You have an elitist perspective because you look at the normal person — and that’s what you mean by “bad player” here, a person of roughly average skill — and you view their very existence as a contamination of the data. I view it as the job of data to reflect reality, not my desires, and that everyone of normal or below average skill who is out there exists and should be reflected in the data in some manner. You want to escape humanity. You can’t escape humanity. This is what we are, warts and all, and our species really isn’t all that intelligent.
Not only that, but also good players who didn’t bother to play enough games to learn its’ ins and outs, as well. So that’s two profiles of players that skew the statistics.
And I respect that part about you, but you’re focused exclusively on big data, instead of data in general. Big data have their own definitions, advantages and disadvantages. You are not taking specific context into account, and I’m sorry, but in this case you are wrong.
You don’t want to believe that data is relevant because you don’t want to believe that you can be caught in a lie. The very idea that anyone but you could prove anyone else wrong is abhorrent to you.
It’s you who’s trying to do that, unfortunately. Everyone else on this forum is aware that big data stats aren’t an accurate representation of what win rate they will have if they pick that deck on their own and play a few hundreds games with it xD
For starters, it’s mathematically proven that the higher the sample, the higher the convergence around expected value (in this case towards 50% for balanced decks, towards 100% for unbalanced)
So that alone is enough to prove my statement. Few hundred games cannt possibly converge that fast.
Also my ladder can be full of 3 specific classes because I’m playing at a specific moment. Big data reflect total games vs all possible matchups during a longer period of time. Again, it will NOT accurately reflect what will happen to you if you pick up that deck.
I’m quite fascinated by how you refuse to accept this obvious fact.
And I repeat again, when we talk about nerfing a class, we should talk about aggregate data, big data, your perspective.
But when we’re talking about individual person’s success on ladder, we don’t. Those data become irrelevant.
1 Like
There’s that elitist thinking again. The only way you could possibly utter something so obviously incorrect is by “othering” a large number of people out of the “everyone” class.
And I am very confident that, once a person wins enough games that they’re actually matched against players of roughly equal skill, that from that point onward their winrate will basically match the big data stats. I get that you’re a T1KL tryhard and that for you that takes WAY more games than the average person who reads these forums, but I’m nevertheless pretty confident that you’re not like top 10 players on the EU server and that eventually you will meet your match and your winrate will normalize to 50% from that point onwards.
I can’t possibly play more than I play these days. It’s impossible that happens to me for sludgelock until the meta shifts again, and if it shifts again, so will the sludgelock data.
I can be stuck in top 200 legend without ever converging the aggregate data. It takes much more than 70% winrate with a deck to hit top 10 legend. I’m nowhere near close to that. And I am aware of that, because I take other things into account than just pure, big data (which we don’t even have a lot ways to analyze right now, at least not without using machine learning algos)
If an average person asks “how can I improve my win rate” and you tell him to pick the highest win rate deck on the ladder currently, he will still hit a wall with that deck because his skills are average.
But the moment we start talking about strategy and how to counter certain matchups, we ARE OUTSIDE of thae domain of average players, and INSIDE a domain of high skills, which is, to be frank, my domain in this conversation. And that’s when you DON’T drop in with big data stats because they will not help an average player play better.
I wouldn’t tell him to play the top netdeck though. I’d tell him to play every netdeck. Understand every matchup. Then pick one of them from there.
Technically, it is possible to understand a matchup while only playing against it, but it’s an order of magnitude easier to understand if you’ve played as it.
In case you didn’t read.
You’re not going to shut me up because you are “better” than me. I respond to evidence (which I think I’ve shown fairness in doing), not authority.
You said you only follow what big data stats say. I am not shutting you up, but refusing to accept big data as relevant in such a conversation.
That is not shutting you up no matter how much you try to twist it.
If you can decide to refuse to take my evidence into account, I think it’s only fair I can do the same.
