Peter Donnelly: How stats fool juries

244,315 views ใƒป 2007-01-12

TED


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืžืชืจื’ื: Yifat Adler ืžื‘ืงืจ: Ran Amitay
00:25
As other speakers have said, it's a rather daunting experience --
0
25000
2000
ื›ืคื™ ืฉืืžืจื• ืœืคื ื™, ื–ื” ื“ื™ ืžืœื—ื™ืฅ --
00:27
a particularly daunting experience -- to be speaking in front of this audience.
1
27000
3000
ื‘ืขืฆื, ืžืœื—ื™ืฅ ืžืื•ื“ -- ืœื“ื‘ืจ ืœืคื ื™ ื”ืงื”ืœ ื”ื–ื”.
00:30
But unlike the other speakers, I'm not going to tell you about
2
30000
3000
ืื‘ืœ ื‘ื ื™ื’ื•ื“ ืœื“ื•ื‘ืจื™ื ืื—ืจื™ื, ืื ื™ ืœื ืขื•ืžื“ ืœืกืคืจ ืœื›ื
00:33
the mysteries of the universe, or the wonders of evolution,
3
33000
2000
ืขืœ ืžืกืชืจื™ ื”ื™ืงื•ื, ืขืœ ืคืœืื™ ื”ืื‘ื•ืœื•ืฆื™ื”,
00:35
or the really clever, innovative ways people are attacking
4
35000
4000
ืื• ืขืœ ื“ืจื›ื™ื ื—ื›ืžื•ืช ื•ื—ื“ืฉื ื™ื•ืช ืœื”ืชืžื•ื“ื“ื•ืช
00:39
the major inequalities in our world.
5
39000
2000
ืขื ื—ื•ืกืจ ื”ืฉื™ื•ื•ื™ื•ืŸ ื”ืขื•ืœืžื™.
00:41
Or even the challenges of nation-states in the modern global economy.
6
41000
5000
ื•ืืคื™ืœื• ืœื ืขืœ ื”ืืชื’ืจื™ื ืฉืœ ืžื“ื™ื ื•ืช-ื”ืœืื•ื ื‘ืขื™ื“ืŸ ื”ื›ืœื›ืœื” ื”ื’ืœื•ื‘ืœื™ืช ื”ืžื•ื“ืจื ื™ืช.
00:46
My brief, as you've just heard, is to tell you about statistics --
7
46000
4000
ื›ืคื™ ืฉืฉืžืขืชื, ืื ื™ ืขื•ืžื“ ืœืกืคืจ ืœื›ื ืขืœ ืกื˜ื˜ื™ืกื˜ื™ืงื” --
00:50
and, to be more precise, to tell you some exciting things about statistics.
8
50000
3000
ื•ืื ื ื“ื™ื™ืง, ืœืกืคืจ ืœื›ื ื›ืžื” ื“ื‘ืจื™ื ืžืจืชืงื™ื ืขืœ ืกื˜ื˜ื™ืกื˜ื™ืงื”.
00:53
And that's --
9
53000
1000
ื•ื–ื” --
00:54
(Laughter)
10
54000
1000
[ืฆื—ื•ืง]
00:55
-- that's rather more challenging
11
55000
2000
-- ื–ื”ื• ืืชื’ืจ ื’ื“ื•ืœ ื™ื•ืชืจ ืžื–ื” ืฉืœ
00:57
than all the speakers before me and all the ones coming after me.
12
57000
2000
ื›ืœ ื”ื“ื•ื‘ืจื™ื ืฉื”ื™ื• ืœืคื ื™ ื•ื›ืœ ืืœื” ืฉื™ื’ื™ืขื• ื‘ื”ืžืฉืš.
00:59
(Laughter)
13
59000
1000
[ืฆื—ื•ืง]
01:01
One of my senior colleagues told me, when I was a youngster in this profession,
14
61000
5000
ืื—ื“ ืžืขืžื™ืชื™ ื”ื‘ื›ื™ืจื™ื ืืžืจ ืœื™ ื‘ื’ืื•ื•ื” ืจื‘ื”, ื‘ืชื—ื™ืœืช ื“ืจื›ื™ ื‘ืžืงืฆื•ืข ื”ื–ื”,
01:06
rather proudly, that statisticians were people who liked figures
15
66000
4000
ืฉืกื˜ื˜ื™ืกื˜ื™ืงืื™ื ื”ื ืื ืฉื™ื ืฉืื•ื”ื‘ื™ื ืžืกืคืจื™ื,
01:10
but didn't have the personality skills to become accountants.
16
70000
3000
ืื‘ืœ ืื™ืŸ ืœื”ื ื›ื™ืฉื•ืจื™ ื”ืื™ืฉื™ื•ืช ื”ื“ืจื•ืฉื™ื ืœื”ื™ื•ืช ืจื•ืื™ ื—ืฉื‘ื•ืŸ.
01:13
(Laughter)
17
73000
2000
[ืฆื—ื•ืง]
01:15
And there's another in-joke among statisticians, and that's,
18
75000
3000
ื•ื™ืฉ ืขื•ื“ ื‘ื“ื™ื—ื” ืคื ื™ืžื™ืช ืฉืœ ืกื˜ื˜ื™ืกื˜ื™ืงืื™ื,
01:18
"How do you tell the introverted statistician from the extroverted statistician?"
19
78000
3000
"ืื™ืš ืžื‘ื“ื™ืœื™ื ื‘ื™ืŸ ืกื˜ื˜ื™ืกื˜ื™ืงืื™ ืžื•ืคื ื ืœืกื˜ื˜ื™ืกื˜ื™ืงืื™ ืžื•ื—ืฆืŸ?"
01:21
To which the answer is,
20
81000
2000
ื•ื”ืชืฉื•ื‘ื” ื”ื™ื,
01:23
"The extroverted statistician's the one who looks at the other person's shoes."
21
83000
5000
"ื”ืกื˜ื˜ื™ืกื˜ื™ืงืื™ ื”ืžื•ื—ืฆืŸ ื”ื•ื ื–ื” ืฉื‘ื•ื—ืŸ ื ืขืœื™ื™ื ืฉืœ ืื ืฉื™ื ืื—ืจื™ื."
01:28
(Laughter)
22
88000
3000
[ืฆื—ื•ืง]
01:31
But I want to tell you something useful -- and here it is, so concentrate now.
23
91000
5000
ืื‘ืœ ืื ื™ ืจื•ืฆื” ืœื“ื‘ืจ ืขืœ ืžืฉื”ื• ืฉื™ืžื•ืฉื™ - - ืื– ื›ื“ืื™ ืœื”ืชืจื›ื– ืขื›ืฉื™ื•.
01:36
This evening, there's a reception in the University's Museum of Natural History.
24
96000
3000
ื”ืขืจื‘ ื ืขืจื›ืช ืงื‘ืœืช ืคื ื™ื ื‘ืžื•ื–ื™ืื•ืŸ ืœื”ื™ืกื˜ื•ืจื™ื” ืฉืœ ื”ื˜ื‘ืข ืฉืœ ื”ืื•ื ื™ื‘ืจืกื™ื˜ื”.
01:39
And it's a wonderful setting, as I hope you'll find,
25
99000
2000
ื–ื”ื• ืืชืจ ื ืคืœื, ื›ืคื™ ืฉืื ื™ ืžืงื•ื•ื” ืฉืชื’ืœื•,
01:41
and a great icon to the best of the Victorian tradition.
26
101000
5000
ื•ืกืžืœ ื—ืฉื•ื‘ ืœืžื™ื˜ื‘ ื”ืžืกื•ืจืช ื”ื•ื™ืงื˜ื•ืจื™ืื ื™ืช.
01:46
It's very unlikely -- in this special setting, and this collection of people --
27
106000
5000
ื–ื” ืžืื•ื“ ืœื ืกื‘ื™ืจ -- ื‘ืืชืจ ื”ืžื™ื•ื—ื“ ื”ื–ื” ื•ืขื ืื•ืกืฃ ื”ืื ืฉื™ื ื”ื–ื” --
01:51
but you might just find yourself talking to someone you'd rather wish that you weren't.
28
111000
3000
ืื‘ืœ ืืชื ืขืœื•ืœื™ื ืœืžืฆื•ื ืืช ืขืฆืžื›ื ืžื“ื‘ืจื™ื ืขื ืžื™ืฉื”ื• ืฉืืชื ืžืขื“ื™ืคื™ื ืœื”ืžื ืข ืžืฉื™ื—ื” ืื™ืชื•.
01:54
So here's what you do.
29
114000
2000
ืื– ื–ื” ืžื” ืฉืืชื ืฆืจื™ื›ื™ื ืœืขืฉื•ืช.
01:56
When they say to you, "What do you do?" -- you say, "I'm a statistician."
30
116000
4000
ื›ืฉืฉื•ืืœื™ื ืืชื›ื, "ื‘ืžื” ืืชื ืขื•ืกืงื™ื?" -- ืชื’ื™ื“ื• "ืื ื™ ืกื˜ื˜ื™ืกื˜ื™ืงืื™."
02:00
(Laughter)
31
120000
1000
[ืฆื—ื•ืง]
02:01
Well, except they've been pre-warned now, and they'll know you're making it up.
32
121000
4000
ืื‘ืœ ืขื›ืฉื™ื• ื”ื ื›ื‘ืจ ืงื™ื‘ืœื• ืื–ื”ืจื” ืžื•ืงื“ืžืช ื•ื™ื“ืขื• ืฉืืชื ืžื‘ืœืคื™ื.
02:05
And then one of two things will happen.
33
125000
2000
ื•ืื– ื™ืฉ ืฉืชื™ ืืคืฉืจื•ื™ื•ืช.
02:07
They'll either discover their long-lost cousin in the other corner of the room
34
127000
2000
ื”ื ื™ื’ืœื• ืืช ื”ื“ื•ื“ืŸ ื”ืื‘ื•ื“ ืฉืœื”ื ื‘ืคื™ื ื” ื”ืจื—ื•ืงื”
02:09
and run over and talk to them.
35
129000
2000
ืฉืœ ื”ื—ื“ืจ ื•ื™ืจื•ืฆื• ืœื“ื‘ืจ ืื™ืชื•.
02:11
Or they'll suddenly become parched and/or hungry -- and often both --
36
131000
3000
ืื• ืฉื”ื ื™ื–ื›ืจื• ื‘ืื•ืคืŸ ืคืชืื•ืžื™ ืฉื”ื ืžืชื™ื ืžืฆืžื ื•/ืื• ืžืจืขื‘
02:14
and sprint off for a drink and some food.
37
134000
2000
ื•ื™ืคืชื—ื• ื‘ืจื™ืฆื” ืงืœื” ืœืขื‘ืจ ื”ืžืฉืงืื•ืช ื•ื”ืžื–ื•ืŸ.
02:16
And you'll be left in peace to talk to the person you really want to talk to.
38
136000
4000
ื•ืื– ืชื”ื™ื• ื—ื•ืคืฉื™ื™ื ืœื“ื‘ืจ ืขื ืžื™ ืฉื‘ืืžืช ืชืจืฆื• ืœื“ื‘ืจ ืื™ืชื•.
02:20
It's one of the challenges in our profession to try and explain what we do.
39
140000
3000
ืื—ื“ ืžืืชื’ืจื™ ื”ืžืงืฆื•ืข ื”ื•ื ืœื ืกื•ืช ืœื”ืกื‘ื™ืจ ืžื” ืื ื—ื ื• ืขื•ืฉื™ื.
02:23
We're not top on people's lists for dinner party guests and conversations and so on.
40
143000
5000
ืื ื—ื ื• ืœื ื‘ืจืืฉ ืจืฉื™ืžื•ืช ื”ืื•ืจื—ื™ื ื•ื”ื ื•ืฉืื™ื ืœืฉื™ื—ื” ื‘ืกืขื•ื“ื•ืช ื—ื’ื™ื’ื™ื•ืช.
02:28
And it's something I've never really found a good way of doing.
41
148000
2000
ื•ื–ื” ืžืฉื”ื• ืฉืžืขื•ืœื ืœื ืžืฆืืชื™ ื“ืจืš ื˜ื•ื‘ื” ืœื‘ืฆืข.
02:30
But my wife -- who was then my girlfriend --
42
150000
3000
ืื‘ืœ ืืฉืชื™ -- ืฉื”ื™ื™ืชื” ืื– ื”ื—ื‘ืจื” ืฉืœื™ --
02:33
managed it much better than I've ever been able to.
43
153000
3000
ื˜ื™ืคืœื” ื‘ื–ื” ื”ืจื‘ื” ื™ื•ืชืจ ื˜ื•ื‘ ืžืžื ื™.
02:36
Many years ago, when we first started going out, she was working for the BBC in Britain,
44
156000
3000
ืœืคื ื™ ื”ืจื‘ื” ืฉื ื™ื, ื›ืฉื”ืชื—ืœื ื• ืœืฆืืช, ื”ื™ื ืขื‘ื“ื” ืขื‘ื•ืจ ื”ื‘ื™.ื‘ื™.ืกื™. ื‘ื‘ืจื™ื˜ื ื™ื”,
02:39
and I was, at that stage, working in America.
45
159000
2000
ื•ื‘ืื•ืชื• ืฉืœื‘, ืื ื™ ืขื‘ื“ืชื™ ื‘ืืžืจื™ืงื”.
02:41
I was coming back to visit her.
46
161000
2000
ื—ื–ืจืชื™ ื›ื“ื™ ืœื‘ืงืจ ืื•ืชื”.
02:43
She told this to one of her colleagues, who said, "Well, what does your boyfriend do?"
47
163000
6000
ื”ื™ื ืกื™ืคืจื” ืขืœ ื›ืš ืœืื—ืช ืžื—ื‘ืจื•ืชื™ื” ืœืขื‘ื•ื“ื” ืฉืฉืืœื”, "ื‘ืžื” ืขื•ืกืง ื”ื—ื‘ืจ ืฉืœืš?"
02:49
Sarah thought quite hard about the things I'd explained --
48
169000
2000
ืฉืจื” ื—ืฉื‘ื” ืขืžื•ืงื•ืช ืขืœ ื”ื“ื‘ืจื™ื ืฉื”ืกื‘ืจืชื™ --
02:51
and she concentrated, in those days, on listening.
49
171000
4000
ื‘ืื•ืชื ื”ื™ืžื™ื ื”ื™ื ื”ืชืจื›ื–ื” ื‘ื”ืงืฉื‘ื”.
02:55
(Laughter)
50
175000
2000
[ืฆื—ื•ืง]
02:58
Don't tell her I said that.
51
178000
2000
ืืœ ืชื’ืœื• ืœื” ืฉืืžืจืชื™ ืืช ื–ื”.
