Overwatch 2 developer blog: Matchmaker and competitive deep dive, part 1

General Discussion
182

It can? I mean, supports aren’t weak. Just not suited and even if they were. The queue times aren’t looking good anyways.

Their pool of players will always be higher on dps, because they’re the most popular role. Giving supports buffs to become popular enough will not be sustainable due hurting their main playerbase. Which 5v5 tries to please, at least right now.

Which reinforces what I stated, 5v5 was a move to make DPS life more bearable and solve the lower tank player count. While supports are the next bottleneck on the equation will eventually lead to 4v4 due the bigger influx of players by the game being F2P.

Similarly mirrors/hero queue could improve the queue times for similar reasons, but at same time the player count for each role will not be balanced anywhere near in the future.

You can’t make supports pleasing if you not hurt damage role in the process and historically damaging damage role didn’t done much on queue times for dps. I mean, those metas you quoted had tanks as bottleneck and severely hindered dps and supports were CC beasts.

Even if you improve a little hurting the damage role and buffing the support role, the problem will be persistent due the ratio of players playing those roles. Eventually making the gap be similar to the old tank problem.

While if you enforce hero queue and mirror matches that can’t have hard counters at least you can have good queue times and could work even if the game becomes 4v4 in the end.

Right now they need to fix problems, that they created, by actually make their step to 5v5 makes sense. Which right now it’s all over the place and I don’t think a simply buff on supports would be enough, not on current ratio on queues on most places and ranks.

I get that having supports more appealing can attract more folks to play, but if this was solveable alone they wouldn’t need to remove one tank in the process either. At same time, right now the game are effectively swapping from 2016 OW to actually enable RQ. After 3 years of RQ.

Matchmaking matching players within certain role against folks at similar MMR were a thing that needed to be adopted on the RQ introduction. Even if they put those ranks you quoted on previous posts. That hurted a lot groups and new players.

The most logical pace would be hero queue, without counters, by enforcing mirrors. Which would make sense on their 3 statements:

  • players playing with 1-2 heroes most of the time (what they introduced on RQ and could fit hero queue)
  • players playing with what they want to play (on higher ranks often folks would play meta anyways and wouldn’t be hit as much)
  • counters becoming less frequent or impactful (which with mirrors and without counters you wouldn’t need to care about it)

At same time would go against:

  • ult charge preservation between heroes (which could work on other modes not the main ones who they stated that folks only plays with 1-2 heroes anyways)
  • swap in the middle of the game (even if they stated that folks often don’t swap they introduced the ult charge)
183

In the past they’ve said they favor 5 stacks vs 5 stacks. I don’t know how that applies to this new system if the 5 stacks have role rank discrepancies. I believe they’ve said their data says that being in a group is a small advantage, but not as much as people think. OW2 makes shot calling a lot easier for pubs.

184

To the extent that Matchmaker/Queue/Support problems lose more players than Support buffs.

That’s not the case.

Also, that kinda assumes making Supports more appealing to DPS/Tank players preferred playstyle, would be universally considered a bad thing to DPS/Tank players.

185

Thank you so much Devs for providing this information. In the future it would be amazing if you could regularly update us with your thoughts on the state of the matchmaker and any stats that you think it is ‘safe’ to release to the public - things like the % of games that are stomps and a breakdown of the reasons these occur from your point of view would be fantastic.

186

The fact that they need “few months” to simply not put Silver and Diamond tanks against each other is a joke.

Just that one change would do wonders. This is not OW1 and there’s no second tank to pick up the slack. Such a huge skill gap in the most impactful role should not be a thing. EVER.

1 Like
187

Nothing but garbage and lies as expected. Every single control map I play is an instant loss right after I get a win. HOW ABOUT DOING SOMETHING USEFUL FOR ONCE AND REMOVE THE ROUNDS FOR THEM? OH BUT GUESS WHAT HAPPENS AFTER I AVOID 3 OF THEM? I EASILY BEAT THEM! NOW WHY IS THAT BLIZZARD? HUH? TELL ME! WHY WERE THEY ON MY SIDE IN THE FIRST PLACE IF NOT TO PURPOSELY MAKE ME LOSE?

3 Likes
188

Hi Colt - I doubt Blizzard will ever admit this because of the risk of ‘stats farming’ but the only way to solve the problem you are describing is to for them to increase your mmr gain if you perform better than the average player at your current rank. This should become easier once they start balancing for player skill across the roles.

When I play my games (particularly the ones where my team is obviously weaker) I’m always trying to make sure that I play in such a way that I’ll maximize my uptime and therefore my stats creation (basically try very hard not to die). An example of this is that when I’m playing tank and have good supports I’ll be very aggressive and try to maximize my elims. When I am not getting the heals I need to do that I’ll swap to a tank that can cope with less heals and focus on doing damage and taking space only when I’m confident I won’t die.

It’s not easy to do from game to game but it’s the only advice I have to give about climbing when a 50:50 win rate is the matchmaker goal.

Cordially,
Ur4nus

1 Like
189

This is why so many people have alt accounts and it’s totally understandable.

I’m decent with a few heroes, but if I wanna try learning a new one I just can’t keep up with my own MMR. The game assumes I can play all heroes at platinum/diamond level and that is far from the truth.

I’m trying to practice Hanzo and I swear I’m Wood Tier material, awful. I think I’m going to get a new account just for Hanzo and maybe some Reinhardt so I don’t have to ruin any more games for my teammates.

190

I think they are referring to Support buffs and reworks.

191

Not if they actually plan correctly and eventually balance the game on where it’s overperforming.

Nerfs on damage by example, are buffs on supports. Similarly to Buffs on supports would imply on nerfs on damage. Instead of introduce more power creep they should tone it down. Powercreep worked prior to RQ, but RQ itself doesn’t work with power creep, as showed on the past 3 years already.

It is the case, sadly it is. They buffed a ton on tanks and yet needed to remove one tank. Their queue times are worst than supports and they need to fit 1/2 of the quota and they’re really strong.

How making the same thing on supports gonna solve? The answer will not, the “bottleneck” was shifted towards supports even requiring 1/2 tanks and buffing them.

Nope, I’m considering that. I’m considering that folks don’t want to play support and neither tanks due damage role being the most popular role and folks wouldn’t leave that even if the other 2 roles are buffed as much.

Instead you should tone down damage role, because would be wiser on the long run. Not because queue times, at same time if you want to folks enjoy tank and supports you make things like junk and pharah not playing against rein and brig. Because is doable and plausible.

When folks feel free to play what they want, they get the fullfiling desire that anything they play will at least be able to play freely.

At same time on higher ranks where META are important, folks eventually would play the “current META”.

Hero queue and mirror matches enables that, at same time will make overwatch more boring for folks who plays with a single hero and more diverse for folks who play with other heroes. Which would lead them to make folks play other heroes without being penalized on their ranks (avoiding smurfing) and enabling them to actually feel motivated to grind BP to unlock the seasonal hero.

Don’t get me wrong, but buffing alone will not solve supports. If was truthful, tanks would have instant queues not supports.

192

Oh now you finally release something when the game is dying . Good riddance. Your new system isn’t gonna fix anything - this game is fundamentally broken

You cannot win 4v5 like you can in other games . When you get 1 troll / incompetent player you basically lose. So making small “changes” won’t do anything .

4 Likes
193

The divide ranks of each ranking category seem to not be working. Maybe make it so matches are only gold 1 thru 5, master 1 thru 5, and etc. within the pools? If it is to fix the que times, this was caused by losing players to bad matches or broken QP/Arcade tank stacking.

194

Yup, one of my first games months ago on comp we had a Zarya just holding W and not waiting for the team to group up. We spammed chat and VC to get the tank’s attention and nothing. We couldn’t get past 20-30 feet from spawn and that was a sign that this game was done.

195

i don’t like this system.

here’s what we know: there are top500 players ranking into plat. those players have a fake plat rank, so the game puts them against other top500 players with fake plat rank. the game will give them extra SR per win until they reach 50% winrate.

but since they’re playing against OTHER top500 players in plat, they will reach 50% winrate in platinum, NOT top500 :joy:

hence the hardstuck grandmasters and masters, stuck in plat, gold, silver, even bronze for those with the bug, who are utterly destroying the matchmaker for the rest of the players.

what chance does someone who was diamond in OW1 but is stuck in gold have to make it back to plat and then diamond, if top500s are stuck in plat?

please refine this system.

2 Likes
196

Now that we know we aren’t = to our SR ranks, why play ranked. Blizz is lying to us about where we stand like streamers have said in the past.

