Your “doubts” are baseless. His intention was to discredit me, and to discredit me he used an incorrect piece of information. That’s lying. And maybe slander.
I edited that part later. As you can see, it’s:
and that’s correct, and that’s why your example is meaningless. If anything, it’s a proof that if you increase the denonimator (or whatever the bottom number is called in a fraction) you decrease the result of the division, which is exactly what I’m trying to explain is very significant when you’re dealing with samples of different orders of magnitude.
Again, I’m not seeing anything in your reply which serves as an argument or a proof of anything. I see an incomplete methodology:
Not saying anything which can serve as a counter to my argument.
Even if they took an average of a billion mini-samples made up from one large sample, it would still be subject to the law of big numbers.
I asked him what rank he was and his EU name. He refused to answer. This is because that name was not on any rank better than 2k at the time. I checked and wanted him to answer. He would not.
My speculation was he wasn’t using the same name and was weirdly dodging to answer.
What he did say was he was above 2k because he just so happened to tank his rank experimenting with rainbow shaman.
Then, within the next few days his name was at 550. Did he change his name? Don’t know. But his name 100% wasn’t better than 2k, he even admitted it himself.
So at no time was I incorrect in that. What I was incorrect was thinking he still no longer was better than 2k, but then his name did appear that day so he improved it.
The question has always been why not say his name and rank when asked. When I did ask, he got defensive and claimed to block me clearly wanting to avoid the situation.
There’s no point in answering seriously to anything you write, you spent all your credit with me
If you wanna make a clown of yourself and accuse people of something, you better have proofs ready, cuz I’ll always just have mine.
Sitting comfortably in top 200 EU and top 240 NA.
Let’s say it’s possible for me to drop below 2k, and that I did, by experimenting with rainbow shaman.
How does that change:
a) the fact that I was top 1k during the whole time of the report
b) that I proved the statement a) with screenshots from tracker
c) that a) and b) happened only because you were grasping at straws to discredit me because you didn’t have proof or arguments to do it fairly so you had to invent conspiracy theories and lies about me?
You have shown your true colors. You mean nothing to me anymore, where once you were an authority on matters of respect, politeness and honesty.
Once again, I have fallen for a narcissist. Nothing new, but also nothing permanent.
That’s not how it works at all. The number of games of matchup data is not equal to the simple product of total games × opponent side popularity. I recently wrote a post explaining how it actually works.
Please note that even this is subject to random variations. It’s less true with smaller samples but becomes more accurate with larger ones.
Speaking of…
… would imply that as a sample size grows, subtracting wins from losses would yield a result which also grows multiplicatively. You don’t understand the law.
He said around 1k, unless you asked him off of the forums somehow. I don’t know how that telephoned into above 2k but somehow it did. The climb to 550 was not surprising.
What your post “explains” is the difference between “games played” and “games recorded”, which is in this case, irrelevant. Or it might actually be relevant, but…highly unlikely.
It’s possible that this discernment explains why the report doesn’t align with my personal experiences and data, but it’s very unlikely, because I also have a tracker installed, and have already PROVED that there are decks represented in the report with less games tracked than what I had on sludgelock (elemental mage, and now I see couple of others, since the meta fragmented even more into many less played decks)
But it’s really not important for this discussion, which is about why you can’t put those two samples in the same basket and make a ranking list with them, just like you can’t put boxers of different weight categories in the same ring and expect a fair fight.
It’s not the law I was referring to. I don’t know the law’s name. I just know it works, and I know it can be described using limits
Like, how limit of n tending to infinity approaches 0 when you have n/n^2? yeah well that’s the law in a nutshell
Since we don’t have a quadratic function in our samples, but it’s something like (n-m)/n, where n>m, our limit when n tends to infinity is 0,5, and that’s what the results of division with growing numbers approaches
P.S. Whatever the number 555 actually represents, keep in mind it’s split into 2 versions of the deck, as well xDDDDD
That may be so but I went to look at the August 1st report to see if it had ‘rainbow’ shaman and that deck wasnt there… So I frontran the experimentation on those decks well before they posted them and I am contesting that ‘rainbow’ shaman is the best shaman list.
