The Math Behind Basketball's Wildest Moves | Rajiv Maheswaran | TED Talks

1,109,023 views ใƒป 2015-07-06

TED


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

ืžืชืจื’ื: Yifat Adler ืžื‘ืงืจ: Shir Ben Asher Kestin
00:12
My colleagues and I are fascinated by the science of moving dots.
0
12954
3583
ืžื“ืข ื”ื ืงื•ื“ื•ืช ื”ื ืขื•ืช ืžืจืชืง ืื•ืชื™ ื•ืืช ืขืžื™ืชื™ื™.
00:16
So what are these dots?
1
16927
1150
ืื– ืžื”ืŸ ื”ื ืงื•ื“ื•ืช ื”ืืœื•?
00:18
Well, it's all of us.
2
18101
1287
ื”ื ืงื•ื“ื•ืช ื”ืŸ ืื ื—ื ื•, ื›ื•ืœื ื•.
00:19
And we're moving in our homes, in our offices, as we shop and travel
3
19412
5085
ืื ื—ื ื• ื ืขื™ื ื‘ื‘ื™ืช, ื‘ืžืฉืจื“, ื‘ืงื ื™ื•ืช ื•ื‘ื—ื•ืคืฉื”,
00:24
throughout our cities and around the world.
4
24521
2066
ื‘ืจื—ื‘ื™ ื”ืขืจื™ื ื•ืžืกื‘ื™ื‘ ืœืขื•ืœื.
00:26
And wouldn't it be great if we could understand all this movement?
5
26958
3669
ื”ืื ืœื ื”ื™ื” ื ืคืœื ืื™ืœื• ื™ื›ื•ืœื ื• ืœื”ื‘ื™ืŸ ืืช ื”ืชื ื•ืขื” ื”ื–ืืช,
00:30
If we could find patterns and meaning and insight in it.
6
30918
2890
ืื™ืœื• ื™ื›ื•ืœื ื• ืœืžืฆื•ื ื‘ื” ืชื‘ื ื™ื•ืช, ืžืฉืžืขื•ื™ื•ืช ื•ืชื•ื‘ื ื•ืช?
00:34
And luckily for us, we live in a time
7
34259
1785
ื”ืชืžื–ืœ ืžื–ืœื ื• ื•ืื ื—ื ื• ื—ื™ื™ื ื‘ืขื™ื“ืŸ ืขืชื™ืจ ืืคืฉืจื•ื™ื•ืช ืชื™ืขื•ื“ ืขืฆืžื™.
00:36
where we're incredibly good at capturing information about ourselves.
8
36068
4497
00:40
So whether it's through sensors or videos, or apps,
9
40807
3663
ื—ื™ื™ืฉื ื™ื, ืกืจื˜ื•ื ื™ื ื•ืืคืœื™ืงืฆื™ื•ืช ืžืืคืฉืจื™ื ืœื ื•
00:44
we can track our movement with incredibly fine detail.
10
44494
2809
ืœืขืงื•ื‘ ืื—ืจื™ ื”ืชื ื•ืขื” ืฉืœื ื• ื•ืœืงื‘ืœ ืขืœื™ื” ื ืชื•ื ื™ื ืžืคื•ืจื˜ื™ื ื‘ื™ื•ืชืจ.
00:48
So it turns out one of the places where we have the best data about movement
11
48092
5032
ื•ืžืชื‘ืจืจ ืฉืื—ื“ ื”ืชื—ื•ืžื™ื ืฉื‘ื”ื ื”ื ืชื•ื ื™ื ืฉื™ืฉ ืœื ื• ืขืœ ื”ืชื ื•ืขื” ื˜ื•ื‘ื™ื ื‘ืžื™ื•ื—ื“
00:53
is sports.
12
53148
1208
ื”ื•ื ื”ืกืคื•ืจื˜.
00:54
So whether it's basketball or baseball, or football or the other football,
13
54682
5333
ื‘ื›ื“ื•ืจืกืœ, ื‘ื‘ื™ื™ืกื‘ื•ืœ, ื‘ื›ื“ื•ืจื’ืœ ืื• ื‘ืคื•ื˜ื‘ื•ืœ,
01:00
we're instrumenting our stadiums and our players to track their movements
14
60039
4402
ืื ื—ื ื• ืžืชืงื™ื ื™ื ืฆื™ื•ื“ ื‘ืืฆื˜ื“ื™ื•ื ื™ื ืฉืœื ื• ื•ืขืœ ื”ืฉื—ืงื ื™ื
ื›ื“ื™ ืœืชืขื“ ืืช ื”ืชื ื•ืขื•ืช ืฉืœื”ื ื‘ื›ืœ ืฉื‘ืจื™ืจ ืฉื ื™ื™ื”.
