oops we had a visual bug on launch, fixed now though :)
28.10.2025 22:17 β π 2 π 0 π¬ 0 π 0@srslypaladin.bsky.social
Matchmaking enjoyer. Technophilosopher. Server Lead at Blizzard Entertainment working on the Overwatch team. My views are entirely my own.
oops we had a visual bug on launch, fixed now though :)
28.10.2025 22:17 β π 2 π 0 π¬ 0 π 0if you feel like you might have what it takes, but don't check all the requirements, i would still encourage you to apply. if you're unsure, please reach out!
28.10.2025 22:11 β π 5 π 0 π¬ 0 π 0two announcements: first, i'm now a lead :)
second and more importantly: we're hiring! if you want to come help us make Overwatch matchmaking the best it can be, please apply!
careers.blizzard.com/global/en/jo...
the hero with the higher vote count gets banned first. each player's earlier votes are worth more. and for each ban we're only considering the votes for the team whose turn it is to ban (we randomly decide which team goes first and then it proceeds A-B-B-A)
01.10.2025 17:09 β π 4 π 0 π¬ 2 π 0does it work if you select a different value in the "ALL" dropdown
26.09.2025 23:10 β π 1 π 0 π¬ 2 π 0other team bans ball, your team bans mercy, other team bans rein, your team can't ban doom because we only allow 2 bans per role. working as intended! (though perhaps we could make this case clearer in the UI)
26.09.2025 23:08 β π 3 π 0 π¬ 1 π 0under the hood, hero SRs use much of the same MMR system that powers role ranks
03.09.2025 01:23 β π 3 π 0 π¬ 0 π 0currently we're just using win-loss. we had been tracking for a while though :)
03.09.2025 01:20 β π 4 π 0 π¬ 0 π 0we've had some of the systems for tracking per-hero SRs for a while, it's been a fairly long-term goal for us!
the 3 divisions is a typo (its still actually 5), will be fixed soon
hard to say, because it's dependent on second-order effects. i'll see if we can do a blog post with data in the future :)
03.09.2025 01:16 β π 4 π 0 π¬ 0 π 0i actually don't know, that's a question for @captainplanet.bsky.social !
28.08.2025 22:31 β π 0 π 0 π¬ 0 π 0we're investigating, thank you!
28.08.2025 22:30 β π 0 π 0 π¬ 1 π 0also note that we rebalanced ranks at the top end, so we expect a few more players achieving Master/GM/Champion over the next few seasons. don't worry though they're still very exclusive, and i'll still be in Plat
28.08.2025 22:29 β π 15 π 0 π¬ 2 π 0happy season 18 launch everyone! this patch I helped build the Hero Skill Rating system. please share any (positive) feedback you have about it.
28.08.2025 22:27 β π 22 π 1 π¬ 9 π 0a while back we changed it so that at high MMR the pins are not guaranteed, since it was causing some issues for queue times. sorry about that, the avoid still makes them less likely to show up than otherwise :(
04.08.2025 18:56 β π 3 π 0 π¬ 1 π 0Idk I think the meta is actually different at different skill levels. optimal strategies change based on aim/mechanics/coordination
02.08.2025 19:06 β π 3 π 0 π¬ 1 π 0yeah that's a hard one, we want players to group, but it can be rough for the solo. Unlike 2/3s we can't try to pack them with other groups into matches. the good news is that 4 stacks are very rare compared to solos, so this should happen very rarely
01.08.2025 23:45 β π 3 π 0 π¬ 3 π 0new ow blogpost with two of my favorite things: philosophy, and sneak peaks at new stats features
overwatch.blizzard.com/en-us/news/2...
matchmaking queues are markets with time as the currency. dont worry though, at least we'll never impose tariffs!
01.08.2025 23:02 β π 6 π 0 π¬ 2 π 0donkey kong is so viscerally joyful! really loving it so far
25.07.2025 20:56 β π 3 π 0 π¬ 0 π 0would love to, and agree AI could likely do better in principle, just a lot of work to build! (though it seems to be getting easier pretty rapidly...)
23.07.2025 17:24 β π 3 π 0 π¬ 0 π 0yea i strongly believe that fairness is the best way to maximize fun (blogpost covered this in some detail)
we're always looking for new ways to detect matches that will be one-sided. some percentage are likely near-impossible to solve though (everyone's human, people have bad days!)
victory is zero-sum, but fun is positive-sum
23.07.2025 17:12 β π 9 π 0 π¬ 0 π 0in my personal experience (tho i mostly play role queue), i'd estimate 90%+ of my matches are winnable at the start.
22.07.2025 23:53 β π 3 π 0 π¬ 1 π 0hard to say! most of the matches that end up one-sided look pretty balanced to us. some of that could be hidden signals we haven't found. mostly i believe its not that tho, and just due to a combination of Overwatch being very decisive + players playing inconsistently
22.07.2025 23:46 β π 3 π 0 π¬ 2 π 0i believe it's very plausible that it feels like something to be an LLM generating tokens.
22.07.2025 23:38 β π 1 π 0 π¬ 0 π 0im pretty happy with match quality in 6v6, statistically most matches look good
21.07.2025 17:39 β π 2 π 0 π¬ 1 π 0back then we just mutated open queue 5v5 into 6v6 instead of making a new queue, felt less disruptive
21.07.2025 17:36 β π 2 π 0 π¬ 0 π 0for 6v6 we only use one MMR to matchmake with, the one associated with your visible rank.
interestingly enough, we do actually track per role MMRs in 6v6 too, we just don't use them for matchmaking. the single MMR ends up being very similar to the average of the 3 role MMRs.
most of my favorite singleplayer games are serious and most of my favorite multiplayer games are silly
21.07.2025 16:26 β π 4 π 0 π¬ 0 π 0