Maybe it’s people using him as a last resort to their desperate situation and they most likely still end up losing. If he is technically losing so much, why do so many people still play him though? Don’t they all want to win?
Still useful, if you keep some of the assumptions they discussed in the blog post in mind.
1 Like
He’s forgetting one thing though.
You can be right about win-rate not dictating hero performance/weighted (fractional) win-rates being a better measure of performance AND simultaneously be unable to identify what a hero needs to be balanced. And it’s the second part that matters for the end-user and it’s the second part that Blizzard continuously struggles on.
They (sometimes) make the right conclusions with the data, sure, but fail to translate that effectively into character design and that’s what the problem is. See: Brig dominating for months and months through multiple iterations.
1 Like
I wonder if they will ever go into the biasing factor in how strong defensive heroes will result in longer play times, and strong offensive heroes will result in shorter play time.
And how this can skew usage rates, if they are weighted by playtime, to make the winrate of defensive heroes look higher, and the winrate of offensive heroes look lower.
8 Likes
I think people will complain because Blizzard seems to only focus on a few heroes. This is what they have done in the past as well. They still heavily favor a set play style thus ignore other styles unless they start to get played to much thus need nerfs… AKA Symm…
Need proof? They literally posted in first patch notes from April that Junkrat’s trap damage was getting a buff from 80 to 100. This was never add to the beta. His trap damage remained at 80 the entire time. They also ignored all feedback regarding this.
11 Likes
Possibly, if not probably. We can’t really tell from this data. What we can say is that he isn’t super overpowered, and that something is wrong.
Someone’s been speaking to a certain Robot 
?
Hey, I informed him about that skewing factor. That was all me 
I noticed it when Orisa had ungodly high winrates.
And how Snipers/Sniperish heroes all have uniformly low winrates.
Because Stall = Higher winrate. AntiStall = Lower winrate.
2 Likes
Is there a specific reason they do it on a per 10 minute basis? Wouldn’t it be more accurate to just use the average length of a match?
You aren’t getting it. I have explained it multiple times already. When a hero is strong overall, it shows in their pick rates. The converse is true also.
When junkrat was briefly buffed, his popularity surged as he got more viable. When the buffs went away, he went back to his struggling state, and the pickrates followed.
You are labeling certain heroes as “generalists” because they are overly strong to the point they are overrepresented in most situations. If they weren’t so overtuned, they wouldn’t be played all/most of the time, allowing other heroes to see the light of day. Do you get it?
Basically you mixed up the cause and effect. How more clear can I be.
Soldier wasn’t so popular until he got net 6 buffs over his lifetime.
Mei wasn’t so rarely picked until she got multiple nerfs.
20 Likes
Even that’s a problem if it includes both attack and defence. Because the team with the longest defence tends to win.
Especially when there’s the tiebreaker rounds where the likely winner is on defence.
Does this mean Dva will be getting much needed buffs next beta?
If orisa is mentioned as such and Dva is right along side her, can’t imagine why they wouldn’t give her buffs next go around.
6 Likes
They probably will, but the devil’s advocate is that she was underplayed merely because she lacked shiny new toys.
Also because despite her not changing. Offtank players are going to need a mentality shift to play in a more MainTankish playstyle.
Or Roadhog, for instance, is basically still an OffTank in OW2. Having received basically nothing to give him a “Tank presence” when his hook is on cooldown. (Not to mention, Ana basically just makes him a hard throw pick, since he has virtually no counterplay against sleep/antiheal).
Gone
6 Likes
Kinda useless article, the game felt totally unbalanced and I don’t like the fact that they have settled down with all the core ideas for ow2 and they just go for balancing plus some reworks.
The sample has probably too much noise from low ranks, people that are playing totally casually those that don’t even have ow1, and still there is a large range between the winrates.
2 Likes
You are confusing OP with JR just magically being more popular.
With how most of these niche heroes are designed you cant just increase damage because that makes them OP, because of their design.
The JR you are talking about had larger projectiles and a 100 MPH rip tire that flew in the sky when it jumped off walls.
You cant just have Torb fire as fast as 76 to make him less niche. Ignoring that his projectiles have larger hitboxes, ignoring that he has a turret to fall back on, ignoring that his bullets do more damage and no fall off than 76.
You cant just increase damage numbers on certain heroes as they would be OP based on their design.
Any small increase to Reaper has made him OP. Weather its his damage, spread or life steal.
Why would they massively buff Sym, when she is performing great on maps she is strong at according to their stats? Why would they prioritize her if they dont see her needing any huge buffs other than maybe reworking her. As she is performing well in situations she was designed to be good at.
But all this is meaningless , what you should take into account is what the developers themselves have said regarding these niche heroes. Maybe they are not as straight out as i am, but what i say is basically in their articles and interviews. Even if you have to read between the lines most of the time.
And i would love them to make these heroes more appealing to play and better in more situations.
The bigger issue here is the people that 1 trick these heroes wanting them to stay the same but just with increased damage numbers like if that will fix everything.
1 Like
Dunno about you, but this is my jam.
I’ve been speculating on the mechanics behind stats for years. So it’s great to hear about the actual methodology they use.
But then again, most people won’t understand that a raw winrate like on Overbuff has some heavy skewing factors on it.
5 Likes
I mean pretty much yeah, sometimes the devs just want a character not to excel too much in one given field, especially when that excelling can massively hinder their ability to buff the character should they need it
Of course, that’s definitely the only reason she’s played so much, there is never a trend with how often you see a character and hod good they are
lets not be sarcastic on this, and act like we don’t make excuses for rein or mercy’s popularity when they’re consistently top 3 overall as well, you cant even say she’s way overrepresented in gm because while she is the top picked, all supports have a good pickrate bar moira, and there’s only a 2% gap between her and the 6th most picked support
she’s also consistently been absent as a dominant force in OWL games until the ow2 beta, where surprise they’ve started using her a lot and its mainly because the meta is using a winston who needs a long range healer that can consistently keep him up on health, and tanks in general needing a lot of healing which ana can provide with one of the highest hps in the game while still having utility unlike moira
ana doesn’t even have a positive unmirrored winrate in the stats they show and by the end of the beta she had the second lowest unmirrored winrate despite her wide usage, whenever she isn’t mirrored she’s losing more than she’s winning, there reason they came to this conclusion is obvious and if she was really overperforming so much the trove of data from pick to winrate data would show it easily, but it doesnt
2 Likes