Okay, well what you’re trying to discuss is skill development. You might notice that my answer wasn’t a “check the big data” answer, it was a “play all the decks” answer.
My only “data” assumption on skill is: that the matchmaking algorithm is efficient (but not perfect) at sorting players according to skill, such that the average player at rank X+1 is higher skill than the average player at rank X, with the understanding that bottom Legend is a lower rank than wherever newly Legend players are put when they graduate from Diamond 1 at the time.
Then there’s I think only two useful concepts that’re derived from that, and that’s pretty much it. Maybe someday I’ll find another, I don’t know. One of these concepts is which decks reward skill the best, which can be empirically measured, but by “reward” here I mean in terms of wins per unit skill and emphatically not that playing a certain deck makes you more skilled than playing another, or that a deck that scores high in this metric is better than a deck that doesn’t. Anywho, my point here is that data has a limited role in the skill discussion. I don’t think it’s nonzero, but the data doesn’t have a lot of "in"s so it’s difficult to have relevance.
Data isn’t going to answer questions like “how do you build skill?” And I already indicated that I don’t really think that big data is relevant to a player until they hit their “deserved” rank, the rank at which they are playing against equally skilled players. I do think that, assuming infinite time played, that everyone would reach that point eventually, but there isn’t infinite time played. Rank resets monthly. So it’s more like I think that the vast majority of players will hit that point within a month. Half of them hit it instantly, without having to climb at all, because half of them are below average (although this is counting bots as “players” so in the current context it’s a little off, but most human players still hit that point well before Legend).
This what you wrote to me. You called my perspective “elitist”, rather than “of above average skills”, which it objectively is. That’s a funny way to start a discussion which should be objective.
And then you went on to say that data reflect reality, not your own desires, implying that I might read from the data what I desire to.
I have to remind you that this whole conversation started when a specific person decided to say that DH is not overpowered, and implied that I should be having no problems to beat it playing sludgelock, which is both wrong objectively (from aggregate data, and my own data) and subjectively (from my experience).
And then you decided to chime on on his side, without reading the whole thread, which you even admitted in your next reply:
That’s inflammatory and offensive, btw. I’m also European. The amount of disrespect you gave me in this thread warrants a report and an ignore.
1 Like
That was directed at him, not you. So far I mostly like you, if it matters. In fact, so much that I’ll remove that comment. I still don’t like Sparky though.
I’ll go do that then maybe respond to the rest of your post.
I see no difference between these two things.
Which I pointed out. 65-35 in Shopper’s favor. So agreed. Sparky wrong once again.
Well I have now, and he’s still essentially correct on that one point. Even a broken clock is right twice per day. I also try to make a point about agreeing with people I dislike when I think they’re right, although I suspect that I’m still woefully inconsistent at it.
60% winrate is, in my opinion, something that basically only happens when players are below their deserved rank, as defined in a previous post. The matchmaking system gets players to their deserved rank very efficiently, meaning that players not being at their deserved rank almost doesn’t exist. I guess it behooves me to clarify that almost not existing is a smidge different from not existing at all, so my agreement with Sparky is less than total here. But your experience here is, to say the least, extremely unusual, because you have an extremely unusual level of piloting skill, which you’ve proven beyond reasonable doubt.
So how would you explain me being stuck between 1k and 3k on EU for more than 700 games with that winrate?
It’s not as efficient as we’d like to think it is. Higher ranks are stacked, lower ranks are just bots. My brother started playing 2 days ago and he’s already silver 1.
They made it like this. We can now enter a guessing game of why they made it like this, but they made it inefficient and stacked in higher ranks.
I believe that should be enough to explain the difference between individual player’s stats vs aggregate data stats, and why context matters.
Let me tell you what my personal experience is like.