If you look at the matchup winrate table for both top1k and legend for rainbow shaman it seems like the deck has barely been played… let alone vs a competitive class distribution. Until that table is actually filled out I think its a unicorn deck that they are grossly misstating its ‘power’ level.
You re so in his head that he had to do an emergency climb dropping the deck he claimed was the best thing since xmas.
All he has going for him is this rank climb and thats fine but along the road he decided he deserves royalty status over it and the damn forum users dont take him seriously.
Why would people put more weight in VS reports than his word is a mystery for the ages.
I am really enjoying this meta. There are so many decks to be played and a couple of skill cap decks to play as well. I cannot think of a meta that has ever existed like this in recent memory. There’s a couple outlier cards but overall everything is in a pretty good spot. I can see at least two more decks emerging just from the mini set stuff as well.
That’s true for the majority of the meta. Over 50% of decks by popularity have matchup sample sizes below 100, therefore VS doesn’t publish.
Considering that they use that data anyway in their overall winrate calculations, they probably should publish them anyway. It’s hypocritical to say “this is a small, potentially misleading sample” then use the number in a formula and publish the result.
He said he dropped out of 1k. To me that means you’re no longer in the 1k bracket (1000-1999). If he meant out of top1k he could have said it.
Either way, his name never appeared anywhere in the top 2500 because I checked. That’s why I requested his rank and name. There’s a reason he didn’t give it. He loves to brag but refused to let anyone see it officially. Probably needed a day to change his name it would seem and doesn’t want us to know his original name.
Speculation for me, sure, but what else explains his name not being on there?
Well that answers the question of how it telephoned. Now I know.
It’s easy to miss with the way the leaderboards are designed. I believe you when you say you tried, but I believe that it was before your eyes and you didn’t observe it.
I don’t believe that Altair would be below top 2500, pretty much ever. He is an exceptional pilot and plays a lot.
I’ve read your post about methodology. I’m aware of:
However, I’m OK with that as it doesn’t change anything regarding my argument, though.
It’s also a pretty decent heuristic to take into account.
I’m also aware of this:
But I guess they’re as objective as it gets, considering they have no access to actual data and have to rely on tracker data.
Still, nothing changes the fact that sooner or later in your calculation you have to divide some numbers to get to a percentage. A percentage is literally defined as a fraction of something multiplied by 100, and you can’t get a fraction of something without dividing a part of that something with that something.
This was very rude and untruthful.
That number is apparently a number of games tracked in top 1k for Rainbow shaman, which means that the actual number of games played isn’t much different, as most of the players in top 1k use a tracker.
Either way, it doesn’t make much of a difference in the grand scheme of things.
1,5% played or 2% played, doesn’t change the fact that to calculate a winrate you have to divide with a number of games, as that’s what fraction, a percentage means, and the higher that number of games, the lower the resulting winrate will be.
The last part of your sentence is actually contradictory.
Playing frequently often leads to larger fluctuations in rank compared to someone who plays less. Significant streaks of winning or losing can greatly impact someone’s rank, especially if they’re not in the Top Legend MMR, where losses typically result in only double-digit rank changes.
I’ve seen Altair’s rank fluctuate dramatically, from 4-digits one day to 3-digits the next, and I don’t mean just a rise from 1000 to 900. I’ve observed him go from 2000 to 200 in a single night and then back to 2000. Don’t underestimate his gaming skills; he’s a quick-deck connoisseur and experiences substantial fluctuations.
I don’t understand why you (both) put so much pressure on him to maintain a certain rank all the time. Your view of High Legend players seems to have distorted your understanding of rank fluctuations within the Legend tier, that isn’t double-digit.
