Forums

General Betting

Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
Equimine.co.uk
23 Dec 10 09:46
Joined:
Date Joined: 11 Oct 07
| Topic/replies: 471 | Blogger: Equimine.co.uk's blog
I came across this question by chance on a maths forum. I would be interested to see what the answer is generally thought to be. I have my view which doesn't match that given.

The journey to or from work has a mean of 67 minutes and a standard deviation of 15 minutes. You can assume a normal distribution of journey times.

a) What's the probability the workers total traveling time to and from work is more than 151 minutes?
Pause Switch to Standard View Question from a Maths Forum, what do...
Show More
Loading...
Report scoop6winner. December 23, 2010 9:54 AM GMT
0.00
Report Bayes. December 23, 2010 10:14 AM GMT
0.211 assuming that the journey times are independent, which in a case like this is unlikely to be true.
Report Contrarian December 23, 2010 10:55 AM GMT
Bayes is right. Mean for total journey is 134 mins, sd is 21.21. So prob = (in Excel) (1 - NORMDIST(151, 134, 21.21, TRUE))
Report Bayes. December 23, 2010 11:49 AM GMT
The standard deviation comes from:

SD(X + X) = sqrt( Var(X) + Var(X))

The above only applies if the covariance of the two events (the trip to work and the trip home) is zero. If the times are independent then this is the case. Are the journey times of trips to and from work independent though? I doubt it!
Report Equimine.co.uk December 23, 2010 2:28 PM GMT
Thanks for your comments.

The experts answer is double the mean time and the SD to get the z score. So z=(151-134)/30 = 0.57, which gives .2843.

I do not think you can answer the question as worded, the journey times are not in, my view, independent.

Bayes, you have lost me (not difficult to do) with your SD calculation, and why, when you have a moment could you explain further.
Report Contrarian December 23, 2010 2:41 PM GMT
Equimine,

Your experts answer is wrong.

What Bayes is saying is basically that if you double the number of samples (in your example, 'samples' are time periods), then the SD increases by a factor of the square root of 2 (because SD is the root of variance).
Report Equimine.co.uk December 23, 2010 4:05 PM GMT
Contrarian & Bayes,

Thanks for that, I know the basic Bayes but wasn't aware of the point about the SD for multiple events.

As I said I don't think they independent events, but pleased to learn something new anyway.

Have a good Xmas.
Post Your Reply
<CTRL+Enter> to submit
Please login to post a reply.

Wonder

Instance ID: 13539
www.betfair.com