Depends on the total sample. If it’s 1k out of 75k, 1,33%, I wouldn’t take them too seriously, but I would keep an eye on that deck just in case it’s a sleeper/metabreaker.
55% is seriously broken, but the sample is too low, and if you have at least half of education as I did, you must know about the variance law (too low sample = too high variance, high sample = low variance)
Ok. Same question same amount of games but instead of 50 players it’s 1 player.
What does that tell you about the deck?
Literally the same to me.
I see no difference between 50 different rank 100 players on 1000 games together and 1 rank 100 player on 1000 games alone.
Same rank, same sample, same skillset.
I honestly don’t want to know why you’re asking me this. I seriously don’t want to know where this is leading. I hope you don’t tell me, because I’m afraid I might lose it and get banned.
Do you want me to lose it and get banned?
Then please explain to me why you insist Sludgelock is good in top1k, despite having admitted that you have less games than what I gave this theoretical deck?
You have basically just admitted that no one on these forums should take anything you say seriously about Sludgelock you play because you don’t apply the same standards to your own deck that you apply to other decks when it comes to data.
It is hubris for you to assume that the 75,000 top Legend games are imaginary and that your pitiful tracker is the truth.
There is no “enough.” There is only better and worse. A larger sample is better.
This is overall true, except Tempostorm doesn’t use data. The site is literally a handful of Altairs Altairing, and even that might be giving Tempostorm too much credit. Tempostorm is even more worthless than HSReplay.
VS report is made for 6 days period of 14-20th august
This patch, I have 1.155 games worth of sample, all in 1k
If you refuse to accept this as valid proof, because VS report is for 6 data so I should only take into account 6 days of my data, I can respect that.
What I can’t respect, is that you trust some random imaginary statistic which is calculated God knows how, based on 0,3% sample (less games in this patch than I have) over me.
That, I cannot respect. It’s both wrong and it’s disrespectful.
No, what is imaginary is “Power rankings” statistic by VS
Are these all with you as Sludgelock?
No that’s literally just winrate. It’s not imaginary at all
How about you dm me the lottery numbers ?
Damn some people live to prove you right,either that or some dark sorcery is a play.
I better start building a pyre just in case.
No, it’s not. Stop lying.
The Power Ranking scores are each deck’s weighted win rate against the field. We calculate a deck’s Power Ranking score by weighting its matchups against other archetypes, factoring each archetype’s frequency. Both matchup win rates and archetype frequencies are factors that can change at different ranks, which is why the Power Ranking table can be filtered by rank groups.
All I see is “We manipulated the data we got putting different weights on it and called it Power rankings”
“oh and it’s not just winrate, it’s also matchup frequencies, rank, bla bla bla…”
Stop trolling.
For a sample of 0,3% of total sample, variance is too high to take the data seriously.
Also, if I have more games on Sludgelock in those 6 days than they have of rainbow shaman, why isn’t there a sludgelock on the tier list?
Biased and hurried report, done absolutely irresponsibly.
EDIT: I figured it out xD How they calculate it xDD
If you take this nonsense seriously, would you like for me to calculate a new table using my own data? It’s literally the same formula, and I can prove it. They just took the number of games in top 1k, divided it with the number of games in all ranks together (weight) and multiplied it with the deck winrate.
Too bad that:
a) the sample is too low, variance still a problem,
b) weights disregard the difference in skills.
Deck with the highest winrate = 100 Power Score, by definition
Deck with exactly 50% winrate = 50 Power Score, by definition
It’s literally winrate, just measured relative to the top deck instead of using percentages.
Edit: source https://www.vicioussyndicate.com/drr/faq-data-reaper-report/
Scroll down to question Q: What is the meaning of the Meta Score and how do you compute it?
No, it’s not. If you don’t see a difference between your one-liner and this block of text:
The Power Ranking scores are each deck’s weighted win rate against the field. We calculate a deck’s Power Ranking score by weighting its matchups against other archetypes, factoring each archetype’s frequency. Both matchup win rates and archetype frequencies are factors that can change at different ranks, which is why the Power Ranking table can be filtered by rank groups.
