One of my graduate degrees is in information science. It's time to stop talking down to me like I have no idea what you're trying to say.
You are not "sampling" a population through VAERS data, because you cannot confirm literally anything about the VAERS population. Samples must be verifiable, and VAERS submissions are not verified.
Law of large numbers states that the larger the number, the more represented the outlier cases.
Which means that you are correct that IF the vaccine is capable of causing damage, then it will be represented in a large enough sample.
What you haven't established, and cannot establish, is what percentage of that sample population will represent outliers.
It may take a hundred million cases to even find a single vaccine injury. It'll be represented in a large population, but it'll still be a vanishingly small risk.
I can give you a bag of 1,000 coins, and you might feel pretty good about the representation of coins in there. Is there a good chance there's a gold dollar? Sure. A fifty-cent piece? Sure. They're rare, but they might be there.
But what about a buffalo penny? What about a misprinted quarter? Just because I have a large population of coins doesn't mean I have any reason to assume these very rare coins are present in the population. I can't make many assumptions about how many there are.
If I have 100,000 coins, or 100,000,000 coins, then I have a better chance of finding a very rare coin, but it'll still be a very small number of them compared to the total number of coins.
You're looking at VAERS data and assuming that if people are submitting, then at least SOME must have a legitimate vaccine injury.
This is incorrect.
There is no reason to believe this outside of the fact that you ALREADY believe that the vaccine causes injury. You are permitting that conclusion to taint how you are evaluating your data and guide your assumptions.
You assume the vaccine is hurting people and it's being covered up, therefore, 95% of the VAERS reports must be legitimate, despite what the actual verified data suggests.
But there is no reason to assume only 5% of the data is garbage. None. Nothing empirical you can provide. The law of large numbers only states that existing outliers will be represented in a large enough sample, not that you can make any assumptions about the proportion of the outliers to the total population based on that data alone.
So how are you verifying that you data is 100% accurate? How can you verify, empirically, even ONE vaccine injury using the VAERS data?
Can you point to something in VAERS that confirms even a single number it reports has been verified as a vaccine injury? Can you show me that in the raw data, please?
Watch how your argument is working.
"There were 15,000 gun deaths in the US last year. Trump supporters are a huge problem."
"What? What does this have to do with Trump supporters?"
"Well, 15,000 gun deaths? Trump supporters are conservatives and lots have guns. Law of large numbers. Trump supporters must be causing a lot of deaths, even if a few of those gun deaths don't apply to them.
"Can you verify that even a single Trump supporter killed anyone with a gun last year? Does the data break out gun crime by political affiliation?"
"Well, no. But seriously, if there are that many gun deaths, and Trump supporters like guns, then it must be Trump supporters that are the problem."
You see the logical flaw here? I can assume all I want about the data outliers, but that's not science. That's me wanting to believe Trump supporters are violent psychopaths, and making assumptions about data that cannot confirm my theory, based on the "law of large numbers."
I have nothing to support the notion that Trump supporters are the big problem of gun crime, and you have nothing to support that vaccine injuries have any significant representation in the VAERS data.
Lots of people died from medical conditions after getting the vaccine. It is impossible for you, using VAERS data, to establish that any single one of them is due to the vaccine, and therefore, the law of large numbers offers you NOTHING you can use to establish a proportion to assume how represented the outliers are.
I don't think you're responding to my point, though.
Let's try this. Do some data science on this.
I just generated a list of 10,000 cars from a random website where people in the US list the cars they own. Please tell me accurately how many times a 2020 Porche GT Carerra shows up on my list.
You get no further information than the fact that there are 10,000 cars, and the people live in the US. Use your law of large numbers.
I don't lie here, but you insist I do, because if I'm not, then that might mean I actually do know what I'm talking about, which means you might be wrong.
I get it. But it's hard for you to be intimidating on the internet, especially since you've already made attempts to scrape my IP address and so forth. I'm more interested in talking than in dick-measuring contests, which tend to be where you want to go with this when you start puffing up your chest and calling people "boy."
One of my graduate degrees is in information science. It's time to stop talking down to me like I have no idea what you're trying to say.
