mr.skin wrote:It s easy statistics, the probability stays the same for each event, but if we have a set of 5 independent events thus the probability that one of them comes true is the sum of each probability.
OK, I will make a liar out of myself. I HOPE this will be my last post on the subject in this thread.
If you take a 20-sided (Dungeons and Dragons) die, the likelihood any one face will come up on a single roll is 5%.
If you roll that 20-sided die 20 times, the likelihood that any one face will ever come up is NOT the sum of 20 5%s - 20 x 5 or 100%, is it?
By my calculations the odds that any one face will come up on any of the 20 rolls is about 64% 36% chance that one face will not appear ever, even once, in the 20-roll passage.
5 revisions means a total risk of 25-50% of ever experiencing an infection.
If I am right, a man undergoing five operations with a 5% risk of infection ON EACH ONE has a 22.6% chance of acquiring an infection. Put another way, he has a 77.4% chance of going through all the operations without an infection.
With a 10% risk of infection on each one, a series of 5 operations would put him at risk for ever having a single infection in the series is 59% or 41% chance of have no infection at all during the series of 5 operations.
Or, flipping a coin at 50% chance of "heads", you have a 25% chance of two flips being all tails of 25%. 3 flips 12.5% 4 flips 6.25%
here is a spreadsheet calculation anyone can cycle through: First a 50% example at 2 flips of a coin, then a 2 cycle example at 5% risk followed by a 5 cycle example at 5% risk
The coin flip example is pretty intuitive
risk of an outcome (infection or "tails") upon a specific event (revision operation or coin flip)
50.00%
1 minus risk %
Likelihood of escaping that event
50.00%
2
number of events (or "n")
1-risk raised to "n"
likelihood of escaping that outcome throughout the series of events (none of four tosses coming up "tails")
25.00%
The infection example uses the same formulas
risk of an outcome (infection or "tails") upon a specific event (revision operation or coin flip)
5.00%
1 minus risk %
Likelihood of escaping that event
95.00%
2
number of events (or "n")
1-risk raised to "n"
likelihood of escaping that outcome throughout the series of events (none of the operations resulting in infection)
90.25%
risk of an outcome (infection or "tails") upon a specific event (revision operation or coin flip)
5.00%
1 minus risk %
Likelihood of escaping that event
95.00% (or 5% chance of experiencing an infection)
5
number of events (or "n")
1-risk raised to "n"
likelihood of escaping that outcome throughout the series of events (none of the operations resulting in infection)
77.38% (or 22.62% chance of experiencing an infection at some time during the series of operations)
so, the more times you have an operation, the more likely you are to experience an infection throughout the entire series, but still nowhere near 100%.
I wonder which is more satisfying, beating this question to death beating a dead horse or beating a dead penis?
or less satisfying.