01:04
every fraction of a second.
15
64465
1313
01:05
So what we're doing is turning our athletes into --
16
65802
4382
ืœืžืขืฉื” ืื ื—ื ื• ื”ื•ืคื›ื™ื ืืช ื”ืืชืœื˜ื™ื ืฉืœื ื• ืœ...
01:10
you probably guessed it --
17
70208
1959
ืืชื ื‘ื•ื•ื“ืื™ ื›ื‘ืจ ืžื ื—ืฉื™ื...
01:12
moving dots.
18
72191
1396
ื ืงื•ื“ื•ืช ื ืขื•ืช.
01:13
So we've got mountains of moving dots and like most raw data,
19
73946
4934
ืื– ื™ืฉ ืœื ื• ื”ืจืจื™ ื ืชื•ื ื™ื ืขืœ ื ืงื•ื“ื•ืช ื ืขื•ืช.
ื•ื›ืžื• ืจื•ื‘ ื”ืžื™ื“ืข ื”ื’ื•ืœืžื™, ืงืฉื” ืœืงืจื•ื ืื•ืชื ื•ื”ื ืœื ืžืื•ื“ ืžืขื ื™ื™ื ื™ื.
01:18
it's hard to deal with and not that interesting.
20
78904
2502
01:21
But there are things that, for example, basketball coaches want to know.
21
81430
3769
ืื‘ืœ ื™ืฉ ื“ื‘ืจื™ื ืฉืžืืžื ื™ ื›ื“ื•ืจืกืœ, ืœื“ื•ื’ืžื”, ื”ื™ื• ืจื•ืฆื™ื ืœื“ืขืช.
01:25
And the problem is they can't know them because they'd have to watch every second
22
85223
3810
ืื‘ืœ ื”ื ืœื ื™ื›ื•ืœื™ื ืœื“ืขืช ืื•ืชื,
ื›ื™ ื”ื ื™ืฆื˜ืจื›ื• ืœืฆืคื•ืช ื‘ื›ืœ ืฉื ื™ื™ื” ืฉืœ ื›ืœ ืžืฉื—ืง, ืœื–ื›ื•ืจ ื•ืœื ืชื— ืื•ืชืŸ.
01:29
of every game, remember it and process it.
23
89057
2589
01:31
And a person can't do that,
24
91804
1930
ื‘ื ื™ ืื“ื ืœื ืžืกื•ื’ืœื™ื ืœืขืฉื•ืช ืืช ื–ื”,
01:33
but a machine can.
25
93758
1310
ืื‘ืœ ืžื›ื•ื ื•ืช ื™ื›ื•ืœื•ืช.
01:35
The problem is a machine can't see the game with the eye of a coach.
26
95661
3410
ื”ื‘ืขื™ื” ื”ื™ื ืฉืžื›ื•ื ื•ืช ืœื ื™ื›ื•ืœื•ืช ืœืจืื•ืช ืืช ื”ืžืฉื—ืง ื›ืžื• ืฉืžืืžืŸ ืจื•ืื” ืื•ืชื•.
01:39
At least they couldn't until now.
27
99363
2261
ืœืคื—ื•ืช ื”ืŸ ืœื ื™ื›ืœื• ืœื‘ืฆืข ืืช ื–ื” ืขื“ ื›ื”.
01:42
So what have we taught the machine to see?
28
102228
2103
ืื– ืžื” ืœื™ืžื“ื ื• ืืช ื”ืžื›ื•ื ื” ืœืจืื•ืช?
01:45
So, we started simply.
29
105569
1787
ื”ืชื—ืœื ื• ื‘ืžื”ืœื›ื™ื ืคืฉื•ื˜ื™ื.
01:47
We taught it things like passes, shots and rebounds.
30
107380
3799
ืœื™ืžื“ื ื• ืืช ื”ืžื›ื•ื ื” ืžื”ื ืžืกื™ืจื•ืช, ื–ืจื™ืงื•ืช ื•ื›ื“ื•ืจื™ื ื—ื•ื–ืจื™ื.
01:51
Things that most casual fans would know.
31
111203
2541
ืžื”ืœื›ื™ื ืคืฉื•ื˜ื™ื ืฉื›ืœ ื—ื•ื‘ื‘ ื›ื“ื•ืจืกืœ ืžื›ื™ืจ.
01:53
And then we moved on to things slightly more complicated.
32
113768
2832
ืžืฉื ืขื‘ืจื ื• ืœืžื”ืœื›ื™ื ืงืฆืช ื™ื•ืชืจ ืžื•ืจื›ื‘ื™ื.
01:56
Events like post-ups, and pick-and-rolls, and isolations.