Then you need glasses, and a miracle if you want to talk to me again, cuz I just can’t stand this blatant lying
Highest winrate in a sample of 0,3%, weighted based on total number of games, archetypes and ranks??
Why cant I see that winrate exactly as it is? Why is it in form of an index?
I can tell you MY winrate with 20 games more of a sample. Why can’t I read that winrate anywhere, even though it’s useless because the variance is too high?
Why am I even replying to such troll posts?
/popcorn
I enjoy watching this, but I think I’ll throw this in to sorta help Altair
So VS tracked 75000 games in top 1k
VS lists 30 different deck types. I don’t think that’s all there is in top 1k, as there may be deck lists that are just not played enough to make it on the list.
But let’s say it is 30
To get data on each deck vs every other deck, it should be (doing napkin math here) 30 choose 2, which is 435
75000/435 = about 172 games for each match up
Over a week, that’s like 24 games of that particular match up a day. On average.
This is assuming equal frequency though. If a deck just isn’t played much, you can’t expect there to be that many games recorded.
So I can see why Altair sees no significant difference between his own numbers and vs.
24 games a day may be a lot for a non-legend/non-rank climbing player, but if you’re in top 1k, I could see you being dedicated enough to do it.
I can probably rack up that many messing around my meme decks. My meme decks lose fast enough that I can get through many games quickly 
Well, now you have broken his heart, I’m sure.
Schyla. I wish you all the best in this most difficult of breakups…
Thank you, sir!
Means a lot to me, considering that two highest-ish posters on the forum simply decided to troll me tonight.
Let’s recap this real fast:
After all this, they still believe VS report over me, probably because, hey, it’s a website!
EDIT: https://www.vicioussyndicate.com/drr/classarchetype-distribution-data-reaper-report/
Link to the archetype distribution - what’s that shaman deck’s frequency and winrate?
EDIT 2: Nevermind, didn’t work for me before, but i can see the frequency of it now (it’s labeled when hovered over)
They put 1,6% for top1k, still no winrate to be found though
THIS is something people need to research. 90% of US mass media is owned by 6 corporations, and the majority shareholder’s views are the ones you will see represented.
Abandoning the Fairness Doctrine was yet another Right wing way to undermine the entire system.
First,
for how Power Score is computed.
Second, VS does not use “actual” winrate, they use “estimated” winrate. (“Actual” and “estimated” are their terms, not mine.) Actual winrate is calculated by just taking all the games played by one deck in the data, divide wins by games played, done. VS does NOT do this.
Instead, they calculate every MATCHUP winrate first. So they’ll have one winrate for Insanitylock vs Handbuff Pally, one winrate for Insanitylock vs Rainbow Shaman, etc. Then they’ll take the data for every opponent that someone with the tracker played against, and use that to determine the popularity of Handbuff Pally, Rainbow Shaman, etc. With those two types of data, they calculate overall winrate as the sum of the products of matchup winrate and opponent deck popularity. They call this “estimated” winrate.
The sumproduct of matchup winrate and opponent deck frequency is a completely valid method for calculating winrate. (More valid, IMHO.) This is the context of the part of the FAQ that you quoted.
This is enough for me. We’re done.
It’s a shame you’re not playing anymore so you can see it with your own eyes. The deck which doesn’t exist found on a tier list in tier 1.
I just wish more trolls here had a chance to play in top 1k to see for themselves, but I guess the game is too hard for them.
Pity they’re still able to write just the same.
If Altair was telling me that Vicious Syndicate was wrong about Sludgelock in T1KL — and specifically Sludgelock in T1KL — I would believe Altair over VS hands down. Not a problem. I have general confidence in Altair’s analytical skills (which are different from statistical skills) and he probably has the higher sample size.
The problem is that Vicious Syndicate isn’t talking about Sludgelock, at all, yet Altair is saying that VS is nonsense. VS will always have a larger sample size than Altair regarding any deck that Altair isn’t maining.