I don’t consider soldier garbage. Because if you look at the rest of the ladder it does help paint a bigger picture. He’s doing well just about everywhere else and is even a top pick in some places. Why Could this be? Well we know higher tiers have better aim, positioning and can coordinate. Which raises mcree and ashes ceiling more than soldier. Soldier has spread limiting him they don’t. Thier hits are big punches where soldier is a bunch of consecutive smaller hits. It also tells you that you may want to look their kits easy value vs high ceiling. Soldier is well rounded making him great for lower ranks. Ashe and mcree have tools that excel for situations that higher level play can take advantage of. If we think this may be the case we can say soldier is fine and is a hero for lower ranks or the answer may be to redistribute his power and not flat buff so he works better at higher ranks and not break lower ranks. Or you can just be blizzard and say screw the lower ranks and flat out buff.
Anecdotal wouldnt tell you that. It just says “soldier isn’t in my game. He sucks” You know nothing about other ranks which would have told you a lot more about his situation. Stats aren’t there to tell you everything. They never have. They are to paint the picture and you use reasoning to try and see why they are what they are. Which falls back to learning how to interpret data.
Anecdotally you could say something like I rather the stun than the self heal or I find that camera shake makes it to hard to use (he doesn’t have spread anymore), or I can never get value with his ult, or I find I can never land rockets
Or the other way around…I can’t get flicks in, or I play with a friend who plays X and we combo Y together effectively…
All the intangible stuff that doesn’t necessarily show up on the page
It all adds to the same picture
I don’t think it should just be dismissed because it’s one persons experience…it’s all still part of playing said hero
Because by Internet Lingo standards, people figured out that they could say “Oh that’s just anecdotal evidence, it doesn’t prove anything,” as a way to write off people’s lived experiences and not have to listen to them or change their thinking/behavior in any way.
It’s convenient, easy and free, plus it’s a safe haven for the mentally weak who don’t like thinking of the world in complex layers of gray.
It depends upon the claim you are making using that experience. If you were to say, “I see double-bubble every 3 games”, then that’s a perfectly fair statement to make based upon your experience, unless those heroes have near-0% pickrates. But it’s important to realize that there’s a difference between what you see and what actually is. If you were simply to say that Winston/Zarya is in every 3 games, I would tell you that you are wrong, because objectively, you would be. Zarya appears more often than once every 3 games right now, and Winston appears less often than once every 3 games right now.
Even if you stick to the “I see” statement, your anecdote can still be easily countered by another anecdote that says something different. Who’s to say one of more valuable than the other?
In short, an anecdote is what someone sees. The statistics are the way things really are.
And that stance is perfectly fine, given that you acknowledge that it is your experience and that said experience may not be representative of the way things are. Maintaining that position is fine, but it’s not going to get you anywhere because in an argument, the experience of the person you are arguing against will always matter more to them than your own experiences. The fact that you would be so adamant about your experience is exactly why anecdotes are so useless without support from other resources.
Depends upon what you’re going for. One player’s experience (or simply their perceptions thereof) can be summed up in a single anecdote. The game shouldn’t be balanced around a single person’s personal experience (once again, who can say that theirs is more valuable than anyone else’s?). The game should be balanced around the way things really are, and you’re not going to arrive at that through anecdotes. You find that through statistics, the patch history, and… historical statistics.
About this:
A clear interpretation of winrate is muddied by other factors that influence it that do not influence pickrate. If a hero is insanely overpowered, they will be picked on both teams quite often. Their pickrate will skyrocket, and their winrate will be… average. Why? If a hero is on both teams every game, every win for that hero is also a loss for that hero.
I think the question is really what’s more valuable: Anecdotal thoughts and opinions of thousands of players’ real experiences, or raw data?
The answer is, frustratingly, neither. They’re both equally important. Anecdotal experience doesn’t always represent the full population, but you also can’t accurately understand the data without proper context.
Pickrates are nothing but a popularity contest and shouldnt dictate balance for mid-low ranks. If you wanna talk about balance, you can perhaps use that as one of your claims for GM for the top and bottom, but dont base off your evidence solely by that.
Even then it’s an extremely large sample though. Likely made up of tens of thousands of profiles. Which larger than a lot of scientific studies to be honest.
The thing is that unless someone studies statistics themselves and why smaller sample size populations can still be accurate for larger populations no one will ever be convinced.
The issue with Overbuff is that the developers of the site never explained what a bunch of stuff means. Your own personal stats are really just say all of the healing you have ever done divided by your wins. That’s how it gives you stuff like damage/g or healing/g. It would be really great if for a given week we could see what competitive stats are being polled.
My issue is with the following.
Anecdotal evidence is a great start for trying to recognize patterns. However, humans are incredibly faulty for memory. Using witness testimony as evidence is incredibly faulty and one of the worst pieces of evidence you could ever use.
You’d ideally like to follow up on the anecdotal pattern you feel you saw and formulate a hypothesis on, “I think Hanzo is in every single one of my games”. Then, you’d ideally follow the scientific method and actually try to log in a large number of games recording was Hanzo ever in my game or not.
Not 10 games as most people would do, but rather closer to a 100 or more. The thing is that people don’t have the patience for that.
But weren’t they the most vocal group in asking for private profiles? It seems counterintuitive to think that they would suddenly just decide to leave their profiles open when they’re the ones who wanted private profiles in the first place, no?
There’s little reason to believe that one tricks with private profiles make up a statistically significant amount of the data to completely throw off all the statistics.
You also said specific mains, people might main one thing but dabble with other heroes sometimes. I’d imagine Mercy mains or even mercy one tricks might still be inclined to hide their profiles if they opt to play Symmetra for a few dps matches, or Dva for a few tank matches.
Applies to “mains” as well. There’s no evidence suggesting the ratio of players who do and do not use private profiles is significantly impacting statistics.