1. Amateur has a question please

My wife and I have been married for 30 years. During all this time neither my wife, nor her father (my father-in-law), have had an auto accident. We live in central Texas and he lives in Kansas City. Today both my wife and my father-in-law each had an auto accident within ten minutes of one another each while driving in their own home town.

Can probability even be applied to an "event" like this? What are the odds? Or is it just a pure coincidence beyond calculation?

Or does it have something to do with Halloween?

Thank you.

2. Re: Amateur has a question please

You can calculate the probability on pretty much everything. However lots of factors come into play in a situation like this and the variables typically mediate each other. You would have to have lots and lots of random data about rates, for areas, days, times, ..., etc.

You may be able to look at accident rates for areas, the population numbers, etc., but then there are many other factors such as do the rates take into account people wiith multiple accidents, etc. You could figure it out, but it would end-up being a dissertation. You can probably just go with, the probability is very very low, but exists.

3. The Following User Says Thank You to hlsmith For This Useful Post:

Forensics1 (11-02-2013)

4. Re: Amateur has a question please

I agree with hlsmith. In theory you could calculate a probability for everything if you had data for it. But so many variables are in play you would have to limit them in practice. You could probably, for example, calculate the probability of two people who had not had accidents for a given number of years having an accident. Whether you could do this for a specific time window as narrow as ten minutes is doubtful because data is likely not collected for that.

Beyond that you have to decide exactly how many of the factors you calculated play into the model. The more you add, the more complex the calculations and the less likely the data exist. Actuarial tables, and Cox Regression, are variations of this.

5. The Following User Says Thank You to noetsi For This Useful Post:

Forensics1 (11-02-2013)

6. Re: Amateur has a question please

One aspect for the calculations is the timing of events.
that these 2 particular persons will have a car accident
on the very same day", and you assume e.g. that each of
them will have just 1 accident in 50 years, then the
chance would have been 1/ [50 (years) * 365.25 (days per year)]
* 1 / [50 * 365].

But you asked the question AFTER the event. So the question
really had to be, what are the odds than ANY two persons you know
will have car accidents on the very same day (you perhaps would
and your wife, but e.g. a close friend of yours and your brother,
or any other combination of significant others). And odds for
this are of course much less extreme.

Just my 2pence

K.

7. The Following User Says Thank You to Karabiner For This Useful Post:

Forensics1 (11-02-2013)

8. Re: Amateur has a question please

Thank you for the input. Yours is a very interesting science that I wish I knew more about. I do digital forensics for a living but I am not very good at math. It would be interesting to look at forensics, which is really looking at human behavior, from a probability standpoint. From doing criminal defense work I can say that most people who are accused of a crime are guilty of that crime. Now we just have to discover, and prove, which are not guilty. I will always wonder what is the probability of an innocent person being executed for that one crime s/he did not commit? I am willing to bet it has happened and will happen more often than any one knows.

Good day.

9. Re: Amateur has a question please

I listened to a state medical examiner lecture a while back and she said they never use liver temp for time of death approximation because it is so imprecise. This obviously is contrary to every CSI episode ever, what is your take.

 Tweet

Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts