Reading the recent Weekly Recall: The Balancing Act post confirms what many of us have suspected for a long time—developer balance decisions are overwhelmingly data-driven.
While stats like win rates, pick rates, and hero bans are useful diagnostic tools, they’re not sufficient on their own to reflect the actual quality of matchmaking or in-game performance. The blog highlights how devs acknowledge feedback after numbers raise a flag—but what about the countless matches where poor player placements, low impact performances, or inflated ranks never trigger a statistical anomaly?
Let’s be real: are the devs playing at the mid-to-high ranks where these issues are most apparent? Because here’s a clear example (image above) from a recent game where one player contributed under 10,000 total damage—and yet is ranked alongside teammates delivering double or triple that. What exactly justifies their placement at this level?
The truth is that balance can’t be maintained by spreadsheets alone. When matchmaking consistently places players where they objectively underperform—game after game—it erodes trust in the system.
If you truly value community input alongside data, then examples like this should spark a deeper look into how performance is assessed and how rank is earned.
My question to this 45-55% winrate is: where do niche characters fall in this?
If a niche character and a generalist all fall within the same winrates why pick 1 over the other? Wouldn’t one be the universal choice because of the chasm between some characters within their role?
How are we supposed to know when the only thing you post is a limited amount of statistics. For all we know, that player was a Widowmaker with 9k damage and more kills and final blows than anyone else.
The irony of this topic complaining about conclusions made from data and then using an argument which only includes a very limited amount of data.
only “data” they pay attention to is data from big streamers…this woudl of never happened under jeff kaplan like seriously aaron and the current dev tam clearly dont pay attention to the entire community and dotn care abut what others int he community say to them only data from streamers is valid
To those asking who the player or hero was—honestly, that’s missing the bigger point.
It shouldn’t matter which DPS hero they picked. The game’s matchmaking system placed that player in a role and a rank where it assumes they’re competent enough to perform. That means regardless of the hero, they should be contributing meaningfully—especially when the red team is running a tank like Orisa with no shield, creating open lanes for pressure and ult generation.
In this case, the player was on Soldier: 76. A solid, generalist hero. No niche limitations. No excuses. Yet the impact was minimal—under 10k damage in a full match, with little to no pressure on an unshielded tank. That’s not just a rough game. That’s a mismatch between rank and actual performance.
This is what many of us are frustrated with. The system claims balance based on aggregate win rates, but ignores the day-to-day realities where a player underperforms, contributes less, yet still ends up matched at a level they can’t sustain.
We’re not asking for perfection—we’re asking for matchmaking and MMR systems that hold players to the standard of the rank they’re placed in, not just whether they float around 50% win rate.
And if the community can’t even question that without being told we “lack data,” while the dev team sees everything and gives us fragments, how exactly are we supposed to trust the process?
I love to hate on balance:
5v5 killed any fun I got from tank
Support is way overturned
The damage passive was a bandaid instead of fixing the issues that stemmed from 5v5
But using data makes sense.
Let’s look at a current common example: freja
She is performing very well and dominating lobbies. The developers noticed so then they looked and saw what people were frustrated by: her ability to stay at long distances and still have impact so that’s what they primarily changed.
They noticed a dip and then her resurgence seeing they didn’t do a good enough job.
There are things to complain about. For example the fact that Brigitte literally exists. She was the answer to blizzard not wanting to nerf the cover girl despite her insane dominance at high level. A dominance she has had for the past 9 years pretty much continuously. But using data if done properly is not a bad thing. Especially since it’s not just “oh Genji play lots…Genji broken…delete Genji”
No, no. You don’t get to make a topic talking about how the game is more context than data and then try to omit all other context except the data.
Okay. So what were his Elims? Final blows?
In that heavily-cropped image you showed, the enemy tank has ridiculous amounts of mitigated damage. That means the rest of your team was shooting the Tank all game and thus heavily-inflating damage numbers.
Again, you can’t make an argument with an image THAT heavily-cropped and expect it to actually support your point.
According to Jeff Kaplan, this was the pro players’ fault. He stated that the dev team suggested nerfing Dive heroes, but pro/GM Overwatch players begged them not to nerf them because they took skill and instead introduce a new hero to deal with Dive instead.
According to Jeff Kaplan, it’s Jeff Kaplan’s fault because he is the one listening to biased people and refusing to do his job. He also blamed me for Brigitte too, by the way. I promise you on A-aron’s soul I never asked for Brigitte. And you know the funniest thing about that? I am not even a launch Brig hater!
You’re absolutely right that context matters—but let’s not pretend this post was trying to build an entire case off a single cropped image. The point wasn’t “look at this one stat and panic,” it was: how does a player perform this poorly in a DPS role, on a favorable hero and map, and still get placed at this rank?
To be specific, the player was Soldier: 76 on Junkertown—a map built for hitscan dominance. He had clear sightlines, vertical positioning, and was up against Orisa, a tank with no shield. This is as free as it gets for a Soldier. Yet he ended the match with under 9.7k damage, and no meaningful pressure, zoning, or tempo contribution.
Here’s where it really breaks down: if you actually look at the scoreboard, the enemy team’s Moira—a support hero—out-damaged him. That’s not a nitpick. That’s a red flag. That suggests this player was either having an extraordinarily poor game or, more likely, was misplaced by the system in hopes they’d be carried to a win.
And sure, people are asking “what about their elims, or final blows?” But let’s not forget—in Overwatch 1, damage was the only stat we had to judge DPS impact. And if we’re being honest, even now, damage is still one of the clearest indicators of how much pressure a DPS is applying. You can’t play Soldier on Junkertown into a no-shield comp and get out-damaged by a Moira and claim you belong in that rank.
This isn’t just about that one player. It’s about what this says about the matchmaking system. If win rate is the gold standard for placement, but that standard allows players with this kind of performance to keep queuing into high-stakes roles, something’s broken.
If you’re more interested in defending the scoreboard format than addressing that reality, then maybe ask yourself why we’re seeing this so often in the first place.
Oh interesting, surprised I never heard this before. But that still furthers my point of them not using data and finding an appropriate response with why a hero is over performing
Nobody said the point was “panic”. You’re trying to say that your image “says something” about the state of the game by your own words:
But the stats you showed don’t seem to tell the story you THINK it does.
That’s not difficult for a Moira to do. Moira’s damage is easy to apply and stat pad. That’s why she’s popular in low ranks.
No it wasn’t. Only to lower-ranked players was this ever true. Even in Overwatch 1, there’s been a concept called “trash damage”, which is what happens when a team is dealing a whole lot of damage to an enemy team but nothing is dying. For most DPS, it’s far better to deal Final Blows than any other stat, because that means their damage actually had meaning.
The fact that you’re making this argument, combined with the fact that you admit that the game had a Moira in it (a character rarely seen above metal ranks) tells me that the rank this game took place in was probably relatively low. And that’s further corroborated by the enemy tanks absurdly high mitigation stat. Your team played against a highly-survivable tank and tunnel-visioned HARD on them, and you’re now trying to make the argument that the DPS who probably shot that Tank least was in the wrong.
Given that the other Tank (probably you, since you once proudly boasted of your Tank numbers) has low Mitigation numbers and drastically-inflated damage also tells me that the Tank (again probably you) was trying to Tank v Tank a lot, but I’d have to see damage and final blows stats to verify…but the fact that your supports had insane healing that game also backs up that theory. Your healers (again, one of whom was a Moira) were working overtime to keep their tank alive. But again…getting an accurate story requires more context than you provided.
So that what this is really about? You think SR gains and rank should be determined by farmed stats rather than by winrate?
I’m REALLY doubting they account for attack/defense sides properly. That’s a complicated thing all on its own when people swap heroes mid-match. The data for hero winrates is likely to be… mis-analyzed.
Let’s address the “inflated tank damage” argument for a second. If I’m a tank holding space, landing shots, and my supports are keeping me alive, that’s exactly what’s supposed to happen. That’s a functional frontline doing its job: taking pressure, sustaining, and creating openings for our DPS to follow through.
So the question isn’t “why was the tank doing damage?”—it’s why wasn’t the Soldier doing more? Especially when:
• He was on Soldier: 76
• The map was Junkertown (a hitscan-friendly map)
• The enemy tank was Orisa (no barrier)
• He had a supported front line and plenty of uptime
What held him back? Poor positioning? Tunnel vision? Lack of game sense? Mechanical issues?
Because from where I stand, if a team is enabling space and heals, and the Soldier still can’t apply pressure to a no-shield tank or secure damage down lanes designed for his kit, it’s either lack of skill, lack of awareness, or both.
And again—this isn’t just about one guy having a bad game. It’s about a matchmaking system that looks at a 50% win rate and says, “this player is fine,” without questioning how those wins happened, or whether the player is consistently contributing.
At some point, the system should recognize when a DPS isn’t doing their job—especially in a match where everything was set up for them to succeed. If we can’t ask those questions without getting deflected with “context” every time, we’ll never get closer to a truly competitive experience.