I started going to these forums about 3 years ago. I’d consistently make Legend every month, but I mostly played Arena, where I’d occasionally leaderboard and got 26th Americas once. At some point I lost interest in Ranked, then in Arena. For about a year and a half I would post here and not even play Ranked. For a long time I just had the game uninstalled, for some time I just played Battlegrounds. I didn’t believe that if was necessary to have play experience to look at the data websites and analyze the meta with more competency than 80% of the people posting here.
About two months ago I had a few people calling me out on not playing. On the one hand, I consider such arguments fundamentally invalid, but on the other hand I figured that making Legend would be pretty easy, so why not? So I started playing Ranked again. Made Legend on the 4th of the month, easily. Didn’t really try past that. I didn’t track my winrate while climbing but I suspect it was over 60%.
This previous month also made Legend, but I waited until after expansion release to climb. Even easier, probably a 70% winrate average. After I got to Legend I got bored and played a couple meme decks but still climbed a bit. Probably finished top 10k, I think I was like 6k when I stopped about a week before the month ended.
I half consider you to be a higher tier player than me. The other half doesn’t really know, because I’m not seriously attempting it. I’m goofing around and not really playing that much. As far as I’m concerned I have fully confirmed my hypothesis that actually playing Ranked isn’t necessary and that one can analyze the meta competently with data websites alone. The only thing I feel stupid about posting about in hindsight was the botting problem; it was way worse than I thought. I enjoy posting more than I enjoy playing. But if I did try, maybe I could match you. The thing is, I don’t actually care enough to try. I will admit that the possibility that I could try hard and possibly not achieve your level of results is paralyzing. I’d rather just say that you’re better than me. It’s far easier.
The point I’m trying to get at is that I understand the 60-70% winrate thing, at least as far as it goes to top 5k Legend or so. I can do that every month. When you say that Americas Legend isn’t that hard, I feel you as far as personal experience goes. But I don’t consider my personal experience to be at all representative of normal. Because it very obviously is not. I think it’s kinda insensitive to say, generally, that Americas Legend isn’t that hard. It’s very hard for most. Who cares if it isn’t for me?
I don’t have any problem at all with you being elite. I haven’t really done the work to ever achieve my deserved rank, maybe it’s below yours, maybe it’s very close. In my best Norman Osborne voice: I’m something of an elite myself. But just because you’re an elite doesn’t mean that you have to be an elitist. I think the appropriate context for this forum is not quite the “average player” experience, but the average long term player experience. Filter out the people who have only been playing this game for a matter of months, focus on the people who have been playing for years, then take the average of that. What you get from that average is far below you and it’s also below me, but that is much more representative of the playerbase.
How foolish of you to try that card (a second person to have done that recently… What is it with with this syndrome: Valeera avatar, delusions of grandeur… 
).
First of all, of course, the ad hominem fallacy — obvious as it is.
Second, you never know who you’re talking and trying to compare the length of your… skill to. I’ve hit numbers higher than that in my rather amateurish Hearthstone ‘career’ (see, for example, also this), yet I haven’t had any delusions of grandeur about them and their significance, putting on airs like I’m someone just because of some stupid figures.
Allow me to tell you a little personal tale — indulge me, I think it’s a useful one.
Had a ‘friend’ in HS years ago, maybe it was open beta, maybe the first release — I don’t rightly remember, and it matters not. You know how I made this one? There I was, sitting, watching some local HS stream, making ironic or even sarcastic remarks (hmm, perhaps that’s my modus operandi after all) about trying to climb to Legend with Priest (back in that era — the original ‘Classic’ or maybe even ‘pre-Classic’) in the chat. Some guy there responded to me calmly that he’d done it. I swallowed my next snarky words, feeling a change of attitude was in order, and asked him how he’d accomplished it, with what deck and so on. The guy replied to me politely and constructively, provided with a deck (codes weren’t even a thing back then — we used some websites with deck builders) and tips how to play it, although he admitted it could be tough, and that’s how I got to know him a bit.