03:00
And she was thinking about the work I did developing mathematical models
52
180000
4000
ื”ื™ื ื—ืฉื‘ื” ืขืœ ื”ืขื‘ื•ื“ื” ืฉืœื™ ื‘ืคื™ืชื•ื— ืžื•ื“ืœื™ื ืžืชืžื˜ื™ื™ื
03:04
for understanding evolution and modern genetics.
53
184000
3000
ืœื”ื‘ื ื” ืฉืœ ืื‘ื•ืœื•ืฆื™ื” ื•ื’ื ื˜ื™ืงื” ืžื•ื“ืจื ื™ืช.
03:07
So when her colleague said, "What does he do?"
54
187000
3000
ื•ื›ืืฉืจ ื”ืขืžื™ืชื” ืฉืœื” ืฉืืœื”, "ื‘ืžื” ื”ื•ื ืขื•ืกืง?"
03:10
She paused and said, "He models things."
55
190000
4000
ื”ื™ื ืขืฆืจื” ืจื’ืข ื•ืืžืจื”, "ื”ื•ื ืžื“ื’ืžืŸ (=ื‘ื•ื ื” ื“ื’ืžื™ื ืฉืœ) ื“ื‘ืจื™ื."
03:14
(Laughter)
56
194000
1000
[ืฆื—ื•ืง]
03:15
Well, her colleague suddenly got much more interested than I had any right to expect
57
195000
4000
ื”ืขืžื™ืชื” ืฉืœื” ื’ื™ืœืชื” ืคืชืื•ื ืขื ื™ื™ืŸ ืจื‘ ืฉืœื ื”ื™ื™ืชื™ ืจืื•ื™ ืœื•
03:19
and went on and said, "What does he model?"
58
199000
3000
ื•ืฉืืœื”, "ืžื” ื”ื•ื ืžื“ื’ืžืŸ?"
03:22
Well, Sarah thought a little bit more about my work and said, "Genes."
59
202000
3000
ืฉืจื” ื—ืฉื‘ื” ืขื•ื“ ืงืฆืช ืขืœ ื”ืขื‘ื•ื“ื” ืฉืœื™ ื•ืขื ืชื” "ื’'ื™ื ืก (=ื’ื ื™ื)."
03:25
(Laughter)
60
205000
4000
[ืฆื—ื•ืง]
03:29
"He models genes."
61
209000
2000
"ื”ื•ื ืžื“ื’ืžืŸ ื’'ื™ื ืก."
03:31
That is my first love, and that's what I'll tell you a little bit about.
62
211000
4000
ื•ื–ื•ื”ื™ ืื”ื‘ืชื™ ื”ืจืืฉื•ื ื”, ืขืœื™ื” ืื“ื‘ืจ ืงืฆืช ืขื›ืฉื™ื•.
03:35
What I want to do more generally is to get you thinking about
63
215000
4000
ื•ื‘ื™ืชืจ ื›ืœืœื™ื•ืช, ืื ื™ ืจื•ืฆื” ืฉืชื—ืฉื‘ื•
03:39
the place of uncertainty and randomness and chance in our world,
64
219000
3000
ืขืœ ื”ืžืงื•ื ืฉืœ ื—ื•ืกืจ ื”ื•ื•ื“ืื•ืช, ื”ืืงืจืื™ื•ืช ื•ื”ืžืงืจื” ื‘ืขื•ืœืžื ื•,
03:42
and how we react to that, and how well we do or don't think about it.
65
222000
5000
ืื™ืš ืื ื—ื ื• ืžื’ื™ื‘ื™ื ืœื”ื, ื•ืขื“ ื›ืžื” ืื ื• ื—ื•ืฉื‘ื™ื ืขืœื™ื”ื ื‘ืฆื•ืจื” ื ื›ื•ื ื”.
03:47
So you've had a pretty easy time up till now --
66
227000
2000
ื‘ื”ืจืฆืื•ืช ืขื“ ืขื›ืฉื™ื• ื”ื™ื” ืœื›ื ื“ื™ ืงืœ --
03:49
a few laughs, and all that kind of thing -- in the talks to date.
67
229000
2000
ืงืฆืช ืฆื—ื•ืงื™ื, ื•ื“ื‘ืจื™ื ื›ืืœื”.
03:51
You've got to think, and I'm going to ask you some questions.
68
231000
3000
ืขื›ืฉื™ื• ืชืฆื˜ืจื›ื• ืœื”ืคืขื™ืœ ืืช ื”ืžื•ื—, ื•ืื ื™ ืขื•ืžื“ ืœืฉืื•ืœ ืืชื›ื ื›ืžื” ืฉืืœื•ืช.
03:54
So here's the scene for the first question I'm going to ask you.
69
234000
2000
ื–ืืช ื”ืชืคืื•ืจื” ืœืฉืืœื” ื”ืจืืฉื•ื ื” ืฉืืฆื™ื’ ื‘ืคื ื™ื›ื.
03:56
Can you imagine tossing a coin successively?
70
236000
3000
ืชื•ื›ืœื• ืœื“ืžื™ื™ืŸ ืืช ืขืฆืžื›ื ืžื˜ื™ืœื™ื ืžื˜ื‘ืข ื‘ืจืฆืฃ?
03:59
And for some reason -- which shall remain rather vague --
71
239000
3000
ืžืกื™ื‘ื” ื›ืœืฉื”ื™ -- ืฉืชืฉืืจ ื“ื™ ืžืขื•ืจืคืœืช --
04:02
we're interested in a particular pattern.
72
242000
2000
ืื ื—ื ื• ืžืชืขื ื™ื™ื ื™ื ื‘ื“ืคื•ืก ืžืกื•ื™ื™ื.
04:04
Here's one -- a head, followed by a tail, followed by a tail.
73
244000
3000
ื”ื ื” ื“ืคื•ืก - ืขืฅ, ืื—ืจื™ื• ืคืœื™, ื•ืื—ืจื™ื• ืคืœื™.
04:07
So suppose we toss a coin repeatedly.
74
247000
3000
ื ื ื™ื— ืฉืื ื—ื ื• ื—ื•ื–ืจื™ื ืขืœ ื”ื˜ืœืช ืžื˜ื‘ืข.
04:10
Then the pattern, head-tail-tail, that we've suddenly become fixated with happens here.
75
250000
5000
ื•ืื– ื”ื“ืคื•ืก ืฉืœ ืขืฅ-ืคืœื™-ืคืœื™, ืฉืคืชืื•ื ื”ืชื—ืœื ื• ืœื’ืœื•ืช ื‘ื• ืขื ื™ืŸ ืจื‘, ืžื•ืคื™ืข ื›ืืŸ.
04:15
And you can count: one, two, three, four, five, six, seven, eight, nine, 10 --
76
255000
4000
ื•ืืคืฉืจ ืœืกืคื•ืจ: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 --
04:19
it happens after the 10th toss.
77
259000
2000
ื”ื•ื ืžื•ืคื™ืข ืื—ืจื™ ื”ื”ื˜ืœื” ื”-10.
04:21
So you might think there are more interesting things to do, but humor me for the moment.
78
261000
3000
ืื•ืœื™ ืชื—ืฉื‘ื• ืฉื™ืฉ ื“ื‘ืจื™ื ื™ื•ืชืจ ืžืขื ื™ื™ื ื™ื ืœืขืฉื•ืช, ืื‘ืœ ืชืฉืืจื• ืื™ืชื™ ืจื’ืข.
04:24
Imagine this half of the audience each get out coins, and they toss them
79
264000
4000
ื“ืžื™ื™ื ื• ืฉื”ื—ืฆื™ ื”ื–ื” ืฉืœ ื”ืงื”ืœ ืžืงื‘ืœ ืžื˜ื‘ืขื•ืช ื•ืžื˜ื™ืœ ืื•ืชื
04:28
until they first see the pattern head-tail-tail.
80
268000
3000
ืขื“ ื”ืคืขื ื”ืจืืฉื•ื ื” ื‘ื” ืžื•ืคื™ืข ืขืฅ-ืคืœื™-ืคืœื™.
04:31
The first time they do it, maybe it happens after the 10th toss, as here.
81
271000
2000
ื‘ืคืขื ื”ืจืืฉื•ื ื” ืื•ืœื™ ื–ื” ื™ื•ืคื™ืข ืื—ืจื™ ื”ื”ื˜ืœื” ื”-10, ื›ืžื• ื›ืืŸ.
04:33
The second time, maybe it's after the fourth toss.
82
273000
2000
ื‘ืคืขื ื”ืฉื ื™ื™ื” ืื•ืœื™ ืื—ืจื™ ื”ื”ื˜ืœื” ื”-4.
04:35
The next time, after the 15th toss.
83
275000
2000
ื‘ืคืขื ื”ื‘ืื”, ืื—ืจื™ ื”ื”ื˜ืœื” ื”-15.
04:37
So you do that lots and lots of times, and you average those numbers.
84
277000
3000
ืชื—ื–ืจื• ืขืœ ื›ืš ื”ืจื‘ื” ืคืขืžื™ื, ื•ืชื—ืฉื‘ื• ืžืžื•ืฆืข ืฉืœ ื”ืžืกืคืจื™ื ื”ืืœื”.
04:40
That's what I want this side to think about.
85
280000
3000
ื–ื” ืžื” ืฉืื ื™ ืจื•ืฆื” ืฉื”ืฆื“ ื”ื–ื” ื™ื—ืฉื•ื‘ ืขืœื™ื•.
04:43
The other half of the audience doesn't like head-tail-tail --
86
283000
2000
ื”ื—ืฆื™ ื”ืฉื ื™ ืฉืœ ื”ืงื”ืœ ืœื ืžื—ื‘ื‘ ืืช ืขืฅ-ืคืœื™-ืคืœื™ --
04:45
they think, for deep cultural reasons, that's boring --
87
285000
3000
ื”ื ื—ื•ืฉื‘ื™ื, ืžืกื™ื‘ื•ืช ืชืจื‘ื•ืชื™ื•ืช ืขืžื•ืงื•ืช, ืฉื–ื”ื• ื“ืคื•ืก ืžืฉืขืžื --
04:48
and they're much more interested in a different pattern -- head-tail-head.
88
288000
3000
ื•ื”ื ืžื’ืœื™ื ืขื ื™ื™ืŸ ืจื‘ ื‘ื“ืคื•ืก ืื—ืจ -- ืขืฅ-ืคืœื™-ืขืฅ.
04:51
So, on this side, you get out your coins, and you toss and toss and toss.
89
291000
3000
ืื–, ื‘ืฆื“ ื”ื–ื”, ืชืงื‘ืœื• ืืช ื”ืžื˜ื‘ืขื•ืช ืฉืœื›ื, ื•ืชื˜ื™ืœื• ืื•ืชื ืฉื•ื‘ ื•ืฉื•ื‘ ื•ืฉื•ื‘.
04:54
And you count the number of times until the pattern head-tail-head appears
90
294000
3000
ื•ืืชื ืชืกืคืจื• ืืช ืžืกืคืจ ื”ื”ื˜ืœื•ืช ืขื“ ืฉื”ื“ืคื•ืก ืขืฅ-ืคืœื™-ืขืฅ ืžื•ืคื™ืข
04:57
and you average them. OK?
91
297000
3000
ื•ืชื—ืฉื‘ื• ืžืžื•ืฆืข ืฉืœื”ืŸ. ืื•ืงื™?
05:00
So on this side, you've got a number --
92
300000
2000
ืื– ื‘ืฆื“ ื”ื–ื”, ื™ืฉ ืœื›ื ืžืกืคืจ --
05:02
you've done it lots of times, so you get it accurately --
93
302000
2000
ืขืฉื™ืชื ื–ืืช ื”ืจื‘ื” ืคืขืžื™ื, ืื– ืงื™ื‘ืœืชื ืชื•ืฆืื” ืžื“ื•ื™ื™ืงืช --
05:04
which is the average number of tosses until head-tail-tail.
94
304000
3000
ืฉื”ื™ื ืžืกืคืจ ื”ื”ื˜ืœื•ืช ื”ืžืžื•ืฆืข ืขื“ ืœืขืฅ-ืคืœื™-ืคืœื™.
05:07
On this side, you've got a number -- the average number of tosses until head-tail-head.
95
307000
4000
ื•ื‘ืฆื“ ื”ื–ื”, ื™ืฉ ืœื›ื ืžืกืคืจ -- ืžืกืคืจ ื”ื”ื˜ืœื•ืช ื”ืžืžื•ืฆืข ืขื“ ืœืขืฅ-ืคืœื™-ืขืฅ.
05:11
So here's a deep mathematical fact --
96
311000
2000
ืขื•ื‘ื“ื” ืžืชืžื˜ื™ืช ืขืžื•ืงื” --
05:13
if you've got two numbers, one of three things must be true.
97
313000
3000
ืื ื™ืฉ ืœื›ื 2 ืžืกืคืจื™ื, ืื—ื“ ืž-3 ื”ื“ื‘ืจื™ื ื”ื‘ืื™ื ื—ื™ื™ื‘ ืœื”ืชืงื™ื™ื.
05:16
Either they're the same, or this one's bigger than this one,
98
316000
3000
ืื• ืฉื”ื ื–ื”ื™ื, ืื• ืฉื–ื” ื™ื•ืชืจ ื’ื“ื•ืœ ืžื–ื”,
05:19
or this one's bigger than that one.
99
319000
1000
ืื• ืฉื–ื” ื™ื•ืชืจ ื’ื“ื•ืœ ืžื–ื”.
05:20
So what's going on here?
100
320000
3000
ืžื” ืงื•ืจื” ื›ืืŸ?
05:23
So you've all got to think about this, and you've all got to vote --
101
323000
2000
ื›ื•ืœื›ื ืฆืจื™ื›ื™ื ืœื—ืฉื•ื‘, ื•ื›ื•ืœื›ื ืฆืจื™ื›ื™ื ืœื”ืฆื‘ื™ืข --
05:25
and we're not moving on.
102
325000
1000
ื•ืื ื—ื ื• ืœื ืžืชืงื“ืžื™ื.
05:26
And I don't want to end up in the two-minute silence
103
326000
2000
ื•ืœื ื”ื™ื™ืชื™ ืจื•ืฆื” ืฉืชื™ื•ื•ืฆืจ ืฉืชื™ืงื” ืฉืœ 2 ื“ืงื•ืช
05:28
to give you more time to think about it, until everyone's expressed a view. OK.
104
328000
4000
ื›ื“ื™ ืฉื™ื”ื™ื” ืœื›ื ื™ื•ืชืจ ื–ืžืŸ ืœื—ืฉื•ื‘ ืขืœ ื›ืš, ืขื“ ืฉื›ื•ืœื ื™ื—ื•ื• ืืช ื“ืขืชื.
05:32
So what you want to do is compare the average number of tosses until we first see
105
332000
4000
ื ืจืฆื” ืœื”ืฉื•ื•ืช ืืช ืžืกืคืจ ื”ื”ื˜ืœื•ืช ื”ืžืžื•ืฆืข ืขื“ ืฉื ืจืื” ืขืฅ-ืคืœื™-ืขืฅ ื‘ืคืขื ื”ืจืืฉื•ื ื”
05:36
head-tail-head with the average number of tosses until we first see head-tail-tail.
