There’s no reason to believe that private profiles would consist of data that is significantly different than public profiles. For example, smurf accounts are really not going to make up a significant portion of the data. Despite what the forum says, games do not have 3-4 smurfs per game. There’s also no reason to believe that specific mains or one tricks overwhelmingly hide their profiles more than others.
While I’m sure it does have a small influence/bias and Overbuff is by no means perfectly accurate, it should still be well within reason.
“Sombra is Meta”
“Sombra is in every game”
“Sombra is OP”
The only reason the haters aren’t worse than they are is that there are actual numbers showing she’s barely in the game when she’s claimed to be in every one. That’s a complete 180 from each other, which would you believe?
When? Because the few times they did release numbers, it matched overbuff.
And the more recent hint was when people were saying that Sombra’s incredibly powerful and should be nerfed, COMPLETELY dismissing what’s on overbuff. The dev’s then said not even GM’s were making her work, validating what was on Overbuff.
the two that we throw around on here (because of that page) are pretty terrible metrics though…its a kids finger painting…not a van gogh (but we treat it like the latter)
like yeah…they give you general trends about whats being played globally…IN GAME they mean like next to nothing…and even in terms of balance they really dont say much either…and thats when theyre used properly with like proper context (hint - most of the time theyre not)
better stats would be a godsend for overwatch
and the anecdotal experience thing ties into this…
its frustrating to see so many people get shot down when they give you their first hand experience…like if someone comes in and says “hey i have a 65% success rate on bastion…think people are not playing him properly and/or are selling him short”…usually the response is “good for you (but he sucks)” instead of “tell us more…perhaps you we can learn something”
god forbid a distinct voice exist in the echo chamber
we allow these numbers to be the end all be all of the state of the game discussion when at the end of the day all that really matters is the actual experience you have when playing…YOUR personal stats…like who cares if everyone thinks sombra suks if youre a damn good sombra (and back it up…not just say it)…
Anecdotal evidence is much easier refuted. Someone’s anecdotal evidence can be completely undone by someone else’s if that other person’s story has the opposite conclusion.
It’s a lot harder to dismiss or refute the numbers as they are.
like just imagine if we could see win rates for entire comps, vs opposing comps, or hell even just hero pairs, rates per map, rates per point, rates per game mode…
i guarantee you would start seeing shifts in how the game is played…and much more variation within the meta itself…
not to mention better balance arguments and balance itself
stats have value…but they have even better value depending on what theyre actually measuring
Literally every time Blizzard has shared statistics, they almost mirror Overbuff’s numbers; furthermore the nerfs and buffs usually line up with who OB shows as statistically over/under performing.
An anecdote borders on useless because it’s a single instance of a particular occurrence, and is easily answered by another anecdote stating the inverse of that occurrence. A <hero 1> main might say that <hero 2> is overpowered because they killed them, but that’s equally valuable to a <hero 2> main saying <hero 1> is overpowered because <hero 1> killed them. Thus, in regards to balance, trading anecdotes is the argumentative equivalent of two people standing there saying: “Yes.” “No.” “Yes.” “No.”
An anecdote provides no objective depiction of reality. Sure, saying something happened to you X number of times in-game is plausible, assuming we turn a blind eye to confirmation bias, but how often does that event actually happen overall? I could find dozens of video clips showing off sextuple kills using D.Va’s Self-Destruct, but any reasonable person would know that this is not a representative illustration of how effective D.Va’s ultimate generally is.
How then, do we find evidence that cannot be countered simply by an occurrence of the inverse? How do we find an objective depiction of how often a particular event happens? Statistics. Sure, someone can say that <hero 1> is overpowered and in every game, but if said hero has a 2% pickrate in the rank they’re playing in, then it’s safe to conclude that their assessment is objectively false.
You can still do correlations off the stats to give a decent indication of how heroes relate to each other. For example, rein’s pickrate skyrockets and then mei and lucio shoot up with him. Ball and tracer’s pickrate goes down and we see brig’s pickrate go down. As lucio’s pickrate goes up, brig’s goes down. Statistically though, brig is now performing worse than lucio in every rank and metric than back when he wasn’t meta. That tells us that lucio is probably both stronger in his niche than brig, as well as being stronger outside his niche. That would indicate that brig might need a buff whereas lucio just needed his meta to come back.
Because by and large it’s all we got. Blizzard like many private launcher based games release next to no info about player statistics like you would see on a steam game.
People ignore the problems with Overbuff… like the fact i Think it still has shield Generator Up time for a stat… or that at times hero’s have higher than 100% pick rates.
The other underlying problem is Overwatch does not in general give us access to the most useful stats in game, and over buff is built on those stats.
It’s a super flawed system but it’s a thing we have. It’s like using a dictionary. They get updated and replaced on a regular basis, some miss words others change their definitions. Doesn’t mean we stop using them though.
Well, not really. Because your individual experience is influenced by your actions in game. A good example I ran into once one tricking Dva was the huge rise of double bubble. Looking at my ranks averages cannot refute that claim. You either take my word for it, or believe I am lying with no logical basis to that belief. It gets us nowhere regardless.
But you know, I am going to adamantly maintain the stance that I run into it once every 3 games or so. And unless Winston has the equivalent of 33% pickrate based on whatever the math is that people trust for pickrate and not winrate… It is apparently verifiable. It really isn’t though.
The point is kind of moot because the stats are only part of the story to begin with, and it is an aspect of hero balance that Blizzard genuinely doesn’t seem to care too much about outside of extremes on relevant heroes. So I don’t really even understand why people have so much passion for this one way or the other.
Ain’t nothin’ but a heartache
Tell me why
Ain’t nothin’ but a mistake
Tell me why
I never wanna hear you say
I want it that way
Am I your fire?
Your one desire
Yes, I know it’s too late
But I want it that way
Tell me why
Ain’t nothin’ but a heartache
Tell me why
Ain’t nothin’ but a mistake
Tell me why
I never wanna hear you say
I want it that way
Now I can see that we’ve fallen apart
From the way that it used to be, yeah
No matter the distance
I want you to know
That deep down inside of me
You are my fire
The one desire
You are (you are, you are, you are)
Don’t wanna hear you say
Ain’t nothin’ but a heartache
Ain’t nothin’ but a mistake (don’t wanna hear you say)
I never wanna hear you say (oh, yeah)
I want it that way
Tell me why
Ain’t nothin’ but a heartache
Tell me why
Ain’t nothing but a mistake
Tell me why
I never want to hear you say (never wanna hear you say)
I want it that way
Tell me why
Ain’t nothin’ but a heartache
Ain’t nothin’ but a mistake
Tell me why
I never want to hear you say (don’t want to hear you say)
I want it that way
What is a “truth” to me rarely reflect the general state of the game - why is that? Because I have biases and preferences that makes me unable to maybe even grasp fully that one situation I’m recounting. There might in my anecdote “always be a PharMercy” on the other team simply because as someone who often plays Junkrat and Sym I will notice this immediately and I will “forget” all the times the enemies were running completely different comps. Stats will in this case show PharMercy factually is not run in every single game.
Thats why you cant only look at pickrates. You need to look at both pick and winrate in conjunction. If a hero is never picked and doesnt win, then that hero is probably bad.
