I don't understand this meme. If the surgery has a 50% chance of killing you and the 20 previous patients survived, doesn't that mean that you have a 0.00005% chance of surviving? the chance of 21 successful operations in a row is 0.00005%

>>16712840 (OP)
No idea what the intended interpretation is, but the 50% survival rate is averaged out across all surgeons. Clearly this surgeon is a gigachad who never misses

>>16712840 (OP)
It's like one of those trick questions on a math olympiad test where it asks you to predict the odds of flipping a coin or rolling a die that's come up the same way 50 times or some shit. You're supposed to intuit that there other information you're not privy to that is skewing the results. In the case of the math question you're supposed to question whether the coin/die/whatever is fair, in the case of the surgery you're supposed to intuit that there is something about this particular facility or surgeon that gives their patients a much-higher-than-average survival rate.

>>16712840 (OP)
No. Think it through. Why would it matter to the patient what the outcome of the previous surgeries was? (Not the success rate implied by them -- I mean literally just the outcomes)
It doesn't matter to him.
It's ONE DICE THROW.

The calculation you are referring to is only interesting to the doctor. It's only, exclusively, about the likelihood of this precise SEQUENCE of events occuring that has such a low chance -- that is, even the rate really were 50%. But the success rate of this doctor isn't actually just 50%, it's currently implied to be near 100%. So that means the chance of a 21 streak of good surgeries is not just 0.00005%, but extremely higher, like 90+%.
It's a SEQUENCE OF DICE THROWS, NOT ONE INDIVIDUAL DICE THROW.

>>16712840 (OP)
It's because of me, sorry. I personally kill every patient I do that surgery on just to keep the stats at 50% and mindfuck everyone.

>>16712974
>Today, Dr. Anon, who people are know calling "The Dentist of Death", was found guilty of first degree murder in the deaths of 1,472 patients who came into their clinic for routine cleanings.

>>16712850

That is the scary part actually. Statistics does not even consider gross incompetency.

>>16712840 (OP)
the surgeries are independent, therefore each surgery has a 50% survival rate

>>16713017
I think its more what >>16712850 said

>>16712868
my first assumption is there was some error in the calculation of the survival rate

same way I would assume that if somebody flipped a coin 20 times and got heads 20 times, I would assume a rigged/unfair coin and not a gigachad coin flipper

>>16713017
>surgeries all performed by the same surgeon are independent

>>16712840 (OP)
>the chance of 21 successful operations in a row is 0.00005%
he got more proficient doe

>>16712840 (OP)
>normalfag
All the previous coin flips landed on heads so this one must land on tails!
>mathematician
Each coin flip is an independent event. There is a 50% chance that it will land on heads.
>scientist
Seeing as the coin landed on heads 20 times in a row, it probably has a higher than 50% chance to land on heads. Maybe the coin is weighted, or both sides are actually heads?

>>16712840 (OP)

normal people see it as there have to now be 20 deaths to align it back. so, for them, odds are bad

mathematician knows that past outcomes should not affect future ones (unless there is some unknown hidden information). for him, odds are 50-50.

scientist knows the rate in real world can not be a fact, but an estimate from past performance only.
20 successes in a row is statistically improbable, which suggests something has likely changed, or perhaps the rate given is incorrect, and odds should be much better than 50%

>>16712840 (OP)
you're falling for the gambler's fallacy. no matter how many times you land a coin on heads in a row, the chance of the next toss being tails stays 50% every time. casinos prey on this stupidity

Normal people:
>damn that means it is sure to fail on my turn because it needs to even out

Mathematician:
>this nigga killed at least 20 people lol

Scientist:
>doctors arent even real phds lol

>>16713147
more like

scientist
>he called me sir!

Gambler's fallacy. The mathematician should be worried because mathematically it's 50% chance you die. The normal people on the other hand might be fooled to think that the operation is safer because of those 20 people who survived. The scientist is happy rightfully because for the scientist experimentation trumps theory.

>>16712840 (OP)
The normie thinks, like you, that the failed surgery is "due" because it hasn't happened in a while. This is purely a logical fallacy and not how it works. The mathematician thinking purely statistically knows that previous trials don't affect the current trial, so he has a 50/50 chance of dying. The scientist thinks that there must be another variable at play making all the patients live so he has a very good chance of living.

>>16713186
That's using the prior instead of the posterior or even the likelihood. You failed stats didn't you?

>>16713186
Dumb meme. The mathematician and scientist would come to the same conclusion, unless the scientist is an idiot social scientist/medfag.

>>16712840 (OP)
The doctor actually uses cryotherapy while their patients are sedated and commits healthcare fraud, billing for fake surgery while getting better outcomes.

>>16713193
No, you did

I always knew mathematicians and scientists have nothing to live for.

>>16712916
In tf2 a melee weapon has a 25% crit chance. If I swing my weapon into the open air three times, then attack an enemy player, why do I get more crits?

It's a simple bayesian model

>>16712840 (OP)
This is accurate.

Normal people:
>previous results means I'm FUCKED
>everybody else all surivived means Imma gonna die to balance out the stats

Mathematicians:
>previous results means NOTHING
>every dice roll is independent; I'm still 50/50 no matter how many people happened to luck out before me.

Scientists:
>previous results means I'm GOLDEN
>whatever experiment that resulted in the documented 50/50 rate clearly didn't model the actual procedure accurately. This many people all surviving means the real rate is probably up there in the 90s. YESSSSSSS.

>>16713460
so you're claiming normal people are dumber than scientists? clearly you have never met a scientist

>>16713423
I don't know. There may be additional factors not relevant to math, like the TF2 devs inserting guaranteed crits after a certain number of non-crits has been reached to increase player satisfaction.

>>16713460
Regarding Scientists, I read it more like:
>previous results mean I'm GOLDEN
>The sampling across all surgeons yields a 50% success rate, but this particular surgeon has a 100% success rate, which means that he is good surgeon bringing up the average rather than a bad surgeon bringing it down.

>>16713141
>>16713161
you guys are kind of stupid. if 20 patients in a row survived, then clearly there are variables that raise it beyond the proposed 50% survival rate (which is a statistical average), like a surgeon developing the skills necessary to perform the surgery with greater competency.

>>16713638
you don't get it

>>16712840 (OP)
normal people = 50%
mathematician applies bayesian prior-posterior method
scientist rejects the entire hypothesis with laughable p-value

>>16712840 (OP)
Each surgery is independent from each other surgery (him performing a successful surgery does not cause him to suddenly become a worse doctor to "even out the odds"). The normal person in the meme thinks like you do (these are also the sorts of people to think that if you roll a die six times, you're guaranteed to roll a 6). The mathematician knows the surgeries are independent events and therefore believes the 50% odds regardless of previous outcomes. The scientist comes back to the probability you calculated, but uses it as a statistical test to conclude that the probability of surviving the surgery is extremely unlikely to actually be 50% (but rather way higher).