4 Likes
197

This doesn’t end with hardstuck gm and masters players in the metal ranks though. If you have a 50% win rate against gm players, your SR rises until it reaches GM level. And we saw this too. There were top 500 players ranking up with win rates below 50%. If your current SR is, say, plat and you have a 40% win rate (or a 33.3% or whatever win rate) against top 500 players, you will rank up past plat.

People literally demonstrated this during season 1. It looked weird at first, because some people were assuming that these players were losing in plat (or diamond or whatever) lobbies, but they were still ranking up.

It makes sense, though, when you recognize that they were actually playing in top 500 lobbies. Since plat players do not win against top 500 players, it makes sense that the system would rank you up even if you didn’t quite hit a 50% win rate in one cycle.

2 Likes
198

Feels like with the Ramattra drop into Comp Mode it got even worse.

The whole matchmaker thing is so bad, the derank after season ends, the huge skill differences in the teams and the fact that our Rank has nothing to do with our actual skill level, it’s just hilarious.

2 Likes
199

The only problem to really worry about there is if compositions become too durable or too flimsy.

And you can balance around that by adjusting

  1. Tank Survival
  2. Support Firepower
  3. Better out of combat healing, weaker in-combat healing
  4. Everybody gets an out-of-combat regen passive, but Tanks/DPS also need to be near Supports , and weaker in-combat healing

And there’s probably other ways to make it so compositions die faster or slower. Such as targeted buffs to heroes to disable the extra durable comps.

That said, I don’t respect the opinion that moving the overall power level of supports is inherently bad, when it hasn’t been defined since OW2 launched.

200

That’s can be a thing for sure, but also if you want to avoid that you could simply play heroes without that compatibility. Even on cases those compositions become too durable those compositions could be worked as single scenario and adjusted accordingly.

Is better having several micro scenarios manageable than a big one mess one with almost nobody playing certain heroes anyways.

The false sense of diversity that exists right now shouldn’t be a thing to be proud of.

So, I’m gonna split in 4 parts:

If current powercreep on tanks aren’t enough you keep powercreeping? How that solved til now? Roadhog is a clear example of there’s a limit to powercreep and isn’t healthy at all. Diablo3 proved that too much power creep can’t solve several core issues in the game, similarly to the past 3 years after RQ introduction not solving things that worked prior to it. RQ goes against OP stuff countering OP stuff, so the thing is. You need less OP stuff to begin with.

Supports don’t exactly need more dps, the game needs less healing overall and less overall damage. Which would fix the need of their firepower. So, tunning down could solve it. You can increase the lethality without rely on damage increase, if you tweak the variables correctly and normatize certain thresholds which the game like OW already uses in certain ways but not in a smart way.

Supports healing should be tonned down for sure, to tone down the overall damage in the game and eventually lead to less overall need of certain perks on tanks like those increased hp pools. While healing effectiveness can be great outside of the combat can generate more stalemates and barrier comps more often. Which would fall on the flimsy and durable perks.

The issue with out-of-combat regen passive is how the game would effectively consider it. Is a radius? is damage received based? is based on not doing damage? Is the kind of mechannic that sounds great but actually generates more problems than solutions. Leading to the problem of the paragraph above.

While the non-damage taken can work like shields already works. Takes 3 seconds to kick in. Those seconds would generate stalemates and flimsy mechannics if everybody had that. Making even less desireable to play supports due they being more gimmicky and less actually useful.

Supports, are supports not healers. When they use that mentality and focus on utility. They can solve several problems. But at same time rely on CC wouldn’t be a thing due their stance of tanks having CC, which is a mistake tbh. But it’s another path and topic to discuss.

This topic is about matchmaking and competitive. For that Buffing will not solve it, Kiriko was on season 1 and queue times weren’t great. Ramattra is on Season 2 and queue times still aren’t great.

Nerfing overperformers is a good way to change metas, but that needs to be aligned to better structure to support the game to work as they want it to be played. So far, there’s no actual reason for them not make a mode, which could be turned to a main mode, with hero queue and mirroredmatch. If you consider the points they stated in this post and several others.

They look data, they work with that data. There’s no easier path to follow than that kind of path. I’m not saying to remove like they did on the other modes, but actually emphatize this path as experiment and maybe core gameplay. Because right now checks more on their checkboxes than current system does.

201

20m radius near to a Support teammate, through walls.