106
336000
4000
ืœืžืกืคืจ ื”ื”ื˜ืœื•ืช ื”ืžืžื•ืฆืข ืขื“ ืฉื ืจืื” ืขืฅ-ืคืœื™-ืคืœื™ ื‘ืคืขื ื”ืจืืฉื•ื ื”.
05:41
Who thinks that A is true --
107
341000
2000
ืžื™ ื—ื•ืฉื‘ ืฉื' ื ื›ื•ืŸ --
05:43
that, on average, it'll take longer to see head-tail-head than head-tail-tail?
108
343000
4000
ื›ืœื•ืžืจ, ื‘ืžืžื•ืฆืข, ื™ื™ืงื— ื™ื•ืชืจ ื–ืžืŸ ืœืจืื•ืช ืขืฅ-ืคืœื™-ืขืฅ ืžืืฉืจ ืขืฅ-ืคืœื™-ืคืœื™?
05:47
Who thinks that B is true -- that on average, they're the same?
109
347000
3000
ืžื™ ื—ื•ืฉื‘ ืฉื‘' ื ื›ื•ืŸ -- ืฉื‘ืžืžื•ืฆืข ื”ื ื–ื”ื™ื?
05:51
Who thinks that C is true -- that, on average, it'll take less time
110
351000
2000
ืžื™ ื—ื•ืฉื‘ ืฉื’' ื ื›ื•ืŸ -- ืฉื‘ืžืžื•ืฆืข ื–ื” ื™ืงื— ืคื—ื•ืช ืคืขืžื™ื
05:53
to see head-tail-head than head-tail-tail?
111
353000
3000
ืœืจืื•ืช ืขืฅ-ืคืœื™-ืขืฅ ืžืืฉืจ ืขืฅ-ืคืœื™-ืคืœื™?
05:57
OK, who hasn't voted yet? Because that's really naughty -- I said you had to.
112
357000
3000
ืžื™ ืขื“ื™ื™ืŸ ืœื ื”ืฆื‘ื™ืข? ืฉื•ื‘ื‘ื™ื -- ืืžืจืชื™ ืฉื—ื™ื™ื‘ื™ื ืœื”ืฆื‘ื™ืข.
06:00
(Laughter)
113
360000
1000
[ืฆื—ื•ืง]
06:02
OK. So most people think B is true.
114
362000
3000
ืจื•ื‘ ื”ืื ืฉื™ื ื—ื•ืฉื‘ื™ื ืฉื‘' ื ื›ื•ืŸ.
06:05
And you might be relieved to know even rather distinguished mathematicians think that.
115
365000
3000
ื•ืื•ืœื™ ืชืฉืžื—ื• ืœืฉืžื•ืข ืฉืžืชืžื˜ื™ืงืื™ื ื“ื™ ืžื›ื•ื‘ื“ื™ื ื—ื•ืฉื‘ื™ื ื›ืžื•ื›ื.
06:08
It's not. A is true here.
116
368000
4000
ืื‘ืœ ืœื. ื' ื”ื™ื ื”ืชืฉื•ื‘ื” ื”ื ื›ื•ื ื”.
06:12
It takes longer, on average.
117
372000
2000
ื–ื” ืœื•ืงื— ื™ื•ืชืจ ื–ืžืŸ, ื‘ืžืžื•ืฆืข.
06:14
In fact, the average number of tosses till head-tail-head is 10
118
374000
2000
ืœืžืขืฉื”, ืžืกืคืจ ื”ื”ื˜ืœื•ืช ื”ืžืžื•ืฆืข ืขื“ ืขืฅ-ืคืœื™-ืขืฅ ื”ื•ื 10
06:16
and the average number of tosses until head-tail-tail is eight.
119
376000
5000
ื•ืžืกืคืจ ื”ื”ื˜ืœื•ืช ื”ืžืžื•ืฆืข ืขื“ ืขืฅ-ืคืœื™-ืคืœื™ ื”ื•ื 8.
06:21
How could that be?
120
381000
2000
ืื™ืš ื–ื” ื™ื™ืชื›ืŸ?
06:24
Anything different about the two patterns?
121
384000
3000
ื™ืฉ ื”ื‘ื“ืœ ื‘ื™ืŸ ืฉื ื™ ื”ื“ืคื•ืกื™ื?
06:30
There is. Head-tail-head overlaps itself.
122
390000
5000
ื•ื”ืชืฉื•ื‘ื” ื”ื™ื ื›ืŸ. ืขืฅ-ืคืœื™-ืขืฅ ื—ื•ืคืฃ ืืช ืขืฆืžื•.
06:35
If you went head-tail-head-tail-head, you can cunningly get two occurrences
123
395000
4000
ืื ื™ื•ืฆื ืœื›ื ืขืฅ-ืคืœื™-ืขืฅ-ืคืœื™-ืขืฅ, ืชื•ื›ืœื• ืœืงื‘ืœ ื‘ืขืจืžื•ืžื™ื•ืช 2 ืžื•ืคืขื™ื
06:39
of the pattern in only five tosses.
124
399000
3000
ืฉืœ ื”ื“ืคื•ืก ื‘-5 ื”ื˜ืœื•ืช ื‘ืœื‘ื“.
06:42
You can't do that with head-tail-tail.
125
402000
2000
ืœื ื ื™ืชืŸ ืœืขืฉื•ืช ื–ืืช ืขื ืขืฅ-ืคืœื™-ืคืœื™.
06:44
That turns out to be important.
126
404000
2000
ื•ืžืกืชื‘ืจ ืฉื–ื” ื—ืฉื•ื‘.
06:46
There are two ways of thinking about this.
127
406000
2000
ื™ืฉ 2 ื“ืจื›ื™ื ืœื—ืฉื•ื‘ ืขืœ ื›ืš.
06:48
I'll give you one of them.
128
408000
2000
ืืฆื™ื’ ื‘ืคื ื™ื›ื ืื—ืช ืžื”ืŸ.
06:50
So imagine -- let's suppose we're doing it.
129
410000
2000
ื“ืžื™ื™ื ื• ืœืขืฆืžื›ื -- ื ื ื™ื— ืฉืื ื—ื ื• ืžื‘ืฆืขื™ื ื–ืืช.
06:52
On this side -- remember, you're excited about head-tail-tail;
130
412000
2000
ื‘ืฆื“ ื”ื–ื” -- ืืชื ื–ื•ื›ืจื™ื, ืืชื ืžืชืœื”ื‘ื™ื ืžืขืฅ-ืคืœื™-ืคืœื™,
06:54
you're excited about head-tail-head.
131
414000
2000
ื•ืืชื ืžืชืœื”ื‘ื™ื ืžืขืฅ-ืคืœื™-ืขืฅ.
06:56
We start tossing a coin, and we get a head --
132
416000
3000
ืื ื—ื ื• ืžืชื—ื™ืœื™ื ืœื”ื˜ื™ืœ ืžื˜ื‘ืข, ื•ืงื™ื‘ืœื ื• ืขืฅ --
06:59
and you start sitting on the edge of your seat
133
419000
1000
ืืชื ื™ื•ืฉื‘ื™ื ืขืœ ืงืฆื” ื”ื›ืกื
07:00
because something great and wonderful, or awesome, might be about to happen.
134
420000
5000
ื›ื™ ื™ื™ืชื›ืŸ ืฉืžืฉื”ื• ืขื ืง ื•ื ืคืœื, ืื• ืขืฆื•ื ื•ื›ื‘ื™ืจ, ืขื•ืžื“ ืœื”ืชืจื—ืฉ.
07:05
The next toss is a tail -- you get really excited.
135
425000
2000
ื”ื”ื˜ืœื” ื”ื‘ืื” ื”ื™ื ืคืœื™ -- ืืชื ืžืžืฉ ืžืชืœื”ื‘ื™ื.
07:07
The champagne's on ice just next to you; you've got the glasses chilled to celebrate.
136
427000
4000
ื”ืฉืžืคื ื™ื” ื‘ืงืจื— ืžื•ื›ื ื”, ื”ื’ื‘ื™ืขื™ื ื›ื‘ืจ ื‘ืงื™ืจื•ืจ ืœืงืจืืช ื”ื—ื’ื™ื’ื”.
07:11
You're waiting with bated breath for the final toss.
137
431000
2000
ืืชื ืžื—ื›ื™ื ื‘ื ืฉื™ืžื” ืขืฆื•ืจื” ืœื”ื˜ืœื” ื”ืกื•ืคื™ืช.
07:13
And if it comes down a head, that's great.
138
433000
2000
ืื ื™ื•ืฆื ืขืฅ, ื–ื” ื ืคืœื.
07:15
You're done, and you celebrate.
139
435000
2000
ืกื™ื™ืžืชื ื•ืืชื ื—ื•ื’ื’ื™ื.
07:17
If it's a tail -- well, rather disappointedly, you put the glasses away
140
437000
2000
ืื ื™ื•ืฆื ืคืœื™ -- ืื–, ื‘ืื›ื–ื‘ื” ืจื‘ื”, ืืชื ืžื ื™ื—ื™ื
07:19
and put the champagne back.
141
439000
2000
ืืช ื”ื’ื‘ื™ืขื™ื ื•ืืช ื”ืฉืžืคื ื™ื” ื‘ืฆื“.
07:21
And you keep tossing, to wait for the next head, to get excited.
142
441000
3000
ื•ืืชื ืžืžืฉื™ื›ื™ื ืœื”ื˜ื™ืœ, ืœื—ื›ื•ืช ืœืขืฅ ื”ื‘ื, ืœื”ืชืœื”ื‘.
07:25
On this side, there's a different experience.
143
445000
2000
ื‘ืฆื“ ื”ื–ื”, ื”ื”ืชื ืกื•ืช ื”ื™ื ืฉื•ื ื”.
07:27
It's the same for the first two parts of the sequence.
144
447000
3000
ื–ื” ืื•ืชื• ื”ื“ื‘ืจ ื‘ืฉื ื™ ื”ื—ืœืงื™ื ื”ืจืืฉื•ื ื™ื ืฉืœ ื”ืกื“ืจื”.
07:30
You're a little bit excited with the first head --
145
450000
2000
ืืชื ืงืฆืช ืžืชืจื’ืฉื™ื ื›ืฉืžื•ืคื™ืข ื”ืขืฅ ื”ืจืืฉื•ืŸ --
07:32
you get rather more excited with the next tail.
146
452000
2000
ืžืชืœื”ื‘ื™ื ืงืฆืช ื™ื•ืชืจ ืขื ื”ืคืœื™ ื”ื‘ื.
07:34
Then you toss the coin.
147
454000
2000
ื•ืื– ืืชื ืžื˜ื™ืœื™ื ืืช ื”ืžื˜ื‘ืข.
07:36
If it's a tail, you crack open the champagne.
148
456000
3000
ืื ื™ืฆื ืคืœื™, ืคื•ืชื—ื™ื ืืช ื”ืฉืžืคื ื™ื”.
07:39
If it's a head you're disappointed,
149
459000
2000
ืื ื™ืฆื ืขืฅ, ืืชื ืžืื•ื›ื–ื‘ื™ื,
07:41
but you're still a third of the way to your pattern again.
150
461000
3000
ืื‘ืœ ืืชื ื›ื‘ืจ ื‘ืฉืœื™ืฉ ื”ื“ืจืš ืœืงืจืืช ื”ื“ืคื•ืก ื”ื‘ื ืฉืœื›ื.
07:44
And that's an informal way of presenting it -- that's why there's a difference.
151
464000
4000
ื–ื•ื”ื™ ื“ืจืš ืœื ืคื•ืจืžืœื™ืช ืœื”ืฆื™ื’ ื–ืืช -- ืื‘ืœ ื–ืืช ื”ืกื™ื‘ื” ืœื”ื‘ื“ืœ.
07:48
Another way of thinking about it --
152
468000
2000
ื“ืจืš ืื—ืจืช ืœื—ืฉื•ื‘ ืขืœ ื›ืš --
07:50
if we tossed a coin eight million times,
153
470000
2000
ืื ื”ื™ื™ื ื• ืžื˜ื™ืœื™ื ืžื˜ื‘ืข 8 ืžื™ืœื™ื•ืŸ ืคืขืžื™ื,
07:52
then we'd expect a million head-tail-heads
154
472000
2000
ื”ื™ื™ื ื• ืžืฆืคื™ื ืœืžื™ืœื™ื•ืŸ ืขืฅ-ืคืœื™-ืขืฅ
07:54
and a million head-tail-tails -- but the head-tail-heads could occur in clumps.
155
474000
7000
ื•ืœืžื™ืœื™ื•ืŸ ืขืฅ-ืคืœื™-ืคืœื™ -- ืื‘ืœ ื”ืขืฅ-ืคืœื™-ืขืฅ ื”ื™ื• ื™ื›ื•ืœื™ื ืœื”ื•ืคื™ืข ื‘ืžืงื‘ืฆื™ื.
08:01
So if you want to put a million things down amongst eight million positions
156
481000
2000
ืื– ืื ืจื•ืฆื™ื ืœืฉื™ื ืžื™ืœื™ื•ืŸ ื“ื‘ืจื™ื ื‘ื™ืŸ 8 ืžื™ืœื™ื•ืŸ ืžืงื•ืžื•ืช
08:03
and you can have some of them overlapping, the clumps will be further apart.
157
483000
5000
ื›ืฉื—ืœืงื ื™ื›ื•ืœื™ื ืœื”ื™ื•ืช ื—ื•ืคืคื™ื, ื”ืจื•ื•ื—ื™ื ื‘ื™ืŸ ื”ืžืงื‘ืฆื™ื ื™ื”ื™ื• ื’ื“ื•ืœื™ื ื™ื•ืชืจ.
08:08
It's another way of getting the intuition.
158
488000
2000
ื–ื•ื”ื™ ื“ืจืš ืื—ืช ืœื”ืกื‘ืจ ืื™ื ื˜ื•ืื™ื˜ื™ื‘ื™.
08:10
What's the point I want to make?
159
490000
2000
ืžื” ืื ื™ ืจื•ืฆื” ืœื”ื’ื™ื“?
08:12
It's a very, very simple example, an easily stated question in probability,
160
492000
4000
ื–ื•ื”ื™ ื“ื•ื’ืžื ืžืื•ื“ ืžืื•ื“ ืคืฉื•ื˜ื”, ืฉืืœื” ื‘ื”ืกืชื‘ืจื•ืช ืฉืžื ื•ืกื—ืช ื‘ืงืœื•ืช,
08:16
which every -- you're in good company -- everybody gets wrong.
161
496000
3000
ืฉื›ื•ืœื -- ืืชื ื‘ื—ื‘ืจื” ื˜ื•ื‘ื” -- ื›ื•ืœื ื˜ื•ืขื™ื ื‘ื”.
