affect negatively your match are not that common.
Again my first 70 games isnt so accurate related to leavers smurfs and throwers, i have to check games again.
I think number of smurfs would be higher as sometimes they are higher to spot on certain heroes. Generaly speaking i typed down Smurfs when i was really suspicious or sure that player do not belong there.
Its not troll stat. T =Thrower.
Its a lot of games also as i explained many times in my topics that I can be biased about certain games where emotions affected my judgement. I usualy dont watch my games back, only in some occasions where i am not sure about something.
I dont want to say that my data arent accurate, just saying there is still a lot of place be more accurate
The general idea of this thread is correct. Playing well or poorly for your rank (relative to other players at your level) prompts Match Making Rating to stack your team, or opposing teams, for subsequent matches. However I don’t think that MMR corresponds to specific ranks the way SR does.
And what happens when your performance is beyond where you should be yet you can’t climb due to all the issues many of us have been talking about since season 2? What happens when you are high plat/low diamond for MANY seasons then all of a sudden can’t get out of gold yet you are always trying to carry the team with golds, 6:1 k:d ratio and higher, etc.?
Not going to list out all the type of players I get most of the time as I have many times but it is funny how people seem to always think it is the player and their bad performance and they belong where they belong which is all 100% false. Sure some do, but plenty also don’t.
I didnt have troubles to climbing to where I belong on my accounts. If you were stable high plat/ low diamond and now cant climb from gold, you have to be finding reasons in your gameplay.
Yeah lol sure, good story, now where is evidence of that? Show me your last 10 games, I am curious to see what you talk about.