A few years later, he won the HS world championship (or whatever it was called), and that was the only time I was more or less watching, since that kind of ‘e-sport’ hasn’t generally been my ‘thing’, so to speak, due to heavy ‘RNG’ and all that. That tournament was no exception, but I was still rooting for that guy, even if I might not had been in touch with him by then. Subsequently, he had a much more successful ‘e-sports’ career than probably most people on these forums would ever have, but you know how I remember him most? Not only as a successful player, but a nice a polite guy with a constructive attitude.
If you’ve gotta learn from the best, maybe learn something from this example?
That was my statement, not Kassadin’s.
Rest assured, I don’t care for your sort at all either, so to speak. Anyway, that doesn’t preclude me from admitting that you’ve had a point in this topic… and acknowledged my statement, which I believe to be logically and factually true — so there’s that.
But anyway, back to the substance of the matter — the data. By the way, sorry — unable to keep up with the rest of the discussion yet. I’ll quote some examples of a general point about the win rate of the ‘most succesful’ and ‘top’ players, then respond to that with a data argument.
One could look, for example, at this kind of data from D0nkey:
https://www.youtube.com/watch?v=978i-YTm2Hg
Yes, I know there’s a tracker bias and all that, but I’d think it’s not that significant if you look, for example, at top 100 Legend — one could reasobably assume that many of them in that league are running one and it doesn’t provide a significant advantage on its own.
So, what have we got there? People from top-100 legend running a particular tracker enjoy a winrate of about 51 or 52 per cent — those ‘extra one or two percent’ (alright, maybe three, although the errors are larger for a smaller sample size) I was talking about. By contrast, suspiciously high win rates in other leagues have been linked with the bot infestation.
If a guy tells me of a ‘global winrate’ of 58,5% and easy 11 stars, I suspect something isn’t quite right here. It doesn’t mean a person is lying — it could be bots or something else, but such a figure itself in a game like this isn’t right already.
Besides, I’d had some unfortunate experiences with free 11 stars when Classic was hit by bot waves (the first time was different, though: after returning to the game and thus having no bonus stars whatsoever, I had to ‘grind’ so hard and win so many games, that the win rate was skyrocketing, I guess; during other months, even landing in top-50 or so would guarantee nothing — and they say the bonus star is determined more by your win rate than by rank…), so yeah… I can’t rule out this possibility as well.
Alright, probably gotta take a break here…
PS One more obvoius thing: of course, a single ‘outlier’ random event is also possible (if I remember correctly, some would-be discoveries e.g. in particle physics at the ‘three sigma’ level of confidence, if you know what I mean, disappeared at ‘five sigmas’) — my point was that in general such numbers look very suspicious.
Oh, but I do, because your replies to me showed it.
Well, I didn’t. Good luck trying to prove it to anyone who’s actually playing with 11 stars now.
If it takes an infinite amount of time for my win rate to converge, for all practical purposes, that convergence is immaterial, useless.
Norwis had a score of 70-47 on his grind to Legend 3 with Mage today. That’s 59,8% winrate, which is slightly higher than my winrate, hence why he’s top 3 and I’m top 200. And that’s not an outlier. The guy’s been top 10 consistently ever since I’m back to this game. And unfortunately, it takes more than 60% to hit #1. Are all the best players fishy?
Source, that same d0nkey top you keep misquoting as a source for any of your delusional conclusions.
Here, I’ll even link it to ya:
https://www.hsguru.com/streamer-decks?twitch_id=549965312
We are done here. I am so sorry another thread has been destroyed by me trying to fix someone’s misconceptions. It will not happen again. I now know at least a few more people whose bait I will never take again.
1 Like
I know.
Oh for Pete’s sake, link to d0nkey directly. This absolute hack of a YouTuber doesn’t need more clicks.
I could nitpick a few things in this, but it’s very reasonable overall. As reasonable as I can hope for. I don’t really get why Kassadin feels compelled to argue against it.
Because of this:
That’s it. Not even reading posts, correcting people with no clue about anything, I’ve had just about enough of it.