08:19
This is my little diversion into my real passion, which is genetics.
162
499000
4000
ื–ื•ื”ื™ ืกื˜ื™ื™ื” ืงื˜ื ื” ืœืชืฉื•ืงื” ื”ืืžื™ืชื™ืช ืฉืœื™ - ื’ื ื˜ื™ืงื”.
08:23
There's a connection between head-tail-heads and head-tail-tails in genetics,
163
503000
3000
ื™ืฉ ืงืฉืจ ื‘ื™ืŸ ืขืฅ-ืคืœื™-ืขืฅ ื•ืขืฅ-ืคืœื™-ืคืœื™ ื‘ื’ื ื˜ื™ืงื”
08:26
and it's the following.
164
506000
3000
ื•ืืฆื™ื’ ืื•ืชื• ื‘ืคื ื™ื›ื.
08:29
When you toss a coin, you get a sequence of heads and tails.
165
509000
3000
ื›ืืฉืจ ืžื˜ื™ืœื™ื ืžื˜ื‘ืข, ืžืงื‘ืœื™ื ืกื“ืจื” ืฉืœ ืขืฆื™ื ื•ืคืœื™ื.
08:32
When you look at DNA, there's a sequence of not two things -- heads and tails --
166
512000
3000
ื›ืฉืžืชื‘ื•ื ื ื™ื ื‘ื“ื™.ืืŸ.ืื™ื™., ื™ืฉ ืกื“ืจื•ืช ืฉืื™ื ืŸ ืฉืœ 2 ื“ื‘ืจื™ื -- ืขืฆื™ื ื•ืคืœื™ื --
08:35
but four letters -- As, Gs, Cs and Ts.
167
515000
3000
ืืœื ืฉืœ 4 ืื•ืชื™ื•ืช -- A-ื™ื, G-ื™ื, C-ื™ื ื•-T-ื™ื.
08:38
And there are little chemical scissors, called restriction enzymes
168
518000
3000
ื™ืฉ ืžืกืคืจื™ื™ื ื›ื™ืžื™ื™ื ืงื˜ื ื™ื ืฉื ืงืจืื™ื ืื ื–ื™ืžื™ ื”ื’ื‘ืœื”
08:41
which cut DNA whenever they see particular patterns.
169
521000
2000
ืฉื—ื•ืชื›ื™ื ืืช ื”ื“ื™.ืืŸ.ืื™ื™. ื›ืฉื”ื ื ืชืงืœื™ื ื‘ื“ืคื•ืกื™ื ืžืกื•ื™ื™ืžื™ื.
08:43
And they're an enormously useful tool in modern molecular biology.
170
523000
4000
ื”ื ื›ืœื™ ืฉื™ืžื•ืฉื™ ื‘ื™ื•ืชืจ ื‘ื‘ื™ื•ืœื•ื’ื™ื” ืžื•ืœืงื•ืœืจื™ืช ืžื•ื“ืจื ื™ืช.
08:48
And instead of asking the question, "How long until I see a head-tail-head?" --
171
528000
3000
ื•ื‘ืžืงื•ื ืœืฉืื•ืœ, "ื›ืžื” ื–ืžืŸ ื™ืงื— ืขื“ ืฉื ืจืื” ืขืฅ-ืคืœื™-ืขืฅ?" --
08:51
you can ask, "How big will the chunks be when I use a restriction enzyme
172
531000
3000
ืืคืฉืจ ืœืฉืื•ืœ, "ืžื” ื™ื”ื™ื” ื’ื•ื“ืœ ื”ื—ืชื™ื›ื•ืช ื›ืฉืืฉืชืžืฉ ื‘ืื ื–ื™ื ื”ื’ื‘ืœื”
08:54
which cuts whenever it sees G-A-A-G, for example?
173
534000
4000
ืฉื—ื•ืชืš ื‘ื›ืœ ืคืขื ืฉื”ื•ื ืžื•ืฆื G-A-A-G, ืœื“ื•ื’ืžื?
08:58
How long will those chunks be?"
174
538000
2000
ืžื” ื™ื”ื™ื” ืื•ืจืš ื”ื—ืชื™ื›ื•ืช?"
09:00
That's a rather trivial connection between probability and genetics.
175
540000
5000
ื–ื”ื• ืงืฉืจ ื“ื™ ื˜ืจื™ื•ื•ื™ืืœื™ ื‘ื™ืŸ ื”ืกืชื‘ืจื•ืช ื•ื’ื ื˜ื™ืงื”.
09:05
There's a much deeper connection, which I don't have time to go into
176
545000
3000
ื™ืฉ ืงืฉืจ ื”ืจื‘ื” ื™ื•ืชืจ ืขืžื•ืง, ืฉืื™ืŸ ืœื™ ื–ืžืŸ ืœื”ื›ื ืก ืืœื™ื•,
09:08
and that is that modern genetics is a really exciting area of science.
177
548000
3000
ื•ื”ื•ื ืฉื’ื ื˜ื™ืงื” ืžื•ื“ืจื ื™ืช ื”ื™ื ืฉื˜ื— ืžื“ืขื™ ืžืื•ื“ ืžืจืชืง.
09:11
And we'll hear some talks later in the conference specifically about that.
178
551000
4000
ื‘ื”ืžืฉืš ื”ื•ืขื™ื“ื” ื ืฉืžืข ื›ืžื” ื”ืจืฆืื•ืช ืขืœ ื”ื ื•ืฉื.
09:15
But it turns out that unlocking the secrets in the information generated by modern
179
555000
4000
ืื‘ืœ ืžืชื‘ืจืจ ืฉื—ืฉื™ืคืช ื”ืกื•ื“ื•ืช ื‘ื ืชื•ื ื™ื ืฉืžื™ื•ืฆืจื™ื ืข"ื™
09:19
experimental technologies, a key part of that has to do with fairly sophisticated --
180
559000
5000
ื˜ื›ื ื•ืœื•ื’ื™ื•ืช ื ื™ืกื™ื•ื ื™ื•ืช ืžื•ื“ืจื ื™ื•ืช, ืžืจื›ื™ื‘ ืžืคืชื— ื‘ื—ืฉื™ืคื” ื–ื• --
09:24
you'll be relieved to know that I do something useful in my day job,
181
564000
3000
ื•ื“ืื™ ืชืฉืžื—ื• ืœืฉืžื•ืข ืฉืื ื™ ืขื•ืกืง ื‘ื“ื‘ืจื™ื ืฉื™ืžื•ืฉื™ื™ื ื‘ืขื‘ื•ื“ืช ื”ื™ื•ื-ื™ื•ื ืฉืœื™,
09:27
rather more sophisticated than the head-tail-head story --
182
567000
2000
ืžื•ืจื›ื‘ื™ื ื™ื•ืชืจ ืžืกื™ืคื•ืจ ื”ืขืฅ-ืคืœื™-ืขืฅ --
09:29
but quite sophisticated computer modelings and mathematical modelings
183
569000
4000
ืžื•ื“ืœื™ื ืžื—ืฉื‘ื™ื™ื ื•ืžื•ื“ืœื™ื ืžืชืžื˜ื™ื™ื
09:33
and modern statistical techniques.
184
573000
2000
ื•ื˜ื›ื ื™ืงื•ืช ืกื˜ื˜ื™ืกื˜ื™ื•ืช ืžื•ื“ืจื ื™ื•ืช ืžืชื•ื—ื›ืžื™ื ืœืžื“ื™.
09:35
And I will give you two little snippets -- two examples --
185
575000
3000
ืืฆื™ื’ ื‘ืคื ื™ื›ื 2 ื“ื•ื’ืžืื•ืช
09:38
of projects we're involved in in my group in Oxford,
186
578000
3000
ืœืคืจื•ื™ื™ืงื˜ื™ื ืฉื”ืงื‘ื•ืฆื” ืฉืœื™ ืžืื•ืงืกืคื•ืจื“ ืžืขื•ืจื‘ืช ื‘ื”ื.
09:41
both of which I think are rather exciting.
187
581000
2000
ืื ื™ ื—ื•ืฉื‘ ืฉืฉื ื™ื”ื ืžืจืชืงื™ื ืœืžื“ื™.
09:43
You know about the Human Genome Project.
188
583000
2000
ื•ื“ืื™ ืฉืžืขืชื ืขืœ ืคืจื•ื™ื™ืงื˜ ื”ื’ื ื•ื ื”ืื ื•ืฉื™.
09:45
That was a project which aimed to read one copy of the human genome.
189
585000
4000
ื–ื”ื• ืคืจื™ื™ืงื˜ ืฉืžื˜ืจืชื• ื”ื™ื™ืชื” ืงืจื™ืืช ืขื•ืชืง ืื—ื“ ืฉืœ ื”ื’ื ื•ื ื”ืื ื•ืฉื™.
09:51
The natural thing to do after you've done that --
190
591000
2000
ืœืื—ืจ ืฉื”ืคืจื•ื™ื™ืงื˜ ื”ื•ืฉืœื - ื”ื“ื‘ืจ ื”ื˜ื‘ืขื™ ืœืขืฉื•ืชื•
09:53
and that's what this project, the International HapMap Project,
191
593000
2000
ื”ื•ื ืžื” ืฉืขืฉื” ืคืจื•ื™ื™ืงื˜ ื”-HapMap ื”ื‘ื™ื ืœืื•ืžื™,
09:55
which is a collaboration between labs in five or six different countries.
192
595000
5000
ืฉื”ื•ื ืฉื™ืชื•ืฃ ืคืขื•ืœื” ื‘ื™ืŸ ืžืขื‘ื“ื•ืช ืž-5 ืื• 6 ืžื“ื™ื ื•ืช ืฉื•ื ื•ืช.
10:00
Think of the Human Genome Project as learning what we've got in common,
193
600000
4000
ื—ื™ืฉื‘ื• ืขืœ ืคืจื•ื™ื™ืงื˜ ื”ื’ื ื•ื ื”ืื ื•ืฉื™ ื›ืขืœ ืœืžื™ื“ื” ืฉืœ ืžื” ืฉืžืฉื•ืชืฃ ืœื ื•.
10:04
and the HapMap Project is trying to understand
194
604000
2000
ืคืจื•ื™ื™ืงื˜ ื”-HapMap ืžื ืกื” ืœื”ื‘ื™ืŸ
10:06
where there are differences between different people.
195
606000
2000
ืื™ืคื” ืงื™ื™ืžื™ื ื”ื”ื‘ื“ืœื™ื ื‘ื™ืŸ ื”ืื ืฉื™ื ื”ืฉื•ื ื™ื.
10:08
Why do we care about that?
196
608000
2000
ืœืžื” ื–ื” ืžืขื ื™ื™ืŸ ืื•ืชื ื•?
10:10
Well, there are lots of reasons.
197
610000
2000
ืžืกื™ื‘ื•ืช ืจื‘ื•ืช.
10:12
The most pressing one is that we want to understand how some differences
198
612000
4000
ื”ืกื™ื‘ื” ื”ื‘ื•ืขืจืช ื‘ื™ื•ืชืจ ื”ื™ื ืฉืื ื• ืจื•ืฆื™ื ืœื”ื‘ื™ืŸ ืื™ืš ื”ื‘ื“ืœื™ื ืžืกื•ื™ื™ืžื™ื
10:16
make some people susceptible to one disease -- type-2 diabetes, for example --
199
616000
4000
ื’ื•ืจืžื™ื ืœืื ืฉื™ื ืžืกื•ื™ื™ืžื™ื ืœื”ื™ื•ืช ืจื’ื™ืฉื™ื ืœืžื—ืœื” ืžืกื•ื™ื™ืžืช - ืกื•ื›ืจืช ืกื•ื’ 2, ืœื“ื•ื’ืžื,
10:20
and other differences make people more susceptible to heart disease,
200
620000
5000
ื•ื”ื‘ื“ืœื™ื ืื—ืจื™ื ื’ื•ืจืžื™ื ืœืื ืฉื™ื ืœื”ื™ื•ืช ื™ื•ืชืจ ืจื’ื™ืฉื™ื ืœืžื—ืœื•ืช ืœื‘,
10:25
or stroke, or autism and so on.
201
625000
2000
ืื• ืœืฉื‘ืฅ, ืื• ืœืื•ื˜ื™ื–ื ื•ื›ื•'.
10:27
That's one big project.
202
627000
2000
ื–ื” ืคืจื•ื™ื™ืงื˜ ื’ื“ื•ืœ ืื—ื“.
10:29
There's a second big project,
203
629000
2000
ืคืจื•ื™ื™ืงื˜ ื’ื“ื•ืœ ื ื•ืกืฃ,
10:31
recently funded by the Wellcome Trust in this country,
204
631000
2000
ืฉื ื•ืกื“ ืœืื—ืจื•ื ื” ืข"ื™ ื•ื•ืœืงื ื˜ืจืกื˜ ื‘ืžื“ื™ื ื” ื”ื–ืืช,
10:33
involving very large studies --
205
633000
2000
ื›ื•ืœืœ ืžื—ืงืจื™ื ืจื—ื‘ื™ ื”ื™ืงืฃ --
10:35
thousands of individuals, with each of eight different diseases,
206
635000
3000
ืืœืคื™ ืื ืฉื™ื, ื”ืกื•ื‘ืœื™ื ืžืื—ืช ืž-8 ืžื—ืœื•ืช ืฉื•ื ื•ืช,
10:38
common diseases like type-1 and type-2 diabetes, and coronary heart disease,
207
638000
4000
ืžื—ืœื•ืช ืฉื›ื™ื—ื•ืช ื›ืžื• ืกื•ื›ืจืช ืกื•ื’ 1 ื•ืกื•ื’ 2, ื•ืžื—ืœืช ืœื‘ ื›ืœื™ืœื™ืช,
10:42
bipolar disease and so on -- to try and understand the genetics.
208
642000
4000
ื”ืคืจืขื” ื“ื•-ืงื•ื˜ื‘ื™ืช ื•ื›ื•'. ืžื—ืงืจื™ื ืืœื• ืžื ืกื™ื ืœื”ื‘ื™ืŸ ืืช ื”ื’ื ื˜ื™ืงื”.
10:46
To try and understand what it is about genetic differences that causes the diseases.
209
646000
3000
ื”ื ืžื ืกื™ื ืœื”ื‘ื™ืŸ ืื™ื–ื” ืจื›ื™ื‘ ื‘ื”ื‘ื“ืœื™ื ื”ื’ื ื˜ื™ื™ื ื’ื•ืจื ืœืžื—ืœื•ืช.
10:49
Why do we want to do that?
210
649000
2000
ืœืžื” ืื ื—ื ื• ืจื•ืฆื™ื ืœืขืฉื•ืช ื–ืืช?
10:51
Because we understand very little about most human diseases.