33
116624
4588
ืžื”ืœื›ื™ื ื›ืžื• ืคื•ืกื˜-ืืค, ืคื™ืง-ืื ื“-ืจื•ืœ ื•ื‘ื™ื“ื•ื“.
02:01
And if you don't know them, that's okay. Most casual players probably do.
34
121377
3543
ื”ืžื”ืœื›ื™ื ื”ืืœื” ืžื•ื›ืจื™ื ืœืจื•ื‘ ื”ืฉื—ืงื ื™ื ื”ื—ื•ื‘ื‘ื ื™ื ื•ื’ื ืื ืื™ื ื›ื ืžื›ื™ืจื™ื ืื•ืชื - ืœื ื ื•ืจื.
02:05
Now, we've gotten to a point where today, the machine understands complex events
35
125560
5340
ื”ื™ื•ื ื”ืžื›ื•ื ื” ืžื‘ื™ื ื” ื›ื‘ืจ ืžื”ืœื›ื™ื ืžื•ืจื›ื‘ื™ื
02:10
like down screens and wide pins.
36
130924
3073
ื›ืžื• Down screens ื•-Wide pins.
02:14
Basically things only professionals know.
37
134021
2726
ื“ื‘ืจื™ื ืฉืจืง ืžืงืฆื•ืขื ื™ื ืžื›ื™ืจื™ื.
02:16
So we have taught a machine to see with the eyes of a coach.
38
136771
4388
ืœื™ืžื“ื ื• ืืช ื”ืžื›ื•ื ื” ืœืจืื•ืช ืžื ืงื•ื“ืช ื”ืžื‘ื˜ ืฉืœ ื”ืžืืžืŸ.
02:22
So how have we been able to do this?
39
142009
1857
ืื™ืš ืขืฉื™ื ื• ืืช ื–ื”?
02:24
If I asked a coach to describe something like a pick-and-roll,
40
144511
3118
ืื ื”ื™ื™ืชื™ ืžื‘ืงืฉ ืžืžืืžืŸ ืœืชืืจ ืžื”ืœืš ื›ืžื• ืคื™ืง-ืื ื“-ืจื•ืœ,
02:27
they would give me a description,
41
147653
1640
ื•ื”ื™ื™ืชื™ ืžืงื•ื“ื“ ืืช ื”ืชื™ืื•ืจ ืฉืœื• ื›ืืœื’ื•ืจื™ืชื,
02:29
and if I encoded that as an algorithm, it would be terrible.
42
149317
2856
ื–ื” ื”ื™ื” ืื™ื•ื ื•ื ื•ืจื.
02:33
The pick-and-roll happens to be this dance in basketball between four players,
43
153026
4278
ืคื™ืง-ืื ื“-ืจื•ืœ ื”ื•ื ืจื™ืงื•ื“ ืฉืœ ืืจื‘ืขื” ืฉื—ืงื ื™ื ื‘ื›ื“ื•ืจืกืœ,
02:37
two on offense and two on defense.
44
157328
1912
ืฉื ื™ ืชื•ืงืคื™ื ื•ืฉื ื™ ืžื’ื™ื ื™ื.
02:39
And here's kind of how it goes.
45
159486
1618
ื”ื•ื ืžืชื ื”ืœ ื‘ืขืจืš ื›ื›ื”:
02:41
So there's the guy on offense without the ball
46
161128
2533
ืฉื—ืงืŸ ืชื•ืงืฃ ืฉื”ื›ื“ื•ืจ ืœื ื‘ื™ื“ื™ื•
02:43
the ball and he goes next to the guy guarding the guy with the ball,
47
163685
3209
ืžืชืงืจื‘ ืœืฉื—ืงืŸ ื”ื”ื’ื ื” ืฉืฉื•ืžืจ ืขืœ ื”ืฉื—ืงืŸ ืขื ื”ื›ื“ื•ืจ,
02:46
and he kind of stays there
48
166918
1257
ื•ืื– ื”ื•ื ื ืฉืืจ ื‘ืกื‘ื™ื‘ื”,
02:48
and they both move and stuff happens, and ta-da, it's a pick-and-roll.
49
168199
3317
ื•ืฉื ื™ื”ื ื ืขื™ื ื•ืงื•ืจื™ื ื›ืœ ืžื™ื ื™ ื“ื‘ืจื™ื ื•ื˜ื”-ื“ื”! ืงื™ื‘ืœื ื• ืคื™ืง-ืื ื“-ืจื•ืœ.
02:51
(Laughter)
50
171540
2215
(ืฆื—ื•ืง)
02:53
So that is also an example of a terrible algorithm.
51
173779
2508
ื–ืืช ื’ื ื“ื•ื’ืžื” ืœืืœื’ื•ืจื™ืชื ื’ืจื•ืข.
