sample size that does not collect enough data has an effect on the score.
What sort of sample size are you thinking?
The larger the sample the better the sample is. But sence I hypothesize the lack of rigging I would say no matter the size of your data it will always have a p-value >0.05.
So the main data set available without a sub contains around 2 million games. It is disingenuous to say this is a statistically insignificant sample.
But I suspect you didnāt look before offering a guarantee.
ā¦otherwise you would realise the data is aggregate and unsuitable for calculating a meaningful p-value.
So it isnāt this simple, actually.
Without writing a treaty on stats, the larger the sample the more likely any difference will be statistically significant.
This is why significance testing is only one part of understanding data. The p value tells you nothing about effect size or meaningfulness of observed differences.
Also because I dont believe the game is rigged, and am only on team rigged due to a severe skills shortage, itās not appropriate for me to execute the experiment as my bias may affect the result
Yaāll could have ended the thread a few dozen comments ago when op told he has no understanding of math by expecting a similar spread betwen enemies.
I ran out of Netflix shows
Ok. Letās see.
In the other thread:
you made a similar claim based on a sample size of 19 that your 11-3 Zoo deck was then countered by 5 consecutive Control Warriors thus āprovingā a rigged system.
Iāve just done you one better. I played 20 games with my Quest Hunter. 10 without Rustrot Vipers and 10 with two Rustrot Vipers. In the 10 games without the Vipers, I faced one Rogue and one Shaman. In the 10 games with the Vipers, I faced one Rogue and two Shamans. How do you explain this? According to your theory, the game couldnāt be rigged because when I added the Vipers, I actually faced more weapon decks.
Of course, all of this is nonsense because a sample size of 20 is woefully inadequate to draw any reliable conclusions. But I was happy to humor you with a brief test.
I should say I made an error in what I said earlier about using data averages of the whole meta it can be used as a base for generating a p-value. You could look at the data percentages as one class vs all other classes to generate a p-value.
Yes I am trying to keep it simple. I should have said the larger sample is more significant, but with smaller sets of data a p-should be able to account for that and will give a larger number.
You donāt need to keep it simple. Iām 47 with multiple science degrees and a long career.
I donāt see any need to calculate P values in this case. Maybe if something weird turned upā¦
It is a lot easier to type on a phone if you keep what you are writing as simple as possible.
Yes there is no similar because the meta changes depending what hero you play
I havenāt played 100 games yet this season, but hereās today.
Looks super rigged, lol.
https://imgur.com/a/V6Twai3
Thats only 1 hero, and yes thats exactly what i said, the game is giving you a mage meta with missiles and ooze when youre playing a libram pala. Now try something thats good vs mage and you dont see them anymore
ā¦ libram paladin has a favorable match up to mages. All of them.
And I havenāt seen a ooze. One rustrot viper, but I won that game with a smite otk, lol.
Wrong, these players you faced were just bad, missiles and mass poly on librams and you lose gg, try quest hunter or quest warrior and you will see 0 mages in next 10 games
Since I am matched by mmr with bonus stars, they are just as bad as I am, my friend. Iām a dumpster legend player when I care to finish the grind to legend, but most months I stop climbing at D5 and play memes.
Your whole point is that I should have been countered on that win streak and nada.
Maybe your bad mouthing them has turned on the switch that makes you not win and I get free wins for days because I am the supreme commander of the defense force for the actiblizzards?
I usually wear a tinfoil hat when I play, so maybe I am just blocking the rigging waves from coming out of the tubes the internet goes in.
Only a bad player stacks his librams into mass poly against mage. You play like they are going to cast it and wait for it to come out. You know itās going to get cast again by hero card. You just play around them, lol.
I can literally choose what i want to play against, if i dont want to play vs mages i play quest warrior or quest hunter
If i dont want to play against paladins i que mage or priest its so simple.
Feeling like not wanting to play against quest warriors? play paladin or taunt druid
Op, there are a hundred people in a room
70 are red.
22 blue
And 1 each from 8 other colors.
Do you think that if Iām pulling 10 people randomly out there Iām gonna have 1 from each color?
Better yet: do you think every class is played by the same amount of players? Meaning, do you think every class is played by 10% of players at any given time?