211
651000
3000
ืžื›ื™ื•ื•ืŸ ืฉืื ื—ื ื• ืžื‘ื™ื ื™ื ืžืขื˜ ืžืื•ื“ ืขืœ ืจื•ื‘ ื”ืžื—ืœื•ืช ื”ืื ื•ืฉื™ื•ืช.
10:54
We don't know what causes them.
212
654000
2000
ืื ื—ื ื• ืœื ื™ื•ื“ืขื™ื ืžื” ื’ื•ืจื ืœื”ืŸ.
10:56
And if we can get in at the bottom and understand the genetics,
213
656000
2000
ื•ืื ื ืฆืœื™ื— ืœืจื“ืช ืœืฉื•ืจืฉ ื”ืขื ื™ื™ืŸ ื•ืœื”ื‘ื™ืŸ ืืช ื”ื’ื ื˜ื™ืงื”,
10:58
we'll have a window on the way the disease works,
214
658000
3000
ื™ื”ื™ื” ืœื ื• ื—ืœื•ืŸ ื”ืฆืฆื” ืœืื•ืคืŸ ื‘ื• ื”ืžื—ืœื” ืคื•ืขืœืช,
11:01
and a whole new way about thinking about disease therapies
215
661000
2000
ื•ื“ืจืš ื—ื“ืฉื” ืœื’ืžืจื™ ืœื—ืฉื•ื‘ ืขืœ ื˜ื™ืคื•ืœ ื‘ืžื—ืœื•ืช
11:03
and preventative treatment and so on.
216
663000
3000
ื•ืขืœ ื˜ื™ืคื•ืœ ืžื•ื ืข ื•ื›ื•'.
11:06
So that's, as I said, the little diversion on my main love.
217
666000
3000
ื•ื›ืžื• ืฉืืžืจืชื™, ื–ืืช ื”ื™ื™ืชื” ืกื˜ื™ื™ื” ืงื˜ื ื” ืืœ ืื”ื‘ืชื™ ื”ืขื™ืงืจื™ืช.
11:09
Back to some of the more mundane issues of thinking about uncertainty.
218
669000
5000
ื ื—ื–ื•ืจ ืœื›ืžื” ืžื”ื ื•ืฉืื™ื ื”ื™ื•ืชืจ ืืจืฆื™ื™ื ืฉืœ ื—ืฉื™ื‘ื” ืขืœ ื—ื•ืกืจ ื•ื“ืื•ืช.
11:14
Here's another quiz for you --
219
674000
2000
ื”ื ื” ืขื•ื“ ื—ื™ื“ื” ืขื‘ื•ืจื›ื --
11:16
now suppose we've got a test for a disease
220
676000
2000
ื ื ื™ื— ืฉื™ืฉ ืœื ื• ื‘ื“ื™ืงื” ืœืžื—ืœื”
11:18
which isn't infallible, but it's pretty good.
221
678000
2000
ืฉืœื ื—ืกื™ื ื” ืžืคื ื™ ื˜ืขื•ื™ื•ืช, ืื‘ืœ ื”ื™ื ื“ื™ ื˜ื•ื‘ื”.
11:20
It gets it right 99 percent of the time.
222
680000
3000
ื”ืชื•ืฆืื•ืช ืฉืœื” ืชืงื™ื ื•ืช 99 ืื—ื•ื– ืžื”ื–ืžืŸ.
11:23
And I take one of you, or I take someone off the street,
223
683000
3000
ื•ืื ื™ ืœื•ืงื— ืื—ื“ ืžื›ื, ืื• ืžื™ืฉื”ื• ืžื”ืจื—ื•ื‘,
11:26
and I test them for the disease in question.
224
686000
2000
ื•ื‘ื•ื“ืง ืื ื”ื•ื ื—ื•ืœื” ื‘ืžื—ืœื” ื”ื ื™ื“ื•ื ื”.
11:28
Let's suppose there's a test for HIV -- the virus that causes AIDS --
225
688000
4000
ื ื ื™ื— ืฉื–ื•ื”ื™ ื‘ื“ื™ืงื” ืœ-HIV -- ื”ื•ื™ืจื•ืก ืฉื’ื•ืจื ืœืื™ื™ื“ืก --
11:32
and the test says the person has the disease.
226
692000
3000
ื•ืฉื”ื‘ื“ื™ืงื” ืื•ืžืจืช ืฉื”ืื“ื ื—ื•ืœื”.
11:35
What's the chance that they do?
227
695000
3000
ืžื” ื”ืกื™ื›ื•ื™ ืฉื”ื•ื ืื›ืŸ ื—ื•ืœื”?
11:38
The test gets it right 99 percent of the time.
228
698000
2000
ืชื•ืฆืื•ืช ื”ื‘ื“ื™ืงื” ืชืงื™ื ื•ืช 99 ืื—ื•ื– ืžื”ื–ืžืŸ.
11:40
So a natural answer is 99 percent.
229
700000
4000
ืœื›ืŸ ืชืฉื•ื‘ื” ื˜ื‘ืขื™ืช ืชื”ื™ื” 99 ืื—ื•ื–.
11:44
Who likes that answer?
230
704000
2000
ืžื™ ืื•ื”ื‘ ืืช ื”ืชืฉื•ื‘ื” ื”ื–ืืช?
11:46
Come on -- everyone's got to get involved.
231
706000
1000
ืงื“ื™ืžื” -- ื›ื•ืœื ืฆืจื™ื›ื™ื ืœืขื ื•ืช.
11:47
Don't think you don't trust me anymore.
232
707000
2000
ืืœ ืชื—ืฉื‘ื• ืฉืืชื ื›ื‘ืจ ืœื ื‘ื•ื˜ื—ื™ื ื‘ื™.
11:49
(Laughter)
233
709000
1000
[ืฆื—ื•ืง]
11:50
Well, you're right to be a bit skeptical, because that's not the answer.
234
710000
3000
ืืชื ืฆื•ื“ืงื™ื ืื ืืชื ืงืฆืช ืกืคืงื ื™ื™ื, ื›ื™ ื–ืืช ืœื ื”ืชืฉื•ื‘ื” ื”ื ื›ื•ื ื”.
11:53
That's what you might think.
235
713000
2000
ื–ื” ืžื” ืฉืืชื ืื•ืœื™ ื—ื•ืฉื‘ื™ื.
11:55
It's not the answer, and it's not because it's only part of the story.
236
715000
3000
ื–ืืช ืœื ื”ืชืฉื•ื‘ื”, ืื‘ืœ ืœื ื‘ื’ืœืœ ืฉื–ื” ืจืง ื—ืœืง ืžื”ืกื™ืคื•ืจ.
11:58
It actually depends on how common or how rare the disease is.
237
718000
3000
ืœืžืขืฉื”, ื–ื” ืชืœื•ื™ ื‘ืžื™ื“ื” ื‘ื” ื”ืžื—ืœื” ื”ื™ื ืฉื›ื™ื—ื” ืื• ื ื“ื™ืจื”.
12:01
So let me try and illustrate that.
238
721000
2000
ืื ืกื” ืœื”ืžื—ื™ืฉ ื–ืืช ืขื‘ื•ืจื›ื.
12:03
Here's a little caricature of a million individuals.
239
723000
4000
ื–ื•ื”ื™ ืงืจื™ืงื˜ื•ืจื” ืงื˜ื ื” ืฉืœ ืžื™ืœื™ื•ืŸ ืื ืฉื™ื.
12:07
So let's think about a disease that affects --
240
727000
3000
ื ื—ืฉื•ื‘ ืขืœ ืžื—ืœื” ื“ื™ ื ื“ื™ืจื”,
12:10
it's pretty rare, it affects one person in 10,000.
241
730000
2000
ืฉืžืฉืคื™ืขื” ืขืœ ืื“ื ืื—ื“ ืžืชื•ืš 10,000.
12:12
Amongst these million individuals, most of them are healthy
242
732000
3000
ื‘ื™ืŸ ืžื™ืœื™ื•ืŸ ื”ืื ืฉื™ื ื”ืืœื”, ืจื•ื‘ื ื‘ืจื™ืื™ื
12:15
and some of them will have the disease.
243
735000
2000
ื•ื—ืœืงื ื™ืกื‘ืœื• ืžื”ืžื—ืœื”.
12:17
And in fact, if this is the prevalence of the disease,
244
737000
3000
ืœืžืขืฉื”, ืื ื–ื•ื”ื™ ื”ืฉื›ื™ื—ื•ืช ืฉืœ ื”ืžื—ืœื”,
12:20
about 100 will have the disease and the rest won't.
245
740000
3000
ื›-100 ื™ื”ื™ื• ื—ื•ืœื™ื ื•ื›ืœ ื”ืฉืืจ ื™ื”ื™ื• ื‘ืจื™ืื™ื.
12:23
So now suppose we test them all.
246
743000
2000
ื ื ื™ื— ืฉืื ื—ื ื• ื‘ื•ื“ืงื™ื ืืช ื›ื•ืœื.
12:25
What happens?
247
745000
2000
ืžื” ื™ืงืจื”?
12:27
Well, amongst the 100 who do have the disease,
248
747000
2000
ืžื‘ื™ืŸ ื”-100 ืฉื—ื•ืœื™ื,
12:29
the test will get it right 99 percent of the time, and 99 will test positive.
249
749000
5000
ื”ื‘ื“ื™ืงื” ืชืชืŸ ืชื•ืฆืื” ืชืงื™ื ื” 99 ืื—ื•ื– ืžื”ื–ืžืŸ, ื•ืœ-99 ืžื”ื ื”ืชืฉื•ื‘ื” ืชื”ื™ื” ื—ื™ื•ื‘ื™ืช.
12:34
Amongst all these other people who don't have the disease,
250
754000
2000
ืžื‘ื™ืŸ ื”ืื—ืจื™ื ืฉืœื ื—ื•ืœื™ื,
12:36
the test will get it right 99 percent of the time.
251
756000
3000
ื”ื‘ื“ื™ืงื” ืชืชืŸ ืชืฉื•ื‘ื” ืชืงื™ื ื” 99 ืื—ื•ื– ืžื”ื–ืžืŸ.
12:39
It'll only get it wrong one percent of the time.
252
759000
2000
ื”ื™ื ืชืชืŸ ืชื•ืฆืื” ืฉื’ื•ื™ื” ืจืง ื‘ืื—ื•ื– ืื—ื“ ืžื”ื–ืžืŸ.
12:41
But there are so many of them that there'll be an enormous number of false positives.
253
761000
4000
ืื‘ืœ ืžื›ื™ื•ื•ืŸ ืฉื™ืฉ ื›ืœ ื›ืš ื”ืจื‘ื” ืื ืฉื™ื - ื™ื”ื™ื” ืžืกืคืจ ืขืฆื•ื ืฉืœ ืชืฉื•ื‘ื•ืช ื—ื™ื•ื‘ื™ื•ืช ืžื•ื˜ืขื•ืช.
12:45
Put that another way --
254
765000
2000
ื•ื‘ื ื™ืกื•ื— ืื—ืจ --
12:47
of all of them who test positive -- so here they are, the individuals involved --
255
767000
5000
ืžื‘ื™ืŸ ื›ืœ ืืœื” ืฉืžืงื‘ืœื™ื ืชืฉื•ื‘ื” ื—ื™ื•ื‘ื™ืช -- ื”ื ื” ื”ื --
12:52
less than one in 100 actually have the disease.
256
772000
5000
ืคื—ื•ืช ืžืื—ื“ ืžืžืื” ืื›ืŸ ื—ื•ืœื”.
12:57
So even though we think the test is accurate, the important part of the story is
257
777000
4000
ื•ืœืžืจื•ืช ืฉืื ื—ื ื• ื—ื•ืฉื‘ื™ื ืฉื”ื‘ื“ื™ืงื” ืžื“ื•ื™ื™ืงืช, ื”ื—ืœืง ื”ื—ืฉื•ื‘ ืฉืœ ื”ืกื™ืคื•ืจ ื”ื•ื
13:01
there's another bit of information we need.
258
781000
3000
ืฉื™ืฉื ื ื ืชื•ื ื™ื ื ื•ืกืคื™ื ืฉืื ื• ื–ืงื•ืงื™ื ืœื”ื.
13:04
Here's the key intuition.
259
784000
2000
ื–ื•ื”ื™ ื”ืื™ื ื˜ื•ืื™ืฆื™ื” ื”ืžื•ื‘ื™ืœื”.
13:07
What we have to do, once we know the test is positive,
260
787000
3000
ื‘ืจื’ืข ืฉืื ื—ื ื• ื™ื•ื“ืขื™ื ืฉื”ื‘ื“ื™ืงื” ื—ื™ื•ื‘ื™ืช,
13:10
is to weigh up the plausibility, or the likelihood, of two competing explanations.
261
790000
6000
ืขืœื™ื ื• ืœืฉืงื•ืœ ืืช ื”ืกื‘ื™ืจื•ืช, ืื• ื”ืกื™ื›ื•ื™, ืฉืœ ืฉื ื™ ื”ืกื‘ืจื™ื ืืคืฉืจื™ื™ื.
13:16
Each of those explanations has a likely bit and an unlikely bit.
262
796000
3000
ืœื›ืœ ืื—ื“ ืžื”ื”ืกื‘ืจื™ื ื™ืฉ ื—ืœืง ืกื‘ื™ืจ ื•ื—ืœืง ืœื ืกื‘ื™ืจ.
13:19
One explanation is that the person doesn't have the disease --
263
799000
3000
ื”ืกื‘ืจ ืื—ื“ ื”ื•ื ืฉื”ืื“ื ืœื ื—ื•ืœื” --
13:22
that's overwhelmingly likely, if you pick someone at random --
264
802000
3000
ื–ื” ืกื‘ื™ืจ ื‘ื™ื•ืชืจ, ืื ื‘ื•ื—ืจื™ื ืžื™ืฉื”ื• ื‘ืื•ืคืŸ ืืงืจืื™ --
13:25
but the test gets it wrong, which is unlikely.
265
805000
3000
ืื‘ืœ ื”ื‘ื“ื™ืงื” ื˜ื•ืขื” - ื“ื‘ืจ ืœื ืกื‘ื™ืจ.
13:29
The other explanation is that the person does have the disease -- that's unlikely --
266
809000
3000
ื”ื”ืกื‘ืจ ื”ืฉื ื™ ื”ื•ื ืฉื”ืื“ื ื—ื•ืœื” -- ื™ืฉ ืœื›ืš ืกื‘ื™ืจื•ืช ื ืžื•ื›ื” --
13:32
but the test gets it right, which is likely.
267
812000
3000
ื•ื”ื‘ื“ื™ืงื” ืชืงื™ื ื” - ืœื›ืš ื™ืฉ ืกื‘ื™ืจื•ืช ื’ื‘ื•ื”ื”.
13:35
And the number we end up with --
268
815000
2000
ื•ื”ืžืกืคืจ ืฉืื ื—ื ื• ืžืงื‘ืœื™ื ื‘ืกื•ืคื• ืฉืœ ื“ื‘ืจ --
13:37
that number which is a little bit less than one in 100 --
269
817000
3000
ื”ืžืกืคืจ ืฉืงืฆืช ื™ื•ืชืจ ืงื˜ืŸ ืž-1 ืœ-100 --
13:40
is to do with how likely one of those explanations is relative to the other.
270
820000
6000
ืื•ืžืจ ืžื” ื”ืกื‘ื™ืจื•ืช ืฉืœ ื”ืกื‘ืจ ืื—ื“ ืœืขื•ืžืช ื”ืฉื ื™.
13:46
Each of them taken together is unlikely.
271
826000
2000
ืœื›ืœ ืื—ื“ ืžื”ื ืกื‘ื™ืจื•ืช ื ืžื•ื›ื”.
13:49
Here's a more topical example of exactly the same thing.
272
829000
3000
ื”ื ื” ื“ื•ื’ืžื ื™ื•ืชืจ ืืงื˜ื•ืืœื™ืช ืœืื•ืชื• ื”ื“ื‘ืจ.
13:52
Those of you in Britain will know about what's become rather a celebrated case
273
832000
4000
ืื ืืชื ืžื‘ืจื™ื˜ื ื™ื” ืืชื ืžื›ื™ืจื™ื ืืช ื”ืžืงืจื” ื”ืžืคื•ืจืกื
13:56
of a woman called Sally Clark, who had two babies who died suddenly.
274
836000
5000
ืฉืœ ืืฉื” ื‘ืฉื ืกืืœื™ ืงืœืืจืง, ืฉื”ื™ื• ืœื” 2 ืชื™ื ื•ืงื•ืช ืฉืžืชื• ื‘ืคืชืื•ืžื™ื•ืช.
14:01
And initially, it was thought that they died of what's known informally as "cot death,"
275
841000
4000
ื‘ื”ืชื—ืœื”, ื—ืฉื‘ื• ืฉื”ื ืžืชื• ืžืžื” ืฉื™ื“ื•ืข ื‘ืื•ืคืŸ ืœื ืคื•ืจืžืœื™ ื›"ืžื•ื•ืช ื‘ืขืจื™ืกื”",
14:05
and more formally as "Sudden Infant Death Syndrome."
276
845000
3000
ื•ื‘ืื•ืคืŸ ื™ื•ืชืจ ืคื•ืจืžืœื™ ื›ืชืกืžื•ื ืช ื”ืžื•ื•ืช ื”ืคืชืื•ืžื™ ื‘ื™ื ืงื•ืช.
14:08
For various reasons, she was later charged with murder.
277
848000
2000
ืžืกื™ื‘ื•ืช ืฉื•ื ื•ืช, ื”ื™ื ื”ื•ืจืฉืขื” ืžืื•ื—ืจ ื™ื•ืชืจ ื‘ืจืฆื—.
14:10
And at the trial, her trial, a very distinguished pediatrician gave evidence
278
850000
4000
ื•ื‘ืžืฉืคื˜ ืฉืœื”, ืจื•ืคื ื™ืœื“ื™ื ืžืื•ื“ ืžื›ื•ื‘ื“ ื”ืขื™ื“
14:14
that the chance of two cot deaths, innocent deaths, in a family like hers --
279
854000
5000
ืฉื”ืกื™ื›ื•ื™ ืœืฉื ื™ ืžืงืจื™ื ืฉืœ ืžื•ื•ืช ื‘ืขืจื™ืกื”, ืžืงืจื™ ืžื•ื•ืช ืชืžื™ืžื™ื, ื‘ืžืฉืคื—ื” ื›ืžื• ืฉืœื”,
14:19
which was professional and non-smoking -- was one in 73 million.
280
859000
6000
ืฉื”ื™ื™ืชื” ืžืงืฆื•ืขื™ืช ื•ืœื ืžืขืฉื ืช, ื”ื•ื 1 ืœ-73 ืžื™ืœื™ื•ืŸ.
14:26
To cut a long story short, she was convicted at the time.
281
866000
3000
ืื ื ืงืฆืจ ืกื™ืคื•ืจ ืืจื•ืš, ื”ื™ื ื”ื•ืจืฉืขื” ื‘ืื•ืชื• ื”ื–ืžืŸ.
14:29
Later, and fairly recently, acquitted on appeal -- in fact, on the second appeal.
282
869000
5000
ืžืื•ื—ืจ ื™ื•ืชืจ, ื•ื“ื™ ืœืื—ืจื•ื ื”, ื”ื™ื ื–ื•ื›ืชื” ืœืื—ืจ ืขืจืขื•ืจ -- ืœืžืขืฉื”, ืœืื—ืจ ื”ืขืจืขื•ืจ ื”ืฉื ื™ ืฉืœื”.
14:34
And just to set it in context, you can imagine how awful it is for someone
283
874000
4000
ื•ื›ื“ื™ ืฉืชืจืื• ืืช ื›ืœ ื”ืชืžื•ื ื”, ื ืกื• ืœื“ืžื™ื™ืŸ ื›ืžื” ื–ื” ื ื•ืจื
14:38
to have lost one child, and then two, if they're innocent,
284
878000
3000
ืœืื‘ื“ ื™ืœื“ ืื—ื“, ื•ืื– ืขื•ื“ ืื—ื“ - ื›ืฉืืชื” ื—ืฃ ืžืคืฉืข,
14:41
to be convicted of murdering them.
285
881000
2000
ื•ืœื”ื™ื•ืช ืžื•ืจืฉืข ื‘ืจืฆื— ืฉืœื”ื.
14:43
To be put through the stress of the trial, convicted of murdering them --
286
883000
2000
ืœืขื‘ื•ืจ ืืช ื›ืœ ื”ืžืชื— ืฉืœ ื”ืžืฉืคื˜, ืœื”ื™ื•ืช ืžื•ืจืฉืข ื‘ืจืฆื—,
14:45
and to spend time in a women's prison, where all the other prisoners
287
885000
3000
ื•ืœื”ื’ื™ืข ืœื‘ื™ืช ื”ืกื•ื”ืจ ืœื ืฉื™ื, ื‘ื• ื›ืœ ืฉืืจ ื”ืืกื™ืจื•ืช
14:48
think you killed your children -- is a really awful thing to happen to someone.
288
888000
5000
ื—ื•ืฉื‘ื•ืช ืฉืจืฆื—ืช ืืช ื”ื™ืœื“ื™ื ืฉืœืš -- ื–ื” ื“ื‘ืจ ื ื•ืจื ื‘ื™ื•ืชืจ.
14:53
And it happened in large part here because the expert got the statistics
289
893000
5000
ื•ื›ืืŸ, ื”ื•ื ื”ืชืจื—ืฉ ื‘ื—ืœืงื• ื”ื’ื“ื•ืœ ื‘ื’ืœืœ ืฉืชื™ ื˜ืขื•ื™ื•ืช ื—ืžื•ืจื•ืช
14:58
horribly wrong, in two different ways.
290
898000
3000
ืฉืœ ื”ืžื•ืžื—ื” ื‘ื—ื™ืฉื•ื‘ื™ ื”ืกื˜ื˜ื™ืกื˜ื™ืงื” ืฉืœื•.
15:01
So where did he get the one in 73 million number?
291
901000
4000
ืื™ืš ื”ื•ื ื”ื’ื™ืข ืœืžืกืคืจ ืฉืœ 1 ืœ-73 ืžื™ืœื™ื•ืŸ?
15:05
He looked at some research, which said the chance of one cot death in a family
292
905000
3000
ื”ื•ื ืžืฆื ืžื—ืงืจ ืฉืืžืจ ืฉื”ืกื™ื›ื•ื™ ืœืžื•ื•ืช ื‘ืขืจื™ืกื” ืื—ื“ ืœืžืฉืคื—ื”
15:08
like Sally Clark's is about one in 8,500.
293
908000
5000
ื›ืžื• ื–ื• ืฉืœ ืกืืœื™ ืงืœืืจืง ื”ื•ื ื›-1 ืœ-8,500.
15:13
So he said, "I'll assume that if you have one cot death in a family,
294
913000
4000
ื•ืื– ื”ื•ื ืืžืจ, "ืื ื™ ืžื ื™ื— ืฉืื ื™ืฉ ืžืงืจื” ืื—ื“ ืฉืœ ืžื•ื•ืช ื‘ืขืจื™ืกื” ื‘ืžืฉืคื—ื”,
15:17
the chance of a second child dying from cot death aren't changed."
295
917000
4000
ื”ืกื™ื›ื•ื™ ืฉื™ืœื“ ื ื•ืกืฃ ื™ืžื•ืช ืžืžื•ื•ืช ื‘ืขืจื™ืกื” ืœื ืžืฉืชื ื”."
15:21
So that's what statisticians would call an assumption of independence.
296
921000
3000
ื–ื•ื”ื™ ืžื” ืฉืกื˜ื˜ื™ืกื˜ื™ืงืื™ื ืžื›ื ื™ื ื”ื ื—ืช ืื™-ืชืœื•ืช.
15:24
It's like saying, "If you toss a coin and get a head the first time,
297
924000
2000
ื–ื” ื›ืžื• ืœื”ื’ื™ื“, "ืื ืžื˜ื™ืœื™ื ืžื˜ื‘ืข ื•ืžืงื‘ืœื™ื ืขืฅ,
15:26
that won't affect the chance of getting a head the second time."
298
926000
3000
ื–ื” ืœื ืžืฉืคื™ืข ืขืœ ื”ืกื™ื›ื•ื™ ืœืงื‘ืœืช ืขืฅ ื‘ืคืขื ื”ืฉื ื™ื™ื”."
15:29
So if you toss a coin twice, the chance of getting a head twice are a half --
299
929000
5000
ื•ืœื›ืŸ ืื ืžื˜ื™ืœื™ื ืžื˜ื‘ืข ืคืขืžื™ื™ื, ื”ืกื™ื›ื•ื™ ืœืงื‘ืœ ืขืฅ ืคืขืžื™ื™ื ื”ื•ื
15:34
that's the chance the first time -- times a half -- the chance a second time.
300
934000
3000
ื—ืฆื™ - ื”ืกื™ื›ื•ื™ ืœืขืฅ ื‘ืคืขื ื”ืจืืฉื•ื ื”, ื›ืคื•ืœ ื—ืฆื™ - ื”ืกื™ื›ื•ื™ ืœืขืฅ ื‘ืคืขื ื”ืฉื ื™ื™ื”.
15:37
So he said, "Here,
301
937000
2000
ืื– ื”ื•ื ืืžืจ,
15:39
I'll assume that these events are independent.
302
939000
4000
"ืื ื™ ืื ื™ื— ืฉืฉื ื™ ื”ืžืื•ืจืขื•ืช ื”ื ื‘ืœืชื™ ืชืœื•ื™ื™ื.
15:43
When you multiply 8,500 together twice,
303
943000
2000
ื›ืฉืžื›ืคื™ืœื™ื 8,500 ื‘-8,500,
15:45
you get about 73 million."
304
945000
2000
ืžืงื‘ืœื™ื ื‘ืขืจืš 73 ืžื™ืœื™ื•ืŸ."
15:47
And none of this was stated to the court as an assumption
305
947000
2000
ื•ื”ื”ื ื—ื” ื”ื–ืืช ืœื ื”ื•ืฆื’ื” ื‘ืคื ื™ ื‘ื™ืช ื”ืžืฉืคื˜
15:49
or presented to the jury that way.
306
949000
2000
ืื• ื‘ืคื ื™ ื—ื‘ืจ ื”ืžื•ืฉื‘ืขื™ื ื‘ืฆื•ืจื” ื”ื–ืืช.
15:52
Unfortunately here -- and, really, regrettably --
307
952000
3000
ืœืจื•ืข ื”ืžื–ืœ ื›ืืŸ, ื•ื‘ืื•ืคืŸ ืžืฆืขืจ ื‘ื™ื•ืชืจ -
15:55
first of all, in a situation like this you'd have to verify it empirically.
308
955000
4000
ืงื•ื“ื ื›ืœ, ื‘ืžืฆื‘ ื›ื–ื” ืฆืจื™ืš ืœื•ื•ื“ื ืืช ื”ื ืชื•ื ื™ื ื‘ืื•ืคืŸ ืืžืคื™ืจื™.
15:59
And secondly, it's palpably false.
309
959000
2000
ื•ื“ื‘ืจ ืฉื ื™, ื–ื” ื‘ืคื™ืจื•ืฉ ืœื ื ื›ื•ืŸ.
16:02
There are lots and lots of things that we don't know about sudden infant deaths.
310
962000
5000
ื™ืฉ ื”ืžื•ืŸ ื“ื‘ืจื™ื ืœื ื™ื“ื•ืขื™ื ืขืœ ืžื•ื•ืช ื‘ืขืจื™ืกื”.
16:07
It might well be that there are environmental factors that we're not aware of,
311
967000
3000
ื™ืชื›ืŸ ืžืื•ื“ ืฉืงื™ื™ืžื™ื ื’ื•ืจืžื™ื ืกื‘ื™ื‘ืชื™ื™ื ืฉืื ื—ื ื• ืœื ืžื•ื“ืขื™ื ืœื”ื,
16:10
and it's pretty likely to be the case that there are
312
970000
2000
ื•ื™ืฉื ื” ืกื‘ื™ืจื•ืช ื’ื‘ื•ื”ื” ืฉืžืขื•ืจื‘ื™ื ื‘ื›ืš
16:12
genetic factors we're not aware of.
313
972000
2000
ื’ื•ืจืžื™ื ื’ื ื˜ื™ื™ื ืฉืื ื—ื ื• ืœื ืžื•ื“ืขื™ื ืœืงื™ื•ืžื.
16:14
So if a family suffers from one cot death, you'd put them in a high-risk group.
314
974000
3000
ื•ืœื›ืŸ, ืื ื‘ืžืฉืคื—ื” ืžืชืจื—ืฉ ืžื•ื•ืช ื‘ืขืจื™ืกื”, ืฆืจื™ืš ืœื”ื›ื ื™ืก ืื•ืชื” ืœืงื‘ื•ืฆื” ื‘ืกื™ื›ื•ืŸ ื’ื‘ื•ื”.
16:17
They've probably got these environmental risk factors
315
977000
2000
ืงืจื•ื‘ ืœื•ื“ืื™ ืฉื™ืฉ ืœื” ื’ื•ืจืžื™ ืกื™ื›ื•ืŸ ืกื‘ื™ื‘ืชื™ื™ื
16:19
and/or genetic risk factors we don't know about.
316
979000
3000
ื•/ืื• ื’ื•ืจืžื™ ืกื™ื›ื•ืŸ ื’ื ื˜ื™ื™ื ืฉืื ื—ื ื• ืœื ืžื›ื™ืจื™ื.
