Subject: | |
From: | |
Reply To: | |
Date: | Wed, 1 Mar 2000 17:34:43 -0500 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Here is just one more way of looking at the 3 doors problem:
Say every contestant chooses not to switch. Then they will win, on average,
1/3 of the time, which is their probability of winning when they first
picked their door, even if Monty reveals one of the empty doors.
The converse to not switching, provided everything else stays the same
(Monty always reveals an empty door), is switching. The converse
probability to 1/3 (not switching) is 1 - 1/3, or 2/3 (switching). It is
better to switch. Marilyn was correct.
I verified this with the following VB program:
Dim w As Integer, p As Integer, h As Integer, i As Integer, x As Integer
Randomize
x = 0
For i = 1 To 20000 '20000 iterations
w = Int(Rnd * 3 + 1) 'winning door (1-2-3)
p = Int(Rnd * 3 + 1) 'player's pick (1-2-3)
h = 1 'host's pick
If (h = w) Or (h = p) Then 'make sure host's pick
h = 2 ' is not the winning
If (h = w) Or (h = p) Then ' door, or the player's
h = 3 ' door.
End If
End If
'* p = 6 - h - p 'switch
p = p 'no switch
If p = w Then 'if player won,
x = x + 1 'then tally
End If
Next i
Debug.Print x
switching produced numbers in the low 13000 range
not switching produced numbers in the high 6000 range
Curt
|
|
|