Monday, March 16, 2009

Backgammon and Safety

As I got older, I began to like backgammon more than chess because it more approximated life. In backgammon, you make decisions, and then roll the dice. The better the decision, the more likely the roll of the dice will bring a good result. It is possible to make good decisions every time but to lose because of the roll of the dice. I had a friend who was infuriating to play against. I would be a position wherein unless this friend rolled double 6's, I would win. More than once, the friend was able to come up with the double 6's.

If you read and study this link, you will know more about analyzing safety than most.

http://www.edcollins.com/backgammon/backlesson.htm

What I gain from this on the topic of safety is this.

1.Everyone makes predictions of the outcomes of behaviors when discussing safety. For instance, driving a car. One wears a seat belt or not, has bald tires or tires with tread, drives the speed limit or not (reducing speed in inclement weather or not), etc
a. So in the case here one has a course of action which results in 17 good possibilities versus 19 bad possibilities
b. In the other case, one has a course of action which results in 15 good possibilities versus 21 bad possibilities
c. In the other case, one has a course of action which results in 19 good possibilities versus 17 bad possibilities

2. How does this have a safety implication? Well it happens that safety is typically evaluated by results - in other words accidents. So if choices a,b result in good results on the next roll and choice c results in a bad result, the layman or uninformed would determine that choice a and b were the best course of action But they were not.

"Safety should be judged by the moves we make which give the best chance to avoid accidents and or to minimize injury and property damage."
Kirk Fechter

3. Check out this website for another application:

Try This, make a choice, track the results
http://www.grand-illusions.com/simulator/montysim.htm
NY Times version
http://www.nytimes.com/2008/04/08/science/08monty.html
Graphs
http://demonstrations.wolfram.com/MontyHallParadox/
This mathematician has further study
http://www.mindspring.com/~jimvb/monthall.htm

No comments: