-->
@Ramshutu
Putting aside the obnoxious quote ladders that make it near impossible to engage In a discussion for a moment;
Tone arguments are unwelcome.
the above post entirely, and the quoted section specifically illustrates that your understanding of statistics is profoundly failing you at a fairly fundamental level - so much so it’s hard to know where to begin. But let me try.
Do your best.
A regular dice with six sides, has a 1/6 chance of turning up any of the individual numbers.
Unlike dice, people exhibit autonomous behavior that aren't restricted to a six-sided parameter.
Just because I don’t know what the true probability of rolling a six given the specific conditions at the time, or for a specific dice for which the probability of its population is known does not alter the fact that the chance is 1/6;
Because "rolling" dice are governed by their six-sides and the act of "rolling".
Probability is not rendered invalid because of unknown conditions and factors
Non sequitur.
probability is literally an expression of those unknown conditions and factors when only composite data is known
No, probability is the estimation of an event's occurrence given known conditions. It doesn't express unknowns.
Or to be more specific to vaccines, given that I can smell a strawman coming : You could take a warehouse full of random second-hand dice; each of which you can only roll only once; take a large random sample; weight half and roll them all. You have to make generalizations about an individual based on population without being able to control for outcome at an individual level.
Why do you "have to"? Do you have to, or is it just a method to which you've grown accustomed?
The most important thing is we know people vary so we need to know whether a sample is appropriate. If I take from one side of the warehouse, I may not get an accurate representation of the dice. Likewise the age of dice, may play a role, top or bottom of a box, Color, size; but we can adjust for factors we know make a difference
And how have you determined which factors "make a difference?"
if we do that and randomly sample account for this; the remaining subset of randomly chosen dice should be representative - and will include enough of various factors of all individuals to be representative -
Presumption with no substantiation. Representative of what? The individual or the parameters dictated by the sample?
This is not to say the sampling is perfect;
No need to express the imperfection of sampling; my contention isn't against the notion of perfection; my contention is against the irresponsible inferences based on fallacious reasoning.
but it’s testable
What is "testable"?
and, thus far, is based on known biological factors yielding predictive power - scientific.
KNOWN BIOLOGICAL FACTORS =/= HOMOGENEOUS IMMUNOLOGICAL RESPONSE.
The second weighted sample allows you to establish the change in baseline probability after the controlled parameter has been added. Both samples will contain similar populations of weighted dice, or ones for which the weighting will not work - you’ve corrected for age and size and colour, so providing picking any random 100 large black dice from the bottom of boxes at the far end of the warehouse is equally likely to turn up loaded size as any other random selection of 100 large black size from the bottom of boxes at the far end of the warehouse - those numbers won’t be appreciably differentIf you pull a new random dice from the pile ; you can say with some confidence that the chance of you rolling a six increases by a certain amount if you weight the dice. This is true even if you don’t know the specific conditions of that dice. This is because the possible specific conditions and their probability of that dice being non standard is part of the baseline.The increase in chances of rolling a six incorporates both of the possibility of you picking up a dice for which weighting has no impact, and one for which it does.For example - if there are no weighted dice in the warehouse, weighting would improve chances of rolling a six by 100x, if there are weighted dice that are unaffected by more weighting - that number would go down to say - 80x.Rolling Any given dice would be 80x more likely to turn up a six: a probability which incorporates the probabilities of a.) a weighted dice not rolling a six by chance, b.) a weighted dice turning up a six because of the intended action of weighting and c.) a weighed dice not showing a 6 because of some ineffectiveness if the weighting process due to the dice. The statistics applied to the individual is based on the principle that this dice is a member of the broader sample population; and is an expression of the break down of various conditions in that population. When you talk about the probability of the individual - that probability incorporates all those unknowns due to that sampling.
Once again, Dice aren't autonomous.
For Covid, if the illness and survival rate for your risk group (age, weight, etc), is 10% to become sick, and 0.1% to die, and that same risk group reduces to 1% and 0.01% - then your risk has dropped by a factor of 10.You could have some unknown predispoition that means you will die with or without the vaccine; but it’s also more likely you have a predisposition that means you will live with the vaccine and die without; the statistics tells us that the prevelance of the former is at most 0.1% and latter is around 0.9%.
We've finally arrived at the actual argument. How is survival rate determined? What is "predictive" about these rates? If 10 percent of those with whom I've been arbitrarily grouped whether based on age, weight, ethnicity, etc. have contracted a virus and 1/100 of them died, what does this predict about me? The sample is heavily reliant on the ASSUMPTIONS of its parameters--notably, ceteris paribus, the results of INSTANCE being reproducible.
Unless you actually know what that predisposition is, and whether you have it - it’s only possible to express it as a probability determinable by a population - a probability incorporated into the quantified risk reduction. Your risk has still reduced by 0.9% because that is the probability of having a vaccine preventable disposition - even while the 0.1% remains.
Limitations, again, don't speak to fact.
Likewise with seatbelts. Seatbelts will improve your chances of survival. Not because any specific crash you are in yield less chance of dying, but because some types of crashes you can get in will be survivable with a seatbelt;
Circular reasoning. Your conclusion is the same as your premise.
the chances of occurrence of those types of crash can be determined, with the probability of risk incorporating your chances of getting into one of those, vs one where a seat belt will not help. As it is not possible to tell or control all the individual factors - it must be expressed as a probability based on occurrence within a population controlled for known factors.
A population which does not speak to individual autonomous behavior. People are not dice.
Likewise, drunk driving ; the absurdity I point out here is down to your failure to appreciate the meaning of the numbers.
That is an impression; it's neither an observation nor a mode of logic.
The statistics do not imply that on a given drive home on Sunday, where a specific accident - say a truck plows into you from behind - would be more or likely to occur whilst drunk or sober; but because there are a subset of accidents which can be caused by being drunk or prevented by being sober that have a given probability of occurring based on population statistics. The increase in risk from drunk driving incorporates that general change in risk given that it’s not possible to calculate or know all the factors to know the exact per journey risk - in exactly the same way that you can’t calculate all relevant physics for a dice.
Once again, people aren't dice.
That’s how probability works. The boiling down of unknown events in terms of a likelihood of occurring. For risk statistics - it’s all baked into the numbers. It’s the mechanism by which insurance companies can consistently and reproducibly make billions of dollars; by accurately assessing risk of an individual by virtue of analyzing the population they are part of.
...I wouldn't know anything about that, now, would I?
I would highly suggest you find an actuary, and suggest that the risk to an individual cannot be determined by assessing occurrence in a population. If you’re lucky, they will be incapacitated from laughter so long you could steal their jaguar.
Why would I need to find an actuary. If I'm going to appeal to authority, then I could do just well, appealing to my own.
You can’t complain
Once again, WHERE DID I COMPLAIN?
that your specific set of conditions that determine whether you will live or die are unknown so you cannot speculate as to the efficacy of a vaccine
Non sequitur. It isn't my contention that a "specific set of conditions that determine whether you will live or die are unknown..." It is my contention that results of population sampling mean squat to an individual because the sampling results are dictated by the parameters of the sample AND NOT THE INDIVIDUAL HIM OR HERSELF. AND FOR THIS REASON, SAMPLING RESULTS CANNOT BE PREDICTIVE for any particular individual.
because expressing things you don’t know based on their chance of occurrence in a particular scenario is the whole freaking point of having probability and statistics in the first place.
Once again, probability doesn't express "unknowns"; it's a method of estimation using known conditions.