I started playing competitive on PC when the role queue beta launched. Before that I had played on PS4 for a long time. I placed 2250 or something like that as a DPS (which I primarily play). During the beta my SR fluctuated but I never experienced any crazy losing or winning streaks.
During this season (18), I lost 4 out of 5 placement matches and ended up in silver with an SR or 1800 or something like that. A player left in 3 of the 4 losses. No big deal I thought. I’ll just climb back into gold.
I was maining Doomfist and switching to counter the other team’s comp when necessary. I had a ton of fun. My win rate was close to 70% at its peak and I quickly climbed to 1990 SR.
That was a few days ago. Since I started keeping track, I have lost 19 of my last 20 games since then. And I can say with confidence less than 5 of my last 30-40 games (conservative estimate). I’ve read a lot in the forums, youtube and other places about being “tilted”. So I’ve been very conscious of that. I believe I’m doing a good job putting my head down and focusing on improving my play. In other words, I’m trying my best to not let winning and losing affect my play.
But I have to admit, this is getting ridiculous. I read about the MMR system here: Overwatch Forums and it is helpful. Does MMR also apply in competitive? Is there a lag in the time MMR gets updated and factored into matchmaking? The other observation I’ve noticed is that while I was approaching gold, the team SR of my team was usually close to my SR. Now I’m finding that in general the team SR is way below my SR even though I’ve dropped significantly. In other words did MMR overshoot when I was winning and now I’m being punished?
Again, I’m aware that a big part of losing is mental and losing streaks can compound if you play angry. Please don’t flame me with “you’re just tilted” responses. I grant you that may be a part of it but it cannot account for the extreme losing. The majority of the losses are not even close and completely out of my control.
Competitive MMR is separate for each role, and separate from other modes.
No.
SR has no effect on matchmaking. In theory, if your MMR gets significantly separated from your SR, you could see effects where your team SR is always higher/lower than the enemy. I don’t think this happens anymore, though, so unless you wrote down all of your data (your SR, your team SR, enemy SR) it is likely just random fluctuations. Even if it did happen, SR doesn’t matter in matchmaking.
Streaks are just a consequence of mathematics. Tilt can exaggerate them, but not being tilted can’t eliminate them.
Win probability changes slowly with rank because there are so many random factors in each individual match. Unfortunately, it follows from this that frequent and long streaks will occur, and a player’s rank will oscillate widely. Essentially, a player will tend to bounce between the range of where he is nearly guaranteed to win and where he is nearly guaranteed to lose. This problem can be analyzed in depth, mathematically: See Overwatch Forums and https://www.reddit.com/r/OverwatchUniversity/comments/aatezy/why_match_quality_is_frequently_poor/
Finally, you will see some pretty crazy ideas on the forums. They are a trap. If you play better than your rank, and play enough games to overcome random factors, you will rank up.
I fully disagree with the streaks argument. People don’t randomly start playing poorly for 5 games in a row, then playing well for 5 games in a row, then playing poorly 5 games in a row again. That’s not even coincidentally believable when you have so many people experiencing the same thing.
Furthermore, even if you’re tilted, you’re only 17% of the team. Even if you’re only performing at 80% efficiency, it’s a minuscule difference in the grand scheme and not enough to influence several games into being stomps against you (or for you).
You posted fast enough that I doubt that your read the links with the supporting mathematics. But anyways, imagine matchmaking was perfect. That would meant that win probability was always 50%. Now go find a coin, flip it a few hundred times, and write down the results. You will see dramatic streaks.
In truth, though, matchmaking isn’t perfect, because a person can win/lose games due to luck rather than skill, and the matchmaker can’t tell the difference. So a person’s rank will rise/fall though luck, and his win probability will start to fall/rise. Hence, a person will tend to bounce between where they are nearly guaranteed to win, and where they are nearly guaranteed to lose.
I’ll take your word for the analysis and data as you’ve clearly done a ton of research to get to the point of posting what you posted.
Regardless, after three years and now 18 seasons I’m still not sold that this complicated, convoluted system is the best system to use. It seems to be that the ranked ladder in games, even as old as CS:GO, are worlds better than OW. Like not even in the same ballpark…
Without doing hours worth of research (so take this for what it’s worth), I’ve thought many times that the main problem with competitive OW is that there are too few rankings. Again, I’ll use CS:GO as an example, they have twice as many possible competitive ranks, and they don’t display a person’s SR. Rather, at the end of the match you either move up, down, or stay at the same rank (though you usually have a pretty good feel for when you’re getting ready to rank up or derank).
I think the addition of tiers to the competitive system may space out the player base a bit better and may result in more even matches. One thing that I have seen for 18 seasons now is numerous complaints that matches tend to be a stomp one way or the other. That said, maybe more ranks would do absolutely nothing since it seems like people are being grouped with significant differences in SR at this point anyway.
Just a thought - but I just don’t quite understand why Blizzard is living and dying on this MMR/SR system that most people don’t understand and honestly leads competitive to be much more frustrating than it has to.
I could also make another post about why the season system seems completely useless and why people should just be able to steadily climb over time without worrying about losing progress in placement, but that’s for another post.
You’re playing it wrong. According to the devs you’re not supposed to have one account. You’re supposed to buy a boosted account then smurf as you please. You’re a delusional idiot if you think otherwise.
I’ve never played CS:GO, so I can’t comment directly. But note that bronze/silver/gold/… means absolutely nothing as far as matchmaking goes. Your matchmaking is based on MMR, which is a floating point number and uncertainty, per role.
30 coins:
H H T H T H H H T H H H T H T T H H H H T H H H T H T T H H
This is unequal to the standard W W W W W L L L L L W W W W W L L L L L W W W W W pattern than many people experience. Doesn’t seem comparable to a coin flip.
That was worded poorly. I meant to say that I started keeping track and lost 19 of 20. But before that I had already lost the majority of 30-40 games total (but I was only keeping track the last 20). 5 out of 40 is definitely a conservative estimate. I think it is 3 or 4.
Hello my friend, others have given you their opinion on how the matchmaker works so now I am giving my opinion which is based on data I have collected over multiple season and forecasting models I have created within Python and a database I have made from my data.
Also, you can add me as friend on the game and I will tell you more about how to rank up 800SR + using this stinky matchmaker which is not accurate. After you hear me explain this you will have bells go off in your head and will realize what I am saying is correct. But you can make your own decision of course base on your own observation.