pt
Joined: Nov 27, 2009
• Posts: 3
November 27th, 2009 at 2:01:23 PM permalink
hi
how do i relate say odds ie 3 to 1 to standard deviation then to probability
i know they are related but i HAVE NEVER seen any simple math relating them all to each other
On a bell curve you always see numbers expressed as standard deviations so can that be expressed as odds and probaility can anybody give me the simple math relating them all to each other

pt
dk
Joined: Nov 2, 2009
• Posts: 139
November 27th, 2009 at 7:23:13 PM permalink
Quote: pt

hi
how do i relate say odds ie 3 to 1 to standard deviation then to probability
i know they are related but i HAVE NEVER seen any simple math relating them all to each other
On a bell curve you always see numbers expressed as standard deviations so can that be expressed as odds and probaility can anybody give me the simple math relating them all to each other

pt

For an outcome with payout 1 for success and 0 for failure, standard deviation is the square root of the probability of success, p, times the probability of failure, q, Where q = 1 - p.

STDEV = SQ ROOT(p*q)
The ratio of people to cake is too big.
pt
Joined: Nov 27, 2009
• Posts: 3
November 28th, 2009 at 11:48:40 PM permalink
thanks

so any spot on the bell curve can be put in terms of odds?
boymimbo
Joined: Nov 12, 2009
• Posts: 5988
November 29th, 2009 at 12:29:10 AM permalink
There are certain rules as to when you can apply the "normal" approximation to those rules. The general rule is that np or nq must be greater than 10 to apply the approximation.

The standard deviation is expressed as np(1-p)^1/2, where n is the number of trials and p is the probability of the occurrence.

That is, say you throw the dice a thousand times. The odds of throwing a 7 are 1/6th.

The standard deviation of that dice roll ((1000)(1/6)(5/6))^1/2 or 11.785. That, is, you are likely to throw a seven 167 +/- 11.785 times. On a bell curve, one standard deviation encompasses about 68% of the entire curve while 2 standard deviations encompasses about 95% of the entire curve. For example, if you see a 7 thrown less than or equal to 144 times out of a thousand rolls, that result should only occur 2.5% of the time.

When probabilities are very small, a Poisson approximation is used instead of the normal approximation.
----- You want the truth! You can't handle the truth!