What's the point of probability if things go completely against odds?

[TL92]

I mean say there was a 99% chance of getting a green apple on the first pick from a hat, picked randomly. But you don't. Doesn't that negate everything?

 

[Pengwuino]

No, that's the point of probabilities, no result is guaranteed (unless something has a 100% probability of happening).

 

[cronxeh]

I mean say there was a 99% chance of getting a green apple on the first pick from a hat, picked randomly. But you don't. Doesn't that negate everything?

No, it just means there are 99 apples out of 100 items in there, and you picked the one item that isn't an apple. Next time you do it you definitely going to get an apple

 

[frenzal_dude]

I mean say there was a 99% chance of getting a green apple on the first pick from a hat, picked randomly. But you don't. Doesn't that negate everything?

It just means that if you had an infinite amount of trials, 99% of those trials you would get a green apple, 1% u wouldn't.

If you flipped a coin 10x, maybe 4 of those 10 times you would get a head, if you tried again 10x, maybe 6 of those times you would get a head, but if you did it an infinite amount of times, you would get a head 50% of the time.

As the no. of trials approaches infinity, the % of times where you get an outcome will approach the % chance of that outcome.

You can do this type of experiment with an excel spread sheet.

 

[vela]

No, it just means there are 99 apples out of 100 items in there, and you picked the one item that isn't an apple. Next time you do it you definitely going to get an apple

How did you fit 99 apples in the hat?

 

[HallsofIvy]

It's a very large hat!

 

[HallsofIvy]

I mean say there was a 99% chance of getting a green apple on the first pick from a hat, picked randomly. But you don't. Doesn't that negate everything?

Probability deals with what happens in the long run. If you repeat this many times you will get very close to 99% of the apples you pick green. Look up the "law of large numbers".

 

[Mentallic]

Proability tells me that if I flip a coin 3 times, I will get 1.5 heads

 

[statdad]

Proability tells me that if I flip a coin 3 times, I will get 1.5 heads

I hope your tongue at the end means you posted this in jest, since probability says no such thing.

 

[xxChrisxx]

Proability tells me that if I flip a coin 3 times, I will get 1.5 heads

That's rubbish, as "tails never fails" so therefore has a 100% probability of success EDIT: As an interesting point, the probability of a human flipping a coin isn't acutally 50/50. There appears to be a slight favour of an even number of flips, it's something like 50.5/49.5 in favour of the side already up (unless you invert it after it's stopped spinning, then it's the side down). This only matters if you see what side of the the coin is up beforehand.

 

[Mentallic]

I hope your tongue at the end means you posted this in jest, since probability says no such thing.

I was hoping I could be given an insightful explanation as to how I end up with the half a head

As an interesting point, the probability of a human flipping a coin isn't acutally 50/50. There appears to be a slight favour of an even number of flips, it's something like 50.5/49.5 in favour of the side already up (unless you invert it after it's stopped spinning, then it's the side down). This only matters if you see what side of the the coin is up beforehand.

I'm very sceptical of this. This would have to have been done experimentally, and of course depending on the person that is flipping the coin and how many times he does so, it won't be exactly 50/50.

 

[xxChrisxx]

I'm very sceptical of this. This would have to have been done experimentally, and of course depending on the person that is flipping the coin and how many times he does so, it won't be exactly 50/50.

It's not just measured on outcome. The height, rate of spin etc were all recorded, when people got into a rhythm they could get it to flip to heads (we used heads up) knocking on for 80% of the time.

It's not groundbreaking, but it's interesting that it indicated as a 'natural flip' tended to be even for most people, for others it also tended to be odd

Although yeah as you say, it's almost impossible to get enough data for a reliable conclusion. But being bound by the limits of time and boredom, it's the best we could do.

 

[DaveC426913]

Probabilites are misused all the time in medical practice. "Only 3 in 100 of our Amnio patients wind up getting emergency Cesarians." The implication here is that you probably don't have to worry about it.

But the question needs to be asked: what if you are that 3%?

Probabilties have their place, but they do not and should not be applied to predict individual situations.

 

[LCKurtz]

I mean say there was a 99% chance of getting a green apple on the first pick from a hat, picked randomly. But you don't. Doesn't that negate everything?

As others have pointed out, no, it doesn't. However, if you repeat this experiment many times and keep not getting a green apple, you might become suspicious that somebody is pulling your leg about the probability being .99 in the first place. This leads to the idea of hypothesis testing. You could set up a hypothesis that p = .99 and test it. If you do enough trials and they all fail to be green, you might wind up rejecting the hypothesis with 95% confidence. And you would probably be correct, unless you are having a bad day.