No it isn’t. Since you tend to get triggered when people criticize Mauga in other topics, I’m going to assume you were playing Mauga that game. Which is corroborated by the high damage and high healing by the Tank and the absurd healing numbers by the supports.
Mauga is the king of the trash damage. Mauga is a stat-padder’s wet dream because all he needs to do is stand in front of an enemy tank, shoot them, and pop his cooldowns to feel like he’s having an impact. The problem was, the enemy Tank has high mitigation, which means they weren’t dying…which means Mauga’s damage was getting ZERO value.
I already knew that based on the mitigation numbers. So that further backs up my point. Based on these numbers, it’s likely that you played Mauga into your biggest hard counter, got repeatedly diffed and blamed your DPS. You played the unstoppable force into the immovable object, and the immovable object won.
Again, this is you assuming they were “held back”. The numbers you provided seem to indicate that they just weren’t baited into shooting the unkillable tank.
So again, it seems that your definition of “context” here, is that whoever has the highest stats should be boosted to higher ranks.
Which is very on-brand for what I’m assuming is a Mauga playing in low ranked lobbies.
You’ve made a lot of assumptions about what I played, what rank this is, and even how the match played out—all while completely ignoring the central point: I was doing my job as tank, holding front, applying pressure, and staying alive with support. Meanwhile, our DPS—on Soldier: 76, on Junkertown, into a no-shield Orisa—couldn’t even break 10k damage and got outdamaged by a Moira.
Let that sink in. Moira. A support whose damage is literally designed to be passive and secondary.
Now you’re saying I was stat-padding because I did my job while the Soldier didn’t shoot a target he was built to punish? Orisa’s defensive kit is practically made to counter tank pressure, not long-range hitscan. When she uses Fortify or Energy Javelin, I can’t crit her, can’t heal through damage, and basically become a sponge. That’s the cost of tanking. But Soldier? Soldier is supposed to punish that from a safe distance—and instead, he contributed nothing.
So let’s stop pretending this is about “trash damage” or tank diffing. The fact is: the Soldier had every advantage, and still failed to meet the absolute minimum for the role. And the system placed him there anyway.
That’s the conversation I’m trying to have. If matchmaking can’t recognize when a DPS is being hard-carried while contributing zero pressure—then what exactly are we ranking?
You can keep making it personal. I’ll keep pointing to the actual failure that mattered: a DPS who had no excuse to be that ineffective in that setup. And a system that let it happen.
Assumptions are all you want to be made here, since all you’ve given us is a highly-cropped image. You want people to make assumptions. It’s just that, like I said, the data doesn’t tell the story you think it does. You’re trying VERY hard to paint a specific narrative here, which is immediately discredited by both the image provided and your very specific complaint about the matchmaker.
Again, it’s not hard. Especially not in a game where she basically has to stand right next to her Mauga and heal him like a madwoman. You’re trying to use inflated Moira damage as a smoking gun, and that’s not what it is.
Your job is to be enough of a threat to the enemy team that they have no choice but to concede space to you so that your DPS can get kills. Since you are now responding to my assumptions with “I did my job” rather that providing more context, I’m going to take that as my assumptions about what you did in that game being accurate.
So like I said, your “job” is to be a threat to the enemy team to make them concede space. But that isn’t what happened. Orisa called your bluff and gambled that she could out-tank your damage with help from her healers. And she was right. So in essence, you made very little actual space.
Actually, it could be argued that what you did this game was take negative space because I can already tell by the healing numbers that the enemy supports probably had crazy amounts of Ults all game. So, if we’re looking at the picture provided by the stats and your own account of the match, you spent the entire match shooting at the enemy Orisa, who just refused to die. In the meantime, every Support in the game was healbotting so hard that they barely had time to do anything else all game. In the meantime, your DPS (one of whom was probably a Mei or a Reaper, based on their healing stat) was also shooting the Orisa that wasn’t dying instead of taking out the Supports keeping her alive.
Your Soldier, in the meantime, was probably the only person on your team doing the smart thing and trying to kill the supports on the enemy team keeping Orisa alive. But the enemy DPS (who are under no threat from you, since you’re hard tunnel-visioned on Orisa) were probably shutting him down and making him run away to self-heal (because the Supports were tunnel-visioned on you).
Nothing “personal” here that you didn’t invite. You posted one of your own matches as “evidence” of a bad matchmaker…and you seem to be defining “bad” as in “doesn’t benefit me when I stat pad”.
YOU posted your own match to blame your teammate for why you lost and why the matchmaker is apparently rigged against you. I haven’t attacked anything except those two arguments.
But hey, since you said this was a recent game, maybe you still have the code for the replay? I will gladly watch it just to be proven wrong.