How does Overbuff work?

General Discussion
1

In ow2 people have an insane hate Bias towards it maybe because people use it to “Win” arguments X person dosen’t agree with, Maybe because they need 100% accurate data.

But people don’t really like “Understand” why it’s used or how to use it.

Your really just looking for trends and in ow2 everything on there has matched whatever Blizzard has said to a T.

“Orisas actually not that good right now and does pretty mediocre”

Checks overbuff: hmmm well this doesn’t agree with my reality so it’s false.

People just hate fighting the hero its annoying and boring.

“Rein is actually over performing in the lower ranks”

Checks overbuff: no… No… These numbers can’t be right why does he have a 52+% win rate?

Blizzard: Genjis actually over performing by like a lot.

Checks overbuff: Okay this is starting to get scary.

Why do they keep giving Zayra these weird and unnecessary buffs?

Checks overbuff: I mean she dosen’t seem like shes been bad just mediocre so I guess they think there “Helping her out”

There’s like a clear cut “Trend” of dva being OP right now so they might do something about that.

Something’s are just obvious “Lifeweaver is trash right now” like you don’t need anyone to tell you that, sometimes the obvious things are just obvious and overbuff has the data for it. Oh Tracer is absolutely dominating the top ranks? What a surprise! What a shock.

Everyone’s just waiting Patiently and not so calmly for overbuff to get something dramatically wrong everyones shooting there shots and hate towards it before a good chance even arises, but as of right now in overwatch 2 the track record is flawless as far as we know.

They haven’t released every statistic for every hero and that’s like the only saving grace to question just how legitimate the site is for the reason people ACTUALLY use the site, i.e. to see what the “Trend” is

But other then that the track record is cleaner the freshly made glass (in overwatch 2), both in terms of what blizzard has told us, and in terms of things that are just obvious, or what’s directly on top 500 which we CAN see give or take slight data fluctuations.

1 Like
2

So in other words if people just wait they’ll have every chance to rub it in how inaccurate and how wrong overbuff can be, then they can mock and belittle people as much as they please…

But right now is like a terrible terrible time.

Another thing to I saw Mauga had a low pick rate but high as hell win rate, but then I thought about it for a second no one usually comes out of spawn on Mauga, I know it could just be a bias thing I know humans like to look for patterns where there aren’t but then I thought about it, every stream I watch and in most of my games that a Mauga appears no one starts on Mauga it’s like he’s only picked in situations he’s good at and when he is picked he DOMINATES those situations.

Like earlier I had a game where my tank stayed Winston after they went Mauga we had ults and everything, we lost and got full held only way to win is if they would of swapped to another hero immediately after they died and just threw away Winstons ult or if me and my co support threw away our ults and went ANA.

Maugas not really a popular hero and every other tank is doing pretty well so it’s like why switch?

But when people do switch to Mauga my good is he unstoppable, it’s like the worst instance of counter watch I’ve ever seen.

And that explains the dudes 61% win rate.

Its like playing current Orisa into an all beam team, but times a million and that person themselves (Mauga) IS the beam team he’s a 1 man army.

1 Like
3

If we’re talking pickrate, it’s based on the fraction of time you played a hero.

If we’re talking winrate, it’s a waste of time, because winrate math is broken for Overbuff and seemingly the devs.

1 Like
4

It’s not just you, the guys waaayy to good at fighting into heros he counters, it be like if Orisa had absolutely no weaknesses against Zarya, only he also counters multiple other heros.

5

The problem with overbuff is that it is cherrypicked data consisting of only those who wans their picks to be noted.

The problem with top500 leaderboards is that… it is just top 500 (and also take long to be truly representative of the meta)

Until blizzard provides official stats at all times, the feel you get in a match is probably more representative than both for your elo.

1 Like
6

The sheer number of people with open profiles is how it works, at best the “Trend” will still become EXTREMALLY visible after a certain amount.

Like I said people need to wait till it’s like “Legitimately” wrong on something then they can discredit it all they like…

It’s just that right now isn’t the best time especially with alllll the matching data they’ve had throughout ow2s current lifespan i.e. for 10 straight seasons.

Yeah you won’t know exactly “How” op or underpowered Reinhardt is.

Butttttt the number of people will give you a clear cut answer on his state in the game at X current season.

For example overbuff works with 40s and 51s blizzard could have so much damage some heros get into the 45 and 60s territory.

But you don’t need that much data to see that there’s a clear cut trend of said hero going towards X direction.

As of right now for the past 10 season it’s proven itself to have PLENTY of data to show the “Trend” of any individual hero

7

But how many bastion, torb, symmetra mains have their profile opened?
And what about all those genji/tracer/widow mercy pocketed smurfs that never have open profiles and their picks are never recorded?

1 Like
8

You shouldn’t need them, as long as the number of people with open profiles tells you what direction that hero is going.

So if you have enough you can say “Hay this heros actually doing really good right now” or maybe this hero needs help.

An individual outlier won’t scew the data to such a ridiculous amount that you override just how much data is available from your available sources.

It need to be like 30/70 or hell 20/80 and even then you can still see what those people prefer and then compare that data to the full data that’s been told to you and you can getter a better grasp on exactly how accurate your data is and just how much of a sample wize your actually working with.

So your 20/80 could actually match the results of 50/50 consistently and at that point you would know that what you have is most likely accurate.

There needs to be such an extreme divation and low sample size in general that you can’t see the general habits of players.

And when you do get information from the offical source it has to be soooooo far off that it’s undeniable, which ow2 has been the exact opposite of.

So you won’t have all the cards, but it’s not like not having all the cards would completely blind you from just the sheer amount of information the data you do have can give you.

I’d say at best… A 0.4% margin of error only hard carried by the fact that a lot of people do have a lot of locked pfps, but even that’s not enough to make “Trends” completely invisible or easily overridden.

Once you have enough data to see the direction something is going in that’s basically a wrap.

9

People with locked pfps again this is just previous data from what Blizzard has told us, but based on everything matching what Blizzard has said locked PFPs don’t deviate to much from open ones.

For example Lifeweaver most people play him the same way, he’s bad right now for most, a healbot Lifeweaver with a locked pfp would still have relatively the same win rate as an open one would.

1.5k damage and 10k healing this is consistent.

Sooooo if you opened up enough Lifeweaver profiles youddddddd basically find the same results anyways. (Unless you accidentally run into a Scarlet Lifeweaver were a completely different ball game).

Butttttt for the most partttttttt abbbsssoollutely nothing would change.

10k healing and 1k damage stilllll at the end of the day leads to a lost for the majority of the people with open profiles, this current consistency wouldn’t change for those with locked profiles either, because there’s like just such a long amount of data on how consistent this is.

You either ran into a scarlet Weaver or like the best healbot since healing in video games was invented.

1 Like
10

I don’t think that is reliable.

The last time when ppl were crying for rein buffs, Blizzard said he is fine so there is definitely biased deviation from the actual stats.

11

I think overbuff is best when looking at anything other then winrate and pickrate.

Want to know the average damage bastion does, or how many times mercy rezes, or how often tracer dies.

That is where overbuff shines imo.

Winrates and pickrates have so many factors that influence it they can easily be skewed.

Take sym for example, she has historically had a high winrate, but that is due to her low pickrate and her being played in situations that favor her or by sym mains.

TLDR, ignore win/pickrates mostly, and use the other data

1 Like