Dear Allstatters,
Thank you very much for all the responses I have received so far. They have
been fascinating and include one that must qualify for the longest reply
ever on ALLSTAT. What I will do with these replies is create a PDF file
with all of them and put a link to that on my website. However, before I do
that, I want to clarify my query a little as your responses showed I hadn't
fully specified the rules of the game.
The reason why I put the query out is that I use the Monty Hall problem as
an ice-breaker in my training courses. It is a great way of getting people
to realise that statistical thinking is essential since normal human
intuition fails in this problem. However, I do like to then make the Monty
Hall problem real by translating it into a context that people are likely to
experience in real-life. My example wasn't the prisoners & king which I
used in my previous email but a competitive tender involving 3 companies A,
B, C who bidding for a contract with client X (could equally be 3 people
applying for a job).
The rules of this game are. You are working for company A and to begin with
all 3 companies tendering can be regarded as equally likely to win. You
decide you want to change the odds by bribing someone who works for X (call
him P) to find out what the person (call her Q) in X who will be making the
decision, is currently thinking about B and C. We will suppose that P and Q
are quite friendly with each other and are having a chat when P casually
says "I know a bit about company A since I've worked them in the past but I
don't know much about B & C. What are your impressions so far of B & C?".
Let's now imagine two responses that Q could give:
1. "Definitely not choosing B! Can't tell you yet which of A & C I
will be choosing but I have made my mind up."
2. "Definitely not choosing B! However, I haven't decided whether C
is better than A."
Both statements contain different information but it is now up to P to
convey the information back to you. In this scenario, it seems to me that
if you are the salesman for A, if you hear P relay statement 1 correctly
then you should conclude that the Monty Hall problem applies i.e. you are
playing a stick strategy and you now know that C has a 2/3 chance of
winning. Therefore you have wasted your money bribing P and you should now
resign from A and go and work for C. But if you hear statement 2 from P,
does the Monty Hall rationale still apply? Finally what if P provides you
with incomplete information and only says "It's not going to be B" and
leaves out the second sentence from Q? What conclusions can you as the
salesman for A make?
Regards
Nigel Marriott
Chartered Statistician
<http://www.marriott-stats.com/> www.marriott-stats.com
Ground Floor, 21 Marlborough Buildings, Bath BA1 2LY, United Kingdom
Tel (mobile) +44 (0)773 4069997
Tel (office) +44 (0)1225 489033
Fax +44 (0)870 6221969
Marriott Statistical Consulting Limited, Company No. 5577275, VAT No.
883304029
Registered in England, Registered Office - Equity House, 4-6 School Road,
Tilehurst, Reading, RG31 5AL
|