So it appears you have to accept that according to their nature these models aren’t to be relied on. — AJJ
The discussion was about people refusing the vaccine out of fear of risks like stroke and death. Those risks are minuscule -- no matter how you slice the data. They remain so.
— Xtrix
Those minuscule risks don't simply translate into minuscule strokes or minuscule deaths. — baker
Because you brought up the fact that people are having strokes. So while you may not make this argument yourself (as I would assume, given you’re vaccinated), I assumed you were bringing it up to demonstrate how others may be reasoning about this. If that’s not true, I wonder why you brought it up at all?
— Xtrix
For one, because your position lacks empathy. — baker
Of course I'm scared. — baker
And what do you have to offer to me as consolation? Luck?!! — baker
Also, what’s the consolation for the millions who have died of coronavirus?
You need consolation for those people? — baker
I'm talking about proper risk analysis, not whatever you just did. My claim is that it's not actually risk-based. What you call it, or think it is, is irrelevant. — Isaac
Maybe it goes slightly above or below overall numbers — but not by much. Why?
Because 150 strokes out of 10 million people, for example, is astronomically low.
— Xtrix
Show me the maths then. What is it about 150/10,000,000 as a prevalence rate which makes it impossible for any cohort to have a high risk. As far as I can see there's a potential cohort of 150 for whom the risk is 1. — Isaac
If it turns out that 90% of those 150 people were over 65, that’s important to know — no doubt (especially if you’re over 65). Does that significantly change the overall odds? As I mentioned before: no, it doesn’t. It simply means if you’re over 65, you have a slightly greater chance of having a stroke after taking the vaccine.
— Xtrix
That is changing the odds. It's literally what changing the odds is. You've taken one odds (the national prevalence), and you've changed them to get the risk for a 65 year old. — Isaac
it doesn’t change the odds much at all — perhaps by 0.00001% or something to that effect.
— Xtrix
For some variables that may well be the case. For others we know it's much higher. Obesity, for example has an OR of over 13. Age above 65 even higher, making your estimate more than a thousand-fold out. — Isaac
How do I support this claim? With mathematics — which can be checked by everyone.
— Xtrix
I've yet to see any mathematics, despite several requests. — Isaac
Since you can always gather more information, by your definition nothing is risk-based. — Xtrix
I was discussing was strokes -- whatever your discussing, I can only guess. — Xtrix
Let me try one more time: 150/10,000,000 = 0.000015%. That's some pretty easy mathematics. — Xtrix
Let's say everyone in that group was over 65 -- what would someone's, age 65 years or older, odds be of getting a stroke in that case? — Xtrix
it's being fed to us by the government. What is worse, we can not communicate with the government, the government does not discuss with us. — baker
That I'm smarter than you is obvious. — Benkei
Make no mistake: I despise China, but I find less fault with China than with the Westeners who in their greed gobble up whatever China throws at them. — baker
The Chinese company that wrote the software, Shanghai Adups Technology Company, says its code runs on more than 700 million phones, cars and other smart devices. The episode shows how companies throughout the technology supply chain can compromise privacy, with or without the knowledge of manufacturers or customers. It also offers a look at one way that Chinese companies — and by extension the government — can monitor cellphone behavior. For many years, the Chinese government has used a variety of methods to filter and track internet use and monitor online conversations …
Since you can always gather more information, by your definition nothing is risk-based.
— Xtrix
Risk is determined by variables. Assessing the impact of those variables is a risk-based decision. Ignoring them is not. It's nothing to do with always being able to get more data, it's about what we do with the data we've already got. — Isaac
I'm asking how you get the risk from the prevalence. You've just divided the total cases by the total population of the sample. That gives the prevalence. I'm asking for the maths you're using to get from there to the risk. — Isaac
Let's say everyone in that group was over 65 -- what would someone's, age 65 years or older, odds be of getting a stroke in that case?