You are not "sampling" a population through VAERS data, because you cannot confirm literally anything about the VAERS population. Samples must be verifiable, and VAERS submissions are not verified.
Law of large numbers states that the larger the number, the more represented the outlier cases.
Which means that you are correct that IF the vaccine is capable of causing damage, then it will be represented in a large enough sample.
What you haven't established, and cannot establish, is what percentage of that sample population will represent outliers.
It may take a hundred million cases to even find a single vaccine injury. It'll be represented in a large population, but it'll still be a vanishingly small risk.
I can give you a bag of 1,000 coins, and you might feel pretty good about the representation of coins in there. Is there a good chance there's a gold dollar? Sure. A fifty-cent piece? Sure. They're rare, but they might be there.
But what about a buffalo penny? What about a misprinted quarter? Just because I have a large population of coins doesn't mean I have any reason to assume these very rare coins are present in the population. I can't make many assumptions about how many there are.
If I have 100,000 coins, or 100,000,000 coins, then I have a better chance of finding a very rare coin, but it'll still be a very small number of them compared to the total number of coins.
You're looking at VAERS data and assuming that if people are submitting, then at least SOME must have a legitimate vaccine injury.
This is incorrect.
There is no reason to believe this outside of the fact that you ALREADY believe that the vaccine causes injury. You are permitting that conclusion to taint how you are evaluating your data and guide your assumptions.
You assume the vaccine is hurting people and it's being covered up, therefore, 95% of the VAERS reports must be legitimate, despite what the actual verified data suggests.
But there is no reason to assume only 5% of the data is garbage. None. Nothing empirical you can provide. The law of large numbers only states that existing outliers will be represented in a large enough sample, not that you can make any assumptions about the proportion of the outliers to the total population based on that data alone.
So how are you verifying that you data is 100% accurate? How can you verify, empirically, even ONE vaccine injury using the VAERS data?
Can you point to something in VAERS that confirms even a single number it reports has been verified as a vaccine injury? Can you show me that in the raw data, please?
Watch how your argument is working.
"There were 15,000 gun deaths in the US last year. Trump supporters are a huge problem."
"What? What does this have to do with Trump supporters?"
"Well, 15,000 gun deaths? Trump supporters are conservatives and lots have guns. Law of large numbers. Trump supporters must be causing a lot of deaths, even if a few of those gun deaths don't apply to them.
"Can you verify that even a single Trump supporter killed anyone with a gun last year? Does the data break out gun crime by political affiliation?"
"Well, no. But seriously, if there are that many gun deaths, and Trump supporters like guns, then it must be Trump supporters that are the problem."
You see the logical flaw here? I can assume all I want about the data outliers, but that's not science. That's me wanting to believe Trump supporters are violent psychopaths, and making assumptions about data that cannot confirm my theory, based on the "law of large numbers."
I have nothing to support the notion that Trump supporters are the big problem of gun crime, and you have nothing to support that vaccine injuries have any significant representation in the VAERS data.
Lots of people died from medical conditions after getting the vaccine. It is impossible for you, using VAERS data, to establish that any single one of them is due to the vaccine, and therefore, the law of large numbers offers you NOTHING you can use to establish a proportion to assume how represented the outliers are.
I don't think you're responding to my point, though.
Let's try this. Do some data science on this.
I just generated a list of 10,000 cars from a random website where people in the US list the cars they own. Please tell me accurately how many times a 2020 Porche GT Carerra shows up on my list.
You get no further information than the fact that there are 10,000 cars, and the people live in the US. Use your law of large numbers.
See, yet another thing you do that makes me question your commitment to this movement.
"Don't question me. I am an authority on this subject."
Doesn't sound very Q of you.
I don't lie here, but you insist I do, because if I'm not, then that might mean I actually do know what I'm talking about, which means you might be wrong.
I get it. But it's hard for you to be intimidating on the internet, especially since you've already made attempts to scrape my IP address and so forth. I'm more interested in talking than in dick-measuring contests, which tend to be where you want to go with this when you start puffing up your chest and calling people "boy."