You can just say Whitemane, it’s fine.
They never said it was accurate, they said it was inaccurate if I remember correctly
A couple years ago at blizcon they said hotslogs had a 15-20% error factor from thier internal data.
Release thrall got a massive talent adjustment and removal of battle momentum in about 2 weeks. Back when bm was 1 second cdr per aa. Even back then when almost every one was broken, he was either stun locked and died, or was unkillable until he was oom
To keep it short, correlation does not equal causation.
This is such a simple thing I didn’t feel a need to explain it. Do you need a more detailed explanation?
Have you considered that part of this is because you guys still haven’t released a public API for the game? Our only access to data for the game are websites like HOTSLOG and HEROESPROFILE which are not 100% accurate and there are many people who would rather use their own experiences vs that flawed data to make judgements about the game. Perhaps access to official data, even if it’s limited to the type of data we’d see on fansites like Hotslog or Heroesprofile, would bring community sentiment and data closer to parity?
6 Likes
I think Samuro is perhaps a poor example to choose for the point you’re making.
For example, removing the hearth trick removes the "looks and feels (and IS) unfair parts from his kit while perfectly maintaining and keeping the swapping between clones to trick people outplays part of his kit that is high skill and should be rewarding for the player.
A character that lacks counterplay and is frustrating to play against should always be changed. See old Chromie as an example. A frustrating hero affects 5 other players in every game, Nerfing that hero affects 1 person. The effect on games is very lopsided.
There are also a myriad of other reasons to play different heros, Raynor for example is picked in Master+ not because his skill floor is low, but because of his safety from his Q and range etc. Hero’s can have reasons to play them unique to their kit that do not need to be broken interactions that only a minority of players know how to do. I do not think it is possible to balance a hero around such strong interactions when such a minority of the playerbase knows them. The hero is either balanced when using the trick and useless for 99% of players, or balanced for 99% and stupidly broken for 1%. I think you’d agree that neither of these situations is attractive compared to removing the trick and being able to balance him for 100% of players.
Your Probius example is also somewhat concerning. A low playrate hero will almost always have a slightly inflated higher winrate. Low playrate means only dedicated hero mains are playing the hero, and they are playing against people who are less likely to be familiar playing against their hero. If Probius has a 1% playrate, and Fenix has 10%. Fenix might have a 5% lower winrate overall, but if you look at the top 10% of fenix players (Meaning roughly the equivalent 1% playerbase) you will obviously find that Fenix’s winrate increases. This is just common sense. Nobody should ever use this as justification for Probius deserving a heavy nerf. If anything, what he needs is changes that will increase his playrate. Heavy nerfs to probius wouldn’t go over well with the community because it would show a fundamental misunderstanding of low popularity heros from the balance devs. A nerf would demonstrate a lack of competence from the balance devs.
2 Likes
Oh god they’re onto my Probius boi nooooooo
Seriously love the response though.
3 Likes
How is that relevant?
He said the data sets match. We know one of them is accurate. Therefore the other is accurate.
Sure he may be exaggerating a bit about how closely they match, and if we ignore all previous knowledge we have about this subject it would seem like he’s extrapolating and over-generalizing, but there’s no questionable-cause fallacy here.
2 Likes
With all due respect , im curious as to why ?
Wouldnt it help everyone and provide a better transparency over any debates or makes any nerfs or buffs more understandable from the community since we would better understand your point of view ?
The only negative i can think about would be if a hero that no one complains about is actually outperforming by a large margin , then suddenly people start complaining as if X hero has been an issue since day 1 but no one ever cared or complained about it before.
Also , regarding your example of samuro.
I wouldnt really touch him really.
I dont think every heroes should be exactly 50% win rates .
Not every players has the same playing style or skills . others have different prefferences which is why not everyone even has the skill to play samuro.
That said it would make sense that the player base able to even play samuro are on average higher skilled than the player base who main raynor or lili, thus would make sense their win rate to be slightly higher.
4 Likes
Brilliant, sound, respectful response. Thank you.
I am glad to know this part specifically 
.
I can understand that even if it’s a little unfair , thank you for the response and more importantly for all the effort you do to achieve fair balance & perfection 
Like I said, to keep it short.
It doesn’t exactly fit, as you note.
But I felt like it was the best way to show how such leaps of logic don’t work. And I do fully believe anyone who puts critical thought into it, after understanding the correlation-causation fallacy would understand how faulty the logic was.
I probably should have just waited until I had time to do a proper or semi-proper post about it.
you should ask how many actual players there are while you are at it. I’m sure they will be just as forthcoming with that information
5 years people have been waiting for an adequate MM system and reporting, and where are they? Moreover, a bad MM generates abuse of the reporting system, and this in its turn affects the MM and the number of players, and therefore the MM itself. Solving these two problems is easy:
- The reporting system should be limited by the number of reports per week.
- And the MM to fix with the button “quick search”/“qualitative MM with long waiting time”.
I want to emphasize just in case that both problems are interconnected and affect each other.
Not on the topic, but the truth.
While I am someone who wish they did/would do this… you’re missing the issue.
Pushing out Queue times even a bit to make all games higher quality had a massive backlash from the community overall.
Making it so some people are willing to wait with tighter MM rules pushes the queue time up for everyone (which is a large negative) and also makes designing a good matchmaker much harder.
I WISH that was a viable solution, or they would just do it anyways… But there are very good reasons not to do this “easy fix”.
Does the data effect things like quick matches or just ranked play? Is it used to determine the matchmaking at all? Like how Murky and Abathur are never selected by the computer to be on the same team and I do mean never. Where can this data be viewed by players?
didn’t know that you mained Xul.
Probius was my second purchase in heroes of the storm. I loved that he danced with lucio during the trailer. Probius dancing with the character of overwatch who my first purchase is.