3 ms·
you guys are calculating different numbers. you just calculated the chances based on the assumption that 1/100 children will be abducted each year. Whereas the
by oculusthrift 9y ago
you guys are calculating different numbers. you just calculated the chances based on the assumption that 1/100 children will be abducted each year. Whereas the other poster was saying the average child will get abducted once in every 100 days
- castle-bravo 9y agoThis sub-thread is an exercise in showing how absurd the 1% chance of getting abducted (on a given day) would be. Don't take it too seriously. That said, you're wrong, and nobody is ever allowed to be wrong on the internet. I will now proceed to correct you. > 99% of the time everything will be okay I took this as meaning that 99/100 trips are without incident. I then assumed that a kid might take a trip every day. The probability of taking consecutive 365.25 trips without accident are then 0.99^365.25 (odds of surviving the first trip times odds of surviving the second trip and so on...). This works out to 0.025, or 2.5%. So, what I calculated is the probability that a kid will avoid accident in a year, given that only 99/100 days are without accident. The takeaway is that a kids chances of getting abducted on a given day are much, much smaller than 1%.
- make3 9y agoHe's wasn't trying to contradict me. The things we say can be (and are) both true at the same time. I took the approximate average of a 365 tries binomial of 1%, and he computed the probability distribution function of this same binomial for X=0.