The meme is talking about a common probability error that surveys have shown even doctors are prone to making.
Why you’re probably ok:
The rarity of the disease far exceeds the error rate of the positive test. Meaning, the disease occurs in 1 out of a million people, so if you are tested at random and show positive, you only have a 1 out of 30,000 chance (the 3% false-positive rate) of being the the 1 person who truly has the disease.
You must log in or # to comment.
Tactical RPGs have basically taught me that anything below 100% is almost always a miss and even 100% isn’t guaranteed.
I missed a very important 96% shot in XCOM last night. I’m still thinking about it.
The doctor is the one with the correct reaction there. Go do the second test.
I mean, yes…
But at 1/30,000 , they should say “get the second test… but be SUPER CAREFUL on the drive”, since at 1/30000 you’re still an order of magnitude more likely to die in an MVA.
This is ideal bedside manner.
I’ve always thought that I’d make an exceptional professional in the field of medicine.
The only thing really holding me back is my unfathomable depth of ignorance regarding the human body, or health in general.
At one point in my life, I believed that to be a deal breaker. Cheers, RFK Jr.
I think RFK can be convinced to believe that Windex has amazing health benefits, so this might be your chance!
Interpret it how you will.
Accuracy and False Positive Rates are two different numbers.
I’m tired and my brain is being dumb right now, but when you said that my first thought was of course American. 97% accuracy grouping bullets is a lot different than 97% sure a gun was fired.
One says Johnny got shot in the kidney, the other says a truck may have misfired down the road.
This is why we use specificity and sensitivity stats for medical tests. If the test has a sensitivity of 97%, you should definitely be worried.
In the case of trying to minimize false positives, you want the specificity to be high, not necessarily the sensitivity, which is associated with false negatives.
And 97% specificity with a very low pretest probability still results in a low probability for disease, which is why screening for so many diseases is difficult, even if diagnosing them can be easy if there are clinical signs and symptoms in addition the the test. The clinical background can increase the pretest probability significantly, allowing the test to do its job.
Another very relevant video from 3Blue1Brown about the problem.
This is one of the main reasons doctors don’t ‘just give you a battery of tests’. Not only is that expensive, but if you are running dozens of tests, the chance one of them gives a false positive is pretty high. So now you not only wasted a pile of money, but you also think you have some rare disease you don’t actually have. So you waste even more time and money treating that disease you don’t have.
Doctors run tests for things they think you might actually have, which diminishes the false positive chance.
I almost died of a dental abscess back in 2008, which led to a multi systemic failure. That was fun, but I’m still alive today.
Fuckall with worrying about life anymore, if I ain’t dead yet, well I’m not dead. I’m doing okay BTW…
Any day we’re all breathing is a good day
Be well, friend
I mean, fuck 2026 and all but I can have a good day if I have good food.
Today I broke my personal record for consecutive days lived!
Whoa, congrats! I should check my high score too… OMG would you believe it?! Me too!
Amazing I love it! Congrats friend!
Outlier?
What statistician is this referring to? Certainly not one who understands probabilities. The first number has nothing to do with it. You tested positive, and there’s only a 3% chance that result is wrong. Time to settle your affairs.
In a sample of 1 million people, 1 person will have the disease, 30,000 however will test positive for having the disease. Notice how the false positives count is way higher than the actual positive count.
Is 97% accuracy rate the same as a 3% false positive rate? It might be a combination of false positive and false negative rate.
Accuracy is defined in relation to a specific population or dataset with a specific rate of disease, not for any individual. To properly characterize the test, you need to know the specificity and sensitivity, and together they tell you how a test will perform on an individual and how much an individual’s pre-test probability increases in the case of a positive test or decreases based on a negative test.
Don’t worry if it’s confusing, Baysean statistics is often counter-intuitive.
If you’re interested, here is a very good 3Blue1Brown video that explains the concept very well.
Thank’s for the link. Probability and statistics in general is not intuitive to me, not just for this type.
How does that matter if I have a 97% chance of actually having the disease? A lot more people than I have won the lottery, doesn’t have a thing to do with whether I will.
Its right 97% of the time. That does not mean you have a 97% chance of having the disease. The 3% error rate accounts for significantly more false positives than it accounts for false negatives on a disease that’s 1 in a million. Again, with a 3% error rate, there will be 30000 false positive test results in a million. 30000 in a million is a larger number than 1 in a million.
As far as I can see, you can’t really fear or rejoice with the results until you know the false positive/negative ratio.
That slop picture in the middle is uncalled for, I mean there are limits