16:22
And to argue, then, that the chance of a second death is as if you didn't know
317
982000
3000
ื•ืœื›ืŸ, ื”ื˜ืขื ื” ืฉื”ืกื™ื›ื•ื™ ืœืžื•ื•ืช ืฉื ื™ ื‘ืžืฉืคื—ื” ื–ื”ื” ืœืžืงืจื” ื‘ื•
16:25
that information is really silly.
318
985000
3000
ื”ื ืชื•ื ื™ื ืœื ื™ื“ื•ืขื™ื - ื”ื™ื ืžื˜ื•ืคืฉืช ื‘ื™ื•ืชืจ.
16:28
It's worse than silly -- it's really bad science.
319
988000
4000
ื•ื™ื•ืชืจ ืžืžื˜ื•ืคืฉืช -- ื–ื”ื• ืžื“ืข ื’ืจื•ืข ื‘ื™ื•ืชืจ.
16:32
Nonetheless, that's how it was presented, and at trial nobody even argued it.
320
992000
5000
ื•ื‘ื›ืœ ื–ืืช, ื–ื• ื”ื“ืจืš ื‘ื” ื”ืขื ื™ื™ืŸ ื”ื•ืฆื’, ื•ื‘ื‘ื™ืช ื”ืžืฉืคื˜ ืืฃ ืื—ื“ ืœื ื ื™ืกื” ืœื˜ืขื•ืŸ ื ื’ื“ื•.
16:37
That's the first problem.
321
997000
2000
ื–ื•ื”ื™ ื”ื‘ืขื™ื” ื”ืจืืฉื•ื ื”.
16:39
The second problem is, what does the number of one in 73 million mean?
322
999000
4000
ื”ื‘ืขื™ื” ื”ืฉื ื™ื” ื”ื™ื, ืžื” ื”ืžืฉืžืขื•ืช ืฉืœ 1 ืœ-73 ืžื™ืœื™ื•ืŸ?
16:43
So after Sally Clark was convicted --
323
1003000
2000
ืื—ืจื™ ืฉืกืืœื™ ืงืœืืจืง ื”ื•ืจืฉืขื”,
16:45
you can imagine, it made rather a splash in the press --
324
1005000
4000
ืืชื ื™ื›ื•ืœื™ื ืœืชืืจ ืœืขืฆืžื›ื ืฉื”ื™ื” ืจืขืฉ ื’ื“ื•ืœ ื‘ืชืงืฉื•ืจืช.
16:49
one of the journalists from one of Britain's more reputable newspapers wrote that
325
1009000
7000
ืื—ื“ ืžื”ืขื™ืชื•ื ืื™ื ืžืื—ื“ ื”ืขื™ืชื•ื ื™ื ื”ืžื•ืขืจื›ื™ื ื™ื•ืชืจ ื‘ื‘ืจื™ื˜ื ื™ื” ื›ืชื‘
16:56
what the expert had said was,
326
1016000
2000
ืฉืžื” ืฉื”ืžื•ืžื—ื” ืืžืจ ื”ื•ื,
16:58
"The chance that she was innocent was one in 73 million."
327
1018000
5000
ืฉ"ื”ืกื™ื›ื•ื™ ืฉื”ื™ื ื—ืคื” ืžืคืฉืข ื”ื•ื 1 ืœ-73 ืžื™ืœื™ื•ืŸ."
17:03
Now, that's a logical error.
328
1023000
2000
ื–ื•ื”ื™ ื˜ืขื•ืช ืœื•ื’ื™ืช.
17:05
It's exactly the same logical error as the logical error of thinking that
329
1025000
3000
ื–ื•ื”ื™ ืื•ืชื” ื˜ืขื•ืช ืœื•ื’ื™ืช ื›ืžื• ื”ื˜ืขื•ืช ืœื—ืฉื•ื‘
17:08
after the disease test, which is 99 percent accurate,
330
1028000
2000
ืฉืื—ืจื™ ื”ื‘ื“ื™ืงื” ืœืžื—ืœื”, ืฉืžื“ื•ื™ื™ืงืช ื‘-99 ืื—ื•ื–,
17:10
the chance of having the disease is 99 percent.
331
1030000
4000
ื”ืกื™ื›ื•ื™ ืœื—ืœื•ืช ื”ื•ื 99 ืื—ื•ื–.
17:14
In the disease example, we had to bear in mind two things,
332
1034000
4000
ื‘ื“ื•ื’ืžื ืฉืœ ื”ืžื—ืœื”, ื”ื™ื™ื ื• ืฆืจื™ื›ื™ื ืœื–ื›ื•ืจ ืฉื ื™ ื“ื‘ืจื™ื,
17:18
one of which was the possibility that the test got it right or not.
333
1038000
4000
ื”ืจืืฉื•ืŸ ื”ื•ื ื”ืืคืฉืจื•ืช ืฉื”ื‘ื“ื™ืงื” ื”ื™ื™ืชื” ืชืงื™ื ื” ืื• ืœื ืชืงื™ื ื”.
17:22
And the other one was the chance, a priori, that the person had the disease or not.
334
1042000
4000
ื•ื”ืฉื ื™ ื”ื•ื ื”ืกื™ื›ื•ื™, ื-ืคืจื™ื•ืจื™, ืฉื”ืื“ื ื—ื•ืœื” ืื• ื‘ืจื™ื.
17:26
It's exactly the same in this context.
335
1046000
3000
ื–ื” ื‘ื“ื™ื•ืง ืื•ืชื• ื“ื‘ืจ ื‘ื”ืงืฉืจ ื”ื–ื”.
17:29
There are two things involved -- two parts to the explanation.
336
1049000
4000
ื™ืฉ ืฉื ื™ ื“ื‘ืจื™ื ืžืขื•ืจื‘ื™ื -- ืฉื ื™ ื—ืœืงื™ื ืœื”ืกื‘ืจ.
17:33
We want to know how likely, or relatively how likely, two different explanations are.
337
1053000
4000
ืื ื—ื ื• ืจื•ืฆื™ื ืœื“ืขืช ืžื” ื”ืกื™ื›ื•ื™, ืื• ืžื” ื”ืกื™ื›ื•ื™ ื”ื™ื—ืกื™ ืฉืœ ืฉื ื™ ื”ืกื‘ืจื™ื ืืคืฉืจื™ื™ื.
17:37
One of them is that Sally Clark was innocent --
338
1057000
3000
ื”ื”ืกื‘ืจ ื”ืจืืฉื•ืŸ ื”ื•ื ืฉืกืืœื™ ืงืœืืจืง ื—ืคื” ืžืคืฉืข --
17:40
which is, a priori, overwhelmingly likely --
339
1060000
2000
ืฉืœื• ื™ืฉ, ื-ืคืจื™ื•ืจื™, ืกื™ื›ื•ื™ ื’ื‘ื•ื” ืžืื•ื“ -
17:42
most mothers don't kill their children.
340
1062000
3000
ืจื•ื‘ ื”ืืžื”ื•ืช ืœื ื”ื•ืจื’ื•ืช ืืช ื”ื™ืœื“ื™ื ืฉืœื”ืŸ.
17:45
And the second part of the explanation
341
1065000
2000
ื”ื—ืœืง ื”ืฉื ื™ ืฉืœ ื”ื”ืกื‘ืจ
17:47
is that she suffered an incredibly unlikely event.
342
1067000
3000
ื”ื•ื ืฉื”ื™ื ืกื‘ืœื” ืžืžืงืจื” ืขื ืกื‘ื™ืจื•ืช ืžืื•ื“ ื ืžื•ื›ื”.
17:50
Not as unlikely as one in 73 million, but nonetheless rather unlikely.
343
1070000
4000
ืœื ืกื™ื›ื•ื™ ื ืžื•ืš ื›ืžื• 1 ืœ-73 ืžื™ืœื™ื•ืŸ, ืื‘ืœ ื‘ื›ืœ ื–ืืช ื‘ืกื‘ื™ืจื•ืช ื“ื™ ื ืžื•ื›ื”.
17:54
The other explanation is that she was guilty.
344
1074000
2000
ื”ื”ืกื‘ืจ ื”ืฉื ื™ ื”ื•ื ืฉื”ื™ื ืืฉืžื”.
17:56
Now, we probably think a priori that's unlikely.
345
1076000
2000
ืงืจื•ื‘ ืœื•ื“ืื™ ืฉืื ื—ื ื• ื—ื•ืฉื‘ื™ื ืžืœื›ืชื—ื™ืœื” ืฉื–ื” ืœื ืกื‘ื™ืจ.
17:58
And we certainly should think in the context of a criminal trial
346
1078000
3000
ื•ื“ืื™ ืฉืื ื—ื ื• ืฆืจื™ื›ื™ื ืœื—ืฉื•ื‘ ื‘ื”ืงืฉืจ ืฉืœ ืžืฉืคื˜ ืคืœื™ืœื™
18:01
that that's unlikely, because of the presumption of innocence.
347
1081000
3000
ืฉื™ืฉ ืœื›ืš ืกื™ื›ื•ื™ ื ืžื•ืš, ื‘ื’ืœืœ ื”ื ื—ืช ื”ื—ืคื•ืช ืžืคืฉืข.
18:04
And then if she were trying to kill the children, she succeeded.
348
1084000
4000
ื•ืฉืื ื”ื™ื ื ื™ืกืชื” ืœื”ืจื•ื’ ืืช ื”ื™ืœื“ื™ื, ื”ื™ื ื”ืฆืœื™ื—ื”.
18:08
So the chance that she's innocent isn't one in 73 million.
349
1088000
4000
ืื ื›ืš, ื”ืกื™ื›ื•ื™ ืฉื”ื™ื ื—ืคื” ืžืคืฉืข ืื™ื ื• 1 ืœ-73 ืžื™ืœื™ื•ืŸ.
18:12
We don't know what it is.
350
1092000
2000
ืื ื—ื ื• ืœื ื™ื•ื“ืขื™ื ืžื” ื”ื•ื.
18:14
It has to do with weighing up the strength of the other evidence against her
351
1094000
4000
ืฆืจื™ืš ืœืงื—ืช ื‘ื—ืฉื‘ื•ืŸ ืืช ื—ื•ื–ืง ื”ืจืื™ื•ืช ื”ืื—ืจื•ืช ื ื’ื“ื”
18:18
and the statistical evidence.
352
1098000
2000
ื•ืืช ื”ืจืื™ื•ืช ื”ืกื˜ื˜ื™ืกื˜ื™ื•ืช.
18:20
We know the children died.
353
1100000
2000
ืื ื—ื ื• ื™ื•ื“ืขื™ื ืฉื”ื™ืœื“ื™ื ืžืชื•.
18:22
What matters is how likely or unlikely, relative to each other,
354
1102000
4000
ืžื” ืฉื—ืฉื•ื‘ ื–ื” ืžื” ื”ืกื‘ื™ืจื•ืช ืื• ืื™-ื”ืกื‘ื™ืจื•ืช ืฉืœ ืฉื ื™ ื”ื”ืกื‘ืจื™ื
18:26
the two explanations are.
355
1106000
2000
ื‘ื™ื—ืก ื–ื” ืœื–ื”.
18:28
And they're both implausible.
356
1108000
2000
ื•ืฉื ื™ื”ื ื‘ืœืชื™ ืกื‘ื™ืจื™ื.
18:31
There's a situation where errors in statistics had really profound
357
1111000
4000
ื–ื”ื• ืžืงืจื” ื‘ื• ื”ื”ืฉืœื›ื•ืช ืฉืœ ื˜ืขื•ื™ื•ืช ื‘ืกื˜ื˜ื™ืกื˜ื™ืงื”
18:35
and really unfortunate consequences.
358
1115000
3000
ื—ืžื•ืจื•ืช ื‘ื™ื•ืชืจ ื•ืžืฆืขืจื•ืช ื‘ื™ื•ืชืจ.
18:38
In fact, there are two other women who were convicted on the basis of the
359
1118000
2000
ืœืžืขืฉื”, ืฉืชื™ ื ืฉื™ื ื ื•ืกืคื•ืช ื”ื•ืจืฉืขื• ืขืœ ื‘ืกื™ืก
18:40
evidence of this pediatrician, who have subsequently been released on appeal.
360
1120000
4000
ื”ืขื“ื•ืช ืฉืœ ืจื•ืคื ื”ื™ืœื“ื™ื ื”ื–ื”, ื•ืฉื•ื—ืจืจื• ืœืื—ืจ ืขืจืขื•ืจ.
18:44
Many cases were reviewed.
361
1124000
2000
ืžืงืจื™ื ืจื‘ื™ื ื ื‘ื—ื ื• ืžื—ื“ืฉ.
18:46
And it's particularly topical because he's currently facing a disrepute charge
362
1126000
4000
ื•ื–ื” ืžืื•ื“ ืืงื˜ื•ืืœื™ ื›ื™ ืขื›ืฉื™ื• ื”ื•ื ืขื•ืžื“ ื‘ืคื ื™ ืชื‘ื™ืขืช ื”ื•ืฆืืช ืฉื ืจืข
18:50
at Britain's General Medical Council.
363
1130000
3000
ื‘ืžื•ืขืฆื” ื”ืจืคื•ืื™ืช ื”ื›ืœืœื™ืช ื‘ื‘ืจื™ื˜ื ื™ื”.
18:53
So just to conclude -- what are the take-home messages from this?
364
1133000
4000
ื•ื›ื“ื™ ืœืกื›ื - ืžื” ื”ืžืกืจ ืฉืชืงื—ื• ืืชื›ื ื”ื‘ื™ืชื”?
18:57
Well, we know that randomness and uncertainty and chance
365
1137000
4000
ืื ื—ื ื• ื™ื•ื“ืขื™ื ืฉืืงืจืื™ื•ืช, ื•ื—ื•ืกืจ ื•ื“ืื•ืช, ื•ืกื™ื›ื•ื™
19:01
are very much a part of our everyday life.
366
1141000
3000
ื”ื ื—ืœืง ื‘ืœืชื™ ื ืคืจื“ ืžื—ื™ื™ ื”ื™ื•ื ื™ื•ื ืฉืœื ื•.
19:04
It's also true -- and, although, you, as a collective, are very special in many ways,
367
1144000
5000
ื‘ื ื•ืกืฃ, ืœืžืจื•ืช ืฉืืชื, ื›ื›ืœืœ, ืžื™ื•ื—ื“ื™ื ืžืื•ื“ ื‘ื“ืจื›ื™ื ืจื‘ื•ืช,
19:09
you're completely typical in not getting the examples I gave right.
368
1149000
4000
ืืชื ืื•ืคื™ื™ื ื™ื™ื ืžืื•ื“ ืฉืœื ืคืชืจืชื ืืช ื”ื“ื•ื’ืžืื•ืช ืฉืœื™ ื‘ืฆื•ืจื” ื ื›ื•ื ื”.
19:13
It's very well documented that people get things wrong.
