-
matchmaker rigging / EOMM isn’t real (no duh)
-
Rank is directly 1:1 equivalent to MMR and therefore…
-
demotion protection is entirely cosmetic
-
0 MMR (middle of the curve) is Gold 2 as of the interview
-
50% of narrow comp matches have 51-49 odds
-
95% of narrow comp matches have odds that are 55-45 or better
-
worst case scenario narrow queue odds are 60-40, are incredibly rare for low ranks (metal ranks), and the modifiers will compensate properly
-
modifiers such as “expected” and “consolation” show up at around 52.5-47.5 odds
-
the loss streak prevention system has been completely disabled, as opposed to just being “scaled back” as the patch notes described
-
The matchmaker is calibrated against itsself ensuring that the teams are as likely to win as the matchmaker thinks they are based on previous matches with the same/similar odds
-
The QP matchmaker and wide comp matches do not have nearly as stringent odds requirements as narrow comp matches do due to grouping shenanigans and queue times so if you’re getting absurdly bad games in qp that’s why
-
more than 3 pinned avoids breaks high level matchmaking so that isn’t going to happen
-
matchmaking doesn’t take into account win/loss streaks, but the MMR increase/decrease after a match does
-
SBMMR is incredibly difficult to calculate, hence why it hasn’t been re-added
-
The matchmaker tries to minimize total delta (i.e. the difference between MMR values in a match)
-
Research is being done into what is causing the overabundance of one-sided stomps, but they have ruled out the matchmaker causing it in most cases unless the matchmaker is doing something odd that they “haven’t yet figured out”
Interesting to see numbers on the concept, but I doubt the conspiracy theorists will be satisfied. MMR is easy to see at work for anyone who plays a lot and pays attention, so this shouldn’t be surprising to anyone. The devs aren’t trying to make you lose.
I thought it was interesting and impressed/grateful that they gave their time. There’s not much reasoning with some of the community.
It was a good interview but people won’t listen because they want to believe it’s rigged
The second one is specifically for comp….everywhere else it’s just a hidden value…I’m not a huge fan of this as I think mmr should reflect recent play and not long term play (what SR measured)
And I’m sure everyone will say 60-40 should never happen but he did specify that that is the hard cap (the most extreme it will ever get) ….and even if that was what you did get every round that still means you’re winning 2 out of every 5 matches (assuming you’re the 40 every time) which is not that different from perfectly 50-50
And as should’ve been obvious…there’s very little from a matchmaking perspective that they can do to prevent stomps from happening
They emphasized at the end…any system that would intentionally want to make your experience awful is detrimental and just leads to people not wanting to play…they gain nothing from it
Yeah, I’ve never been able get an answer from the usual suspects on here when I ask why the devs would directly manipulate the outcome of their match. Just them, by the way. They’re the main character and the 9 other people in the match are just there to force their win/loss.
Is it a trend to play with no monitor these days?
Because when my tank is 0-11, while enemy is 29-1, I feel like I gotta start worring about these interviews saying matchmaking is all good
Well, there is some advantage to altering perception somehow.
I actually read something that was kinda interesting.
Definition of “Marketing”:
- “Changing things with Placebo Effects”
And you don’t get that with raw accuracy.
Sure. Smurfs says “hi”
Judge: Did you do it?
Criminal: No. Since I would never do that.
Forum poster: Ok. See he didnt do it. Case closed.
actually matchmaking is trash , i lose more than win , and with little group of friends
“I think they spit in my food every time I order from this place”
Proceeds to eat there 7 days a week
Right, but what wasn’t really addressed is that forced 50/50 does NOT make the game awful. In fact, the whole time they were saying “we want to make the game feel good, we don’t want to make players feel bad,” which is what forced 50/50 actually does. It’s THE WHOLE POINT of forcing players to win games sometimes.
I actually watched the whole interview, and I feel like they just kind of “played dumb,” pretending not to really know what EOMM is. Either that, or somehow they are so incompetent that their job is to design a matchmaker, and they have literally no knowledge of the field.
…you can put 5 players on a team that are all “owed” a win, and simply place them against lower players that are all due for a loss, or just not “owed” a win. It would really not be hard to design a system that does this.
Funny they havent figures this one out when its obvious: 5v5 and OW 2 causes more stomps cause every player is more important, in OW 1 you could win a 5v6 easier than a 4v5 in ow 2
Who is spitting in your food?
I know that he said he would but you should trust the good human nature.
You know what a metaphor is since your comment was one but yea
oh really?!! ah you wanted to give me a wise methaphor. sorry I thought somebody spat in your food
It doesn’t require research to know why stomps happen. It’s a simple matter of people not being equally at every mode/map, with/against every hero, and that with a 5v5 format a single outlier can drastically change the match.
Even if over time someone is Gold 2, they might play Flashpoint like a Silver but Hybrid like a Plat.
Well, kinda.
I actually was suggesting to Morgan to track the perceived “Tilt Level” of the player during their play session. And to cater the matchmaking around giving people easier games when they are tilted.
Bruh, they don’t even know their matchmaking is rigged for 50% winrate but hey, keep searching the “reason” lol 