Serving the Quantitative Finance Community

farmer
Topic Author
Posts: 40
Joined: December 16th, 2002, 7:09 am

### Probability Distribution For Event (Bayesian?)

This phone is broken. Notice it says "No SIM". The news story came in between 5:30 and 6:29 before the phone was broken. Assume the latest the phone was broken is 6:36. What is the distribution of likely times the phone was broken?

Alan
Posts: 2505
Joined: December 19th, 2001, 4:01 am
Location: California
Contact:

### Re: Probability Distribution For Event (Bayesian?)

Whatever the distribution is, it's probably a defective distribution, meaning the area is less than one. The remaining mass is the probability the phone isn't actually broken.

farmer
Topic Author
Posts: 40
Joined: December 16th, 2002, 7:09 am

### Re: Probability Distribution For Event (Bayesian?)

Here is the phone from the side. It is assumed broken. It is at least unable to receive news alerts at the same time as saying "No SIM", and this condition is presumed stable. I think I could construct the probabilities using a P(A|B) Bayesian type thing. But I also suspect the probability has a linear increase from 5:30 to 6:29. In that case no probability math is necessary, as the height at any point in the right trapezoid can be determined from a total area of 1.

tagoma
Posts: 2186
Joined: February 21st, 2010, 12:58 pm

### Re: Probability Distribution For Event (Bayesian?)

One has to retrieve the first Fox News notification you received —as opposed to the latest— to get some sense of when your Iphone got broken. This is just my view.