369
1153000
3000
ื™ืฉื ื• ืชืขื•ื“ ืจื—ื‘ ื”ื™ืงืฃ ืœื›ืš ืฉืื ืฉื™ื ื˜ื•ืขื™ื.
19:16
They make errors of logic in reasoning with uncertainty.
370
1156000
3000
ื”ื ืžื‘ืฆืขื™ื ืฉื’ื™ืื•ืช ืœื•ื’ื™ื•ืช ื‘ื–ืžืŸ ื”ืกืงืช ืžืกืงื ื•ืช ื‘ืชื ืื™ ื—ื•ืกืจ ื•ื“ืื•ืช.
19:20
We can cope with the subtleties of language brilliantly --
371
1160000
2000
ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื”ืชืžื•ื“ื“ ืขื ื”ื“ืงื•ื™ื•ืช ืฉืœ ื”ืฉืคื” ื‘ืื•ืคืŸ ืžื–ื”ื™ืจ --
19:22
and there are interesting evolutionary questions about how we got here.
372
1162000
3000
ื•ื™ืฉื ืŸ ืฉืืœื•ืช ืื‘ื•ืœื•ืฆื™ื•ื ื™ื•ืช ืžืขื ื™ื™ื ื•ืช ืขืœ ืื™ืš ื”ื’ืขื ื• ืœื›ืืŸ.
19:25
We are not good at reasoning with uncertainty.
373
1165000
3000
ืื ื—ื ื• ืœื ืžื•ืฆืœื—ื™ื ื‘ื”ืกืงืช ืžืกืงื ื•ืช ื‘ืชื ืื™ ื—ื•ืกืจ ื•ื“ืื•ืช.
19:28
That's an issue in our everyday lives.
374
1168000
2000
ื–ื”ื• ื ื•ืฉื ืฉืงื™ื™ื ื‘ื—ื™ื™ ื”ื™ื•ื ื™ื•ื ืฉืœื ื•.
19:30
As you've heard from many of the talks, statistics underpins an enormous amount
375
1170000
3000
ื›ืคื™ ืฉืฉืžืขืชื ื‘ื”ืจืฆืื•ืช ืจื‘ื•ืช, ื”ืกื˜ื˜ื™ืกื˜ื™ืงื” ืžื”ื•ื•ื” ื‘ืกื™ืก ืœื›ืžื•ืช ืขืฆื•ืžื”
19:33
of research in science -- in social science, in medicine
376
1173000
3000
ืฉืœ ืžื—ืงืจื™ื ืžื“ืขื™ื™ื -- ื‘ืžื“ืขื™ ื”ื—ื‘ืจื”, ื‘ืจืคื•ืื”,
19:36
and indeed, quite a lot of industry.
377
1176000
2000
ื•ืœืžืขืฉื”, ื‘ืชื—ื•ืžื™ื ืชืขืฉื™ื™ืชื™ื™ื ืจื‘ื™ื.
19:38
All of quality control, which has had a major impact on industrial processing,
378
1178000
4000
ื›ืœ ื‘ืงืจืช ื”ืื™ื›ื•ืช, ืฉื”ื™ื ื‘ืขืœืช ื”ืฉืคืขื” ืžื›ืจืขืช ืขืœ ืชื”ืœื™ื›ื™ื ืชืขืฉื™ื™ืชื™ื™ื,
19:42
is underpinned by statistics.
379
1182000
2000
ืžื‘ื•ืกืกืช ืขืœ ืกื˜ื˜ื™ืกื˜ื™ืงื”.
19:44
It's something we're bad at doing.
380
1184000
2000
ื•ื–ื” ืžืฉื”ื• ืฉืื ื—ื ื• ื’ืจื•ืขื™ื ื‘ื‘ื™ืฆื•ืข ืฉืœื•.
19:46
At the very least, we should recognize that, and we tend not to.
381
1186000
3000
ืœื›ืœ ื”ืคื—ื•ืช, ืขืœื™ื ื• ืœื”ื›ื™ืจ ื‘ื›ืš, ื•ืื ื—ื ื• ื ื•ื˜ื™ื ืœื ืœืขืฉื•ืช ื–ืืช.
19:49
To go back to the legal context, at the Sally Clark trial
382
1189000
4000
ื•ืื ื ื—ื–ื•ืจ ืœื”ืงืฉืจ ื”ืžืฉืคื˜ื™, ื‘ืžืฉืคื˜ ืฉืœ ืกืืœื™ ืงืœืืจืง
19:53
all of the lawyers just accepted what the expert said.
383
1193000
4000
ื›ืœ ืขื•ืจื›ื™ ื”ื“ื™ืŸ ืคืฉื•ื˜ ืงื™ื‘ืœื• ืืช ืžื” ืฉื”ืžื•ืžื—ื” ืืžืจ.
19:57
So if a pediatrician had come out and said to a jury,
384
1197000
2000
ื•ื›ืš, ืื ืจื•ืคื ื™ืœื“ื™ื ื”ื™ื” ืื•ืžืจ ืœื—ื‘ืจ ื”ืžื•ืฉื‘ืขื™ื,
19:59
"I know how to build bridges. I've built one down the road.
385
1199000
3000
"ืื ื™ ื™ื•ื“ืข ืœื‘ื ื•ืช ื’ืฉืจื™ื. ื‘ื ื™ืชื™ ืื—ื“ ื‘ืžื•ืจื“ ื”ื“ืจืš.
20:02
Please drive your car home over it,"
386
1202000
2000
ืื ื™ ืžื‘ืงืฉ ืฉืชืกืขื• ืขืœื™ื• ื‘ื“ืจื›ื›ื ื”ื‘ื™ืชื”,"
20:04
they would have said, "Well, pediatricians don't know how to build bridges.
387
1204000
2000
ื”ื ื”ื™ื• ืื•ืžืจื™ื, "ื•ื‘ื›ืŸ, ืจื•ืคืื™ ื™ืœื“ื™ื ืœื ื™ื•ื“ืขื™ื ืœื‘ื ื•ืช ื’ืฉืจื™ื.
20:06
That's what engineers do."
388
1206000
2000
ื–ื”ื• ืชืคืงื™ื“ื ืฉืœ ื”ืžื”ื ื“ืกื™ื."
20:08
On the other hand, he came out and effectively said, or implied,
389
1208000
3000
ื•ืžืฆื“ ืฉื ื™, ื”ืฉืชืžืข ืžื“ื‘ืจื™ื• ืฉื”ื•ื ืื•ืžืจ,
20:11
"I know how to reason with uncertainty. I know how to do statistics."
390
1211000
3000
"ืื ื™ ื™ื•ื“ืข ืื™ืš ืœื”ืกื™ืง ืžืกืงื ื•ืช ื‘ืชื ืื™ ื—ื•ืกืจ ื•ื“ืื•ืช. ืื ื™ ื™ื•ื“ืข ืื™ืš ืขื•ืฉื™ื ืกื˜ื˜ื™ืกื˜ื™ืงื”.
20:14
And everyone said, "Well, that's fine. He's an expert."
391
1214000
3000
ื•ื›ื•ืœื ืืžืจื•, "ื‘ืกื“ืจ ื’ืžื•ืจ. ื”ื•ื ืžื•ืžื—ื”."
20:17
So we need to understand where our competence is and isn't.
392
1217000
3000
ืื ื—ื ื• ื—ื™ื™ื‘ื™ื ืœื”ื‘ื™ืŸ ื‘ืื™ื–ื” ืชื—ื•ืžื™ื ื™ืฉ ืœื ื• ื™ื›ื•ืœืช ื•ื‘ืื™ื–ื” ืœื.
20:20
Exactly the same kinds of issues arose in the early days of DNA profiling,
393
1220000
4000
ืื•ืชื ื ื•ืฉืื™ื ื”ืชืขื•ืจืจื• ื‘ืชื—ื™ืœืช ื”ื“ืจืš ืฉืœ ืฉื™ืžื•ืฉ ื‘ืคืจื•ืคื™ืœื™ื ื’ื ื˜ื™ื™ื,
20:24
when scientists, and lawyers and in some cases judges,
394
1224000
4000
ื›ืืฉืจ ืžื“ืขื ื™ื ื•ืขื•ืจื›ื™ ื“ื™ืŸ ื•ื‘ืžืงืจื™ื ืžืกื•ื™ื™ืžื™ื ืฉื•ืคื˜ื™ื,
20:28
routinely misrepresented evidence.
395
1228000
3000
ื ื”ื’ื• ืœื”ืฆื™ื’ ืจืื™ื•ืช ื‘ืฆื•ืจื” ืžืกื•ืœืคืช.
20:32
Usually -- one hopes -- innocently, but misrepresented evidence.
396
1232000
3000
ื‘ื“ืจืš ื›ืœืœ, ื™ืฉ ืœืงื•ื•ืช, ื‘ืชื•ื ืœื‘, ืื‘ืœ ื”ืจืื™ื•ืช ื”ื•ืฆื’ื• ื‘ืฆื•ืจื” ืžืกื•ืœืคืช.
20:35
Forensic scientists said, "The chance that this guy's innocent is one in three million."
397
1235000
5000
ืžื“ืขื ื™ ื–ื™ื”ื•ื™ ืคืœื™ืœื™ ืืžืจื•, "ื”ืกื™ื›ื•ื™ ืฉื”ื‘ื—ื•ืจ ื”ื–ื” ื—ืฃ ืžืคืฉืข ื”ื•ื 1 ืœ-3 ืžื™ืœื™ื•ืŸ.
20:40
Even if you believe the number, just like the 73 million to one,
398
1240000
2000
ื•ื’ื ืื ืืชื ืžืืžื™ื ื™ื ืœืžืกืคืจ ื”ื–ื”, ื›ืžื• ืœ-73 ืžื™ืœื™ื•ืŸ ืœ-1,
20:42
that's not what it meant.
399
1242000
2000
ื–ืืช ืœื ื”ืžืฉืžืขื•ืช ืฉืœื•.
20:44
And there have been celebrated appeal cases
400
1244000
2000
ื•ื”ื™ื• ืขืจืขื•ืจื™ื ืžืคื•ืจืกืžื™ื
20:46
in Britain and elsewhere because of that.
401
1246000
2000
ื‘ื‘ืจื™ื˜ื ื™ื” ื•ื‘ืžืงื•ืžื•ืช ืื—ืจื™ื ืžื”ืกื™ื‘ื” ื”ื–ืืช.
20:48
And just to finish in the context of the legal system.
402
1248000
3000
ื•ื›ื“ื™ ืœืกื™ื™ื ื‘ื”ืงืฉืจ ืฉืœ ื”ืžืขืจื›ืช ื”ืžืฉืคื˜ื™ืช.
20:51
It's all very well to say, "Let's do our best to present the evidence."
403
1251000
4000
ื–ื” ื™ืคื” ืžืื•ื“ ืœื”ื’ื™ื“, "ื ืขืฉื” ืืช ืžื™ื˜ื‘ ื™ื›ื•ืœืชื ื• ื‘ื”ืฆื’ืช ื”ืจืื™ื•ืช."
20:55
But more and more, in cases of DNA profiling -- this is another one --
404
1255000
3000
ืื‘ืœ ื™ื•ืชืจ ื•ื™ื•ืชืจ, ื‘ืžืงืจื™ื ืฉืœ ื‘ื ื™ื™ืช ืคืจื•ืคื™ืœื™ื ื’ื ื˜ื™ื™ื -- ื–ื” ื“ื‘ืจ ื ื•ืกืฃ --
20:58
we expect juries, who are ordinary people --
405
1258000
3000
ืื ื—ื ื• ืžืฆืคื™ื ืžื—ื‘ืจ ื”ืžื•ืฉื‘ืขื™ื, ืฉื”ื ืื ืฉื™ื ืจื’ื™ืœื™ื --
21:01
and it's documented they're very bad at this --
406
1261000
2000
ื•ื–ื” ืžืชื•ืขื“ ืฉื”ื ืžืื•ื“ ื’ืจื•ืขื™ื ื‘ื›ืš --
21:03
we expect juries to be able to cope with the sorts of reasoning that goes on.
407
1263000
4000
ืื ื—ื ื• ืžืฆืคื™ื ืฉื—ื‘ืจ ื”ืžื•ืฉื‘ืขื™ื ื™ื”ื™ื” ืžืกื•ื’ืœ ืœื”ืชืžื•ื“ื“ ืขื ืชื”ืœื™ืš ื”ืกืงืช ื”ืžืกืงื ื•ืช ืฉื›ืจื•ืš ื‘ื›ืš.
21:07
In other spheres of life, if people argued -- well, except possibly for politics --
408
1267000
5000
ื‘ืชื—ื•ืžื™ ื—ื™ื™ื ืื—ืจื™ื, ืื ืื ืฉื™ื ื”ื™ื• ื˜ื•ืขื ื™ื -- ื—ื•ืฅ ืื•ืœื™ ืžื‘ืคื•ืœื™ื˜ื™ืงื”,
21:12
but in other spheres of life, if people argued illogically,
409
1272000
2000
ืื‘ืœ ื‘ืชื—ื•ืžื™ ื—ื™ื™ื ืื—ืจื™ื, ืื ืื ืฉื™ื ื”ื™ื• ื˜ื•ืขื ื™ื ื‘ื—ื•ืกืจ ื”ื’ื™ื•ืŸ,
21:14
we'd say that's not a good thing.
410
1274000
2000
ื”ื™ื™ื ื• ืื•ืžืจื™ื ืฉื–ื” ื“ื‘ืจ ื’ืจื•ืข.
21:16
We sort of expect it of politicians and don't hope for much more.
411
1276000
4000
ืื•ืœื™ ืื ื—ื ื• ืžืฆืคื™ื ืœื›ืš ืžืคื•ืœื™ื˜ื™ืงืื™ื - ื”ืฆื™ืคื™ื•ืช ืฉืœื ื• ืžื”ื ืœื ื’ื‘ื•ื”ื•ืช.
21:20
In the case of uncertainty, we get it wrong all the time --
412
1280000
3000
ื‘ืžืงืจื™ื ืฉืœ ื—ื•ืกืจ ื•ื“ืื•ืช, ืื ื—ื ื• ื˜ื•ืขื™ื ื›ืœ ื”ื–ืžืŸ --
21:23
and at the very least, we should be aware of that,
413
1283000
2000
ื•ืœื›ืœ ื”ืคื—ื•ืช, ืื ื—ื ื• ืฆืจื™ื›ื™ื ืœื”ื™ื•ืช ืžื•ื“ืขื™ื ืœื›ืš.
21:25
and ideally, we might try and do something about it.
414
1285000
2000
ื•ื‘ืื•ืคืŸ ืื™ื“ื™ืืœื™, ื’ื ืœื ืกื•ืช ืœืขืฉื•ืช ืžืฉื”ื• ื‘ืงืฉืจ ืœื›ืš.
21:27
Thanks very much.
415
1287000
1000
ืชื•ื“ื” ืจื‘ื”.
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7