it’s just one of many factors, accurate information can never be bad
it’s too risky to leave to chance, publish the MMR formula!
Even if I’d recorded the SR changes, you yourself state that decay of 50SR takes 12 games to fix. Add in PBSR and that variance would be undetectable for larger variances. There is a reason why you did it specifically SR3000+ after all.
They’ve left it “to chance” for the past 3 years.
I think we’ll be fine.
2 Likes
I was thinking today, after i won a game where i just kind of guarded the point as Roadhog. I just thought his presence there would be good, and i didn’t chase the enemies around for stats. So, i would not be surprised if my MMR after that game was a bit lower. We won though.
And i guess that might be the danger of showing people’s MMR. People might start judging others, or themselves, based on some extremely specific number that may or may not really MEAN you are better or worse than the others as a player.
MMR is supposed to be stats based? Well, stats don’t mean everything in OW, ALL the time. Sometimes. Just, not every game you need to be a stat monster.
So, i don’t know. On one hand, i hate that we have a hidden rating that determines the matchmaking. It’d be nice to SEE what your “actual” rating is.
On the other hand, i’d like to think that SR is close enough. Seems more based on wins and losses, and since you can win by playing smart, and not putting up massive stats all the time, maybe that’s good.
1 Like
PBSR and other effects are why I need 20 games of SR changes to see the effect. The leaver data that I have was high plat, so PBSR was in effect.
The theory that MMR is based on stats and SR is based on wins is a myth. They are both calculated very similarly, and typically are very close to each other (after rescaling MMR to be in units of SR).
See How Competitive Skill Rating Works - Season 17 → “Summary” and “Popular Myths” → “MMR is determined by statistical performance” for more.
I think it could be cool to see the uncertainty value next to your SR, where the system thinks you might be.
2 Likes
yes however, they have left it to chance where forming teams yourself is not feasible. If they had all this QoL for the LFG, then they would need to make the MMR formula public. QoL for the LFG
How did you take into account the certainty factor with leaver vs streak?
I don’t understand your question.
idk about MMR specifically, but I think people just want to see their actual skill/performance level as it is perceived by the system.
Not a number so heavily based on wins/losses and our ability to play meta heroes. (SR)
Certainty factor of mmr.
Leaver/disconnect/decay would not affect certainty.
Win/loss streak would affect mmr certainty.
I am just wondering, how did you take this consideration into account?
There isn’t really much to “take into account”. If a streak is causing uncertainty to increase, then SR gains will start to go up after the streak has gone on long enough. If you have recorded your SR for enough games, it will be apparent that that is going on.
As you say, leaving and decay do not affect uncertainty.
This is the position I wanted to see if you agreed with. If we both agree on this, then we assume mmr can become uncertain if a player improves in play quality drastically.
As such uncertainty could be affecting SR changes differently to decay/leavers, and mmr would take more games to become certain again at a new value for SR to chase. This would be visible in tracking like you would like. We can somewhat confirm this as it perfectly matches how season 1 movement showed.
Are you following me so far?
The exaggerated SR gains with win streaking is extremely difficult to trigger now, unlike season 1. For a while I thought it was gone completely, but then I gathered some data and saw some believable forum posts. I talk about it (with links to data) in How Competitive Skill Rating Works - Season 17 → Win Streak Bonus / Loss Streak Penalty.
It is difficult to know the exact mechanism that causes SR bonuses with win streaking, but “then we assume mmr can become uncertain if a player improves in play quality drastically” and SR chases the new MMR is a reasonable hypothesis.
This is also matching your “hidden sr” mechanic described in your 3000SR+ decay calculations thread. It feels like the calculations work when certainty has been reached.
From those threads you mentioned, do you feel you have a number of games that can be taken as the general rule for certainty to be reached? (Ie That point where the gains noticably jump?)
I’d say it takes about six games in a row to get clear win streak bonuses (if it is going to happen).
This is why you see so many crazy theories about matchmaking and why people believe in ELO Hell. There is nothing that annoys the human brain more than massive variance.
Ok, let’s go with that assumption. I believe 5 to 6 game win streak (where consistently putting out notably higher performance than current mmr indicates) to be the start as well, with a small 3 to 4 SR boost on the next win growing a little with each subsequent win as long as the player performs consistently.
So the question I always have, is how does it handle someone performing super well by chance for 3 or 4 games in a row?
This would be a sweet spot for mmr to lose certainty, and would put the player in notably higher mmr matches without changing SR gains yet. If this player does not actually belong higher (mmr is an overestimation), then they will most likely lose - but how many awful games until their mmr becomes certain again?
Note though, this is definitely a rare scenario.
Streak bonuses used to a problem, and were toned down (Overwatch Forums). I think the tuning is in a good place now, striking the balance between moving people out that don’t belong there and not throwing people to a new rank by random chance.