It illuminates how the devs go about balancing the game and I was really surpised with there extensive use of win-rate statistics.
On the surface this makes sense win-rate should equal hero strength. In reality I don’t think it does. The devs do state how they use unmirrored win-rate and break it down a win/loss fractionally. However, I still don’t think it works as a good stat for actually determing balance on its own.
From their own unmirrored win-rate stats, Symmetra is one of the best heroes in the game. In reality, stating this in-game people would laugh at you. She is absurdly niche and nearly never worth the effort of trying to make her work. OW2 just makes it worse.
In contrast from their own stats, Ana is one of the worst heroes in terms of unmirrored win-rates. Having solid negative win-rate stats which only got worse as the OW2 beta progressed. However, once again in reality if anyone stated that Ana was weak in OW2, they would laugh. She is extremely powerful which becomes even more painfully obvious when looking at any OWL game currently.
So why? We got this elaborate blog breaking down the stats but the stats provided don’t actually justify the balance changes they made. It doesn’t make sense to me. Zen got buffed, cool he felt weak. Devs release blog pre-buff and he is sitting at solid positive win-rate and statistically stronger than most other supports only losing to Lucio and Brig.
Honestly I want an extremely extensive essay in a scientific paper format to breakdown everything. I want all the details! I did appreciate the blog but it just left me wanting more.
I don’t think they use unmirrored winrate alone. With Sym this is where pick rate adds more context. And I’m sure they can break it down even further by map and see she only gets value on certain niche situations.
I agree that is probably how they do it but in the blog they makes out that they don’t really use pick-rate in determing hero strength.
"Using weighted usage rate helps us more accurately portray the relative interest in new heroes or reworks like Orisa and Doomfist compared to heroes like Ana, who is already a popular hero.
When we go into the balancing portion of the blog, literally only win-rate is used but it fails to give us a good understanding of their decisions. It literally makes things even more confusing because the stats they decided to show do not reflect the balancing decisions they made.
All I am saying is the blog lacks context. Just showing us the win-rate stats and expecting us to understand is not helpful. You need to give us the full picture.
It’s because Blizzard has a very poor handling of stats.
For high pickrate heroes, heroes being used in bad situations are weighted heavily by their system. For instance, if Ana is used in 90% of King’s Row matches, if you play Ana you only have a 10% chance for your pick to be included in the unmirrored winrate. Meanwhile if Ana has a 50% chance of being played on KOTH, then your Ana selection has a 50% chance of being included in an unmirrored winrate.
So, basically Ana’s winrate in areas where she’s weaker (like push) was over-represented, whilst her winrate in areas where she’s stronger (like payload) was under-represented.
Likewise, Symmetra is so infrequently picked that her winrate is nearly entirely a representation of how she is in areas that are advantageous to her.
If you want a better representation of hero strength, you really want an unmirrored winrate that is averaged equally across all maps and points. That’s what siege does with their stats, and you typically find that the point-specific but infrequently picked operators (like Mira) have appropriately represented strength, whilst overpowered, overpicked operators (like Finka) are appropriately represented.
I hope that in any upcoming Blizzard articles the developers remind the player base again (i forget when they mentioned this previously), and go more into why some heroes are niche. They have mentioned this once in a while here and there.
Because they way people are confused about Blizzards balancing, whether its pickrates or winrates, they fail to understand hero designs and how they play into balancing.
People still dont understand that heroes that have certain designs that work only in some niche will NEVER have high pickrates. As long as they continue ot have their current designs.
Its always the same thing, people confused why Sym, Junk, Bastion, Torb, etc cant have high pickrates. And why Blizzard doesnt acknowledge that they are weak. And why doesnt Blizzard buff these heroes if pickrates are showing they are hardly picked.
Again asking for buffs so these heroes can be popular and seen in games and for the “generalists” to not be as popular in games , instead replaced with niche heroes.
People with this mindset dont understand the dynamic of generalist designs and heroes designed with certain flaws that makes them niche. And this is all based on Blizzard designing some heroes this way for some strange reason.
For these niche heroes, its not so much about pick rates or win rates, and more so about their designs.
Buffing damage numbers wont fix faulty designs.
EDIT: Also these "niche heroes are not meant to be 1 tricked. They are to be used in places they are strong and then switched off.
I dont want to search for old posts all day but here is one that sort of explains this.
This is concerning hero bans but it can also be applied to people 1 tricking heroes like Sym, Junk, etc and then claiming that are weak because they dont want to switch in areas those heroes are weak in.
100% agree. I really wish they’d publish weekly hero stats in game. Even if they don’t fully elaborate their design decisions it’s far better than being left completely in the dark.
This all makes sense actually. There’s a big difference between strength and versatility. Sym is niche, but she’s powerful within her niche. As Blizz stated, players are more likely to pick her in her niche where they know she has a strong advantage, resulting in a high WR.
However, Blizz’s WR stats are map specific, allowing them to see that while Sym wins often, those wins are concentrated on a small selection of maps. This is how they know that she needs reworks to increase her flexibility, despite that her WR makes her appear already strong to the untrained eye.
Nearly EVERYONE said Ana was weak in OW2. Were you not here for the great support outcry when fans insisted en masse that the entire role was “unplayable” for the entirety of beta? This was so universally accepted that the few of us who claimed otherwise had to preface it with the dreaded “hot take” disclaimer. No, really…
Thankfully Blizz knew that Ana was a strong hero whose poor performance could be chalked up to the growing pains of players adapting to a new format. They had the wisdom not to provide buffs because they knew we’d learn and improve.
I played a lot of brig in beta as it still felt she was strong. This shows she is still probably the queen. I am not sure what they can do to dethrone her.
This was so universally accepted that the few of us who claimed otherwise had to preface it with the dreaded “hot take” disclaimer. No, really…