ABCDEFGHIJKLMNOPQRSTUVWX
1
2
Problem
3
Mark is a thin man from Germany with glasses who likes to listen to Mozart. Which is more likely? That Mark is A) a truck driver or (B) a professor of literature in Frankfurt?
4
From: The Art of Thinking Clearly, Rolf Dobelli
5
6
7
STEP 1: Create estimates
8
9
Professors (πŸ§‘β€πŸ«) who fit the description (πŸ‘“)Truck drivers (πŸš›) who fit the description (πŸ‘“)
10
11
25%
of professors who wear glasses and fit the description
0.2%
of truck drivers who wear glasses and fit the description
12
0.0002%
of people in Germany who are professors of literature in Frankfurt
0.1%
of people in Germany who are truck drivers
13
0.2%
of people in Germany who wear glasses and fit the description
0.2%
of people in Germany who wear glasses and fit the description
14
15
16
17
STEP 2: Add the estimates to the contingency table
18
19
Contingency table: p(πŸ§‘β€πŸ«) x p(πŸ‘“)Contingency table: p(πŸš›) x p(πŸ‘“)
20
21
πŸ‘“NOTMargin
(total counts)
πŸ‘“NOTMargin
(total counts)
22
πŸ‘“πŸ‘“
23
πŸ§‘β€πŸ«43170πŸš›17085,000
24
(25% of 170)(0.00002% of 85m)(0.2% of 85,000)(0.1% of 85m)
25
NOT πŸ§‘β€πŸ«NOT πŸš›
26
27
Margin
(total counts)
170,00085,000,000Margin
(total counts)
170,00085,000,000
28
(0.2% of 85m)(German population)(0.2% of 85m)(German population)
29
30
31
32
STEP 3: Calculate the blanks (all rows and columns should sum to their margins)
33
34
Contingency table: p(πŸ§‘β€πŸ«) x p(πŸ‘“)Contingency table: p(πŸš›) x p(πŸ‘“)
35
36
πŸ‘“NOTMargin
(total counts)
πŸ‘“NOTMargin
(total counts)
37
πŸ‘“πŸ‘“
38
πŸ§‘β€πŸ«43128170πŸš›17084,83085,000
39
(25% of 170)(0.00002% of 85m)(0.2% of 85,000)(0.1% of 85m)
40
NOT πŸ§‘β€πŸ«169,95884,829,87384,999,830NOT πŸš›169,83084,745,17084,915,000
41
42
Margin
(total counts)
170,00084,829,74585,000,000Margin
(total counts)
170,00084,660,34085,000,000
43
(0.2% of 85m)(German population)(0.2% of 85m)(German population)
44
45
46
47
STEP 4: Calculate the probabilities
48
49
ProbabilityCalculationProbabilityCalculation
50
p(πŸ‘“ | πŸ§‘β€πŸ«)0.25000043 / (43 + 128)p(πŸ‘“ | πŸš›)0.002000170 / (170 + 84,830)
51
p(πŸ‘“)0.002000170,000 / 85,000,000p(πŸ‘“)0.002000170,000 / 85,000,000
52
p(πŸ§‘β€πŸ« | πŸ‘“)0.00025043 / (43 + 169,958)p(πŸš› | πŸ‘“)0.001000170 / (170 + 169,830)
53
p(πŸ§‘β€πŸ«)0.000002170 / 85,000,000p(πŸš›)0.00100085,000 / 85,000,000
54
55
56
57
STEP 5: Finalise inferences
58
59
400%
greater likelihood of πŸš› fitting the description (πŸ‘“) vs πŸ§‘β€πŸ«fitting the description
60
12500%
greater likelihood of the description (πŸ‘“) fitting πŸ§‘β€πŸ« vs fitting πŸš›
61
62
Conclusion:
Despite the Frankfurt literature professors being x125 more likely than truck drives to fit the description, a person with the description is x4 more likely to be a trick driver than a professor.
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100