— Xtrix
It would depend on their measures for any known variables affecting the likelihood of strokes - high blood pressure, atrial fibrillation, smoking, drinking too much alcohol, poor diet, a close relative who has had a stroke, high cholesterol, diabetes, being overweight, sickle cell disease, frequency of migraine with aura. All of these factors have ORs, you multiply the prevalence by the combined ORs for the person (combined dependant on co-variant factors). That's the risk. If we don't know the ORs, then failing to take them into account is irrelevant since they could be anything. If we know the ORs but ignore them, you're not basing your decision on risk anymore. — Isaac
At what point does the data on death from heart disease go from "prevalence" to "risk", exactly? As I said before, if you narrow the range it's still the prevelance -- just a more specific prevalence (like the prevalence of dying from heart disease for people over 65 and male versus overall prevalence). — Xtrix
If we don't know the ORs, then failing to take them into account is irrelevant since they could be anything. If we know the ORs but ignore them, you're not basing your decision on risk anymore. — Isaac
If I know all the factors determining the fall of a coin — Isaac
So the odds (chance/risk whatever term we use) are a measure of my uncertainty, whist the prevalence is a measure of the occurrence in a population. — Isaac
You asked where we stop adding variables. Never. We include all variables. — Isaac
If that’s what’s are restricting “risk analysis” to, then it doesn’t exist. What you’re talking about in that case is certainly. — Xtrix
They’re both odds. — Xtrix
No. You cannot include all variables because, as I mentioned before, there is a nearly infinite range of variables we can control for. — Xtrix
You know, the guy named Bob who’s got red hair and saw Star Wars in the theaters— all known variables. What about him? What’s HIS specific odds? — Xtrix
They’re both odds.
— Xtrix
No. Here's a primer on the differences. https://www.cdc.gov/csels/dsepd/ss1978/lesson3/section2.html — Isaac
No. You cannot include all variables because, as I mentioned before, there is a nearly infinite range of variables we can control for.
— Xtrix
So? How does that affect the maths I provided? Each one of the infinite range of variables which we don't know about has an equal chance of increasing the risk as it does of decreasing the risk, so including them is a matter of multiplying each probability by the uncertainty (0.5*p + 0.5*p). I did write all this out in my reply, if you're not going to bother even reading it, there's no point in replying. We are including all the unknown variables in our measure of uncertainty (risk). What matters here is deliberately not including a known variable. — Isaac
it’s an excellent primer indeed, and saying exactly what I’ve been saying the entire time — Xtrix
There’s an infinite number of KNOWN variables as well —or at the very least in the hundreds of millions of combinations for an individual. — Xtrix
X and y are both odds of dying of a heart attack. — Xtrix
Where does it say that the prevalence and the risk are the same? Provide the quote that you think supports your view. — Isaac
Synonyms for incidence proportion
Attack rate
Risk
Probability of developing disease
Cumulative incidence
Incidence proportion is the proportion of an initially disease-free population that develops disease, becomes injured, or dies during a specified (usually limited) period of time. Synonyms include attack rate, risk, probability of getting disease, and cumulative incidence. Incidence proportion is a proportion because the persons in the numerator, those who develop disease, are all included in the denominator (the entire population).
Example A: In the study of diabetics, 100 of the 189 diabetic men died during the 13-year follow-up period. Calculate the risk of death for these men.
Numerator = 100 deaths among the diabetic men
Denominator = 189 diabetic men
10n = 102 = 100
Risk = (100 ⁄ 189) × 100 = 52.9%
There’s an infinite number of KNOWN variables as well —or at the very least in the hundreds of millions of combinations for an individual.
— Xtrix
OK, so for a stroke, say, give me the first twenty or so, a list with the ORs for each. — Isaac
X and y are both odds of dying of a heart attack.
— Xtrix
You can't have two different odds of the same event. — Isaac
I've got a question I'm far more interested in, if you'll indulge me - What do you think is happening here? This conversation we're having. — Isaac
If it's all about risk profiles, then help me make my choice. What are my numbers? [...]
Because if you can't produce figures for my risk then my decision is not risk based is it? — Isaac
You've made clear what you imagine my politics and motives to be, but you've left out my education level, profession, age... I'm just intrigued as to how you're putting this all together. — Isaac
Also, whilst I'm just asking, what's your role in this storyline? How do you see this ending, for example, what's the coup de grâce with which the hero slays the dragon here? — Isaac
Notice the last line. Also recall my repeating the 150/10,000,000 as a measure of risk. This is saying exactly the same thing. — Xtrix
Winning the NBA Championship is an event. Lebron James' odds of doing so are much greater than mine, alas. Same event, different odds. — Xtrix
it's important to note that your odds of contracting ovarian cancer are zero if you're male. — Xtrix
The data is there -- look it up yourself. — Xtrix
What is the ultimate thesis here? That you cannot measure the risk of COVID? That looking at the "prevalence" of a disease is unrelated to risk? I have no real idea — Xtrix
Get involved in philosophical discussions about knowledge, truth, language, consciousness, science, politics, religion, logic and mathematics, art, history, and lots more. No ads, no clutter, and very little agreement — just fascinating conversations.