02:56
So, if the player who's the interferer -- he's called the screener --
52
176913
4204
ืื ื”ืฉื—ืงืŸ ืฉืžืคืจื™ืข, ืฉื ืงืจื ื”ืฉื—ืงืŸ ื”ื—ื•ืกื,
03:01
goes close by, but he doesn't stop,
53
181278
2872
ืžืชืงืจื‘ ืœืฉื—ืงืŸ ืขื ื”ื›ื“ื•ืจ ืื‘ืœ ืœื ืขื•ืฆืจ,
03:04
it's probably not a pick-and-roll.
54
184174
1765
ื–ื” ื›ื ืจืื” ืœื ืคื™ืง-ืื ื“-ืจื•ืœ.
03:06
Or if he does stop, but he doesn't stop close enough,
55
186560
3945
ืื ื”ื•ื ืขื•ืฆืจ, ืื‘ืœ ืœื ืขื•ืฆืจ ืžืกืคื™ืง ืงืจื•ื‘,
03:10
it's probably not a pick-and-roll.
56
190529
1761
ื’ื ื–ื” ื›ื ืจืื” ืœื ืคื™ืง-ืื ื“-ืจื•ืœ.
03:12
Or, if he does go close by and he does stop
57
192642
3237
ืื ื”ื•ื ืขื•ืฆืจ ืžืกืคื™ืง ืงืจื•ื‘, ืื‘ืœ ืฉื ื™ ื”ืฉื—ืงื ื™ื ืžืชื—ืช ืœืกืœ,
03:15
but they do it under the basket, it's probably not a pick-and-roll.
58
195903
3324
ื’ื ื–ื” ื›ื ืจืื” ืœื ืคื™ืง-ืื ื“-ืจื•ืœ.
03:19
Or I could be wrong, they could all be pick-and-rolls.
59
199462
2524
ื•ืื•ืœื™ ืื ื™ ื‘ื›ืœืœ ื˜ื•ืขื”, ื•ื›ืœ ื”ืžื”ืœื›ื™ื ื”ืืœื” ื™ื›ื•ืœื™ื ืœื”ื™ื•ืช ืคื™ืง-ืื ื“-ืจื•ืœ.
03:22
It really depends on the exact timing, the distances, the locations,
60
202010
4568
ื–ื” ืชืœื•ื™ ื‘ืชื–ืžื•ืŸ ื”ืžื“ื•ื™ืง, ื‘ืžืจื—ืงื™ื, ื‘ืžื™ืงื•ืžื™ื,
03:26
and that's what makes it hard.
61
206602
1495
ื•ื–ื” ืžื” ืฉืžืกื‘ืš ืืช ื”ืขื ื™ื™ืŸ.
03:28
So, luckily, with machine learning, we can go beyond our own ability
62
208579
4944
ืœืžืจื‘ื” ื”ืžื–ืœ, ืœืžื™ื“ื” ื—ื™ืฉื•ื‘ื™ืช ืžืืคืฉืจืช ืœื ื• ืœืคืจื•ืฅ ืืช ื’ื‘ื•ืœื•ืช ื”ื™ื›ื•ืœืช ืฉืœื ื•
03:33
to describe the things we know.
63
213547
1743
ืœืชืืจ ืืช ื”ื“ื‘ืจื™ื ืฉืื ื—ื ื• ื™ื•ื“ืขื™ื.
03:35
So how does this work? Well, it's by example.
64
215314
2280
ืื– ืื™ืš ื–ื” ืขื•ื‘ื“? ื‘ืขื–ืจืช ื“ื•ื’ืžืื•ืช.
03:37
So we go to the machine and say, "Good morning, machine.
65
217759
2830
ืื ื—ื ื• ื ื™ื’ืฉื™ื ืœืžื›ื•ื ื” ื•ืื•ืžืจื™ื ืœื”, "ื‘ื•ืงืจ ื˜ื•ื‘, ืžื›ื•ื ื”.
03:41
Here are some pick-and-rolls, and here are some things that are not.
66
221077
3359
ื”ื ื” ื›ืžื” ืžื”ืœื›ื™ื ืฉืœ ืคื™ืง-ืื ื“-ืจื•ืœ, ื•ื”ื ื” ื›ืžื” ืžื”ืœื›ื™ื ืฉื”ื ืœื ืคื™ืง-ืื ื“-ืจื•ืœ.
03:44
Please find a way to tell the difference."
67
224720
2252
ืชืžืฆืื™ ื‘ื‘ืงืฉื” ื“ืจืš ืœื–ื”ื•ืช ืืช ื”ื”ื‘ื“ืœื™ื ื‘ื™ื ื™ื”ื".
03:47
And the key to all of this is to find features that enable it to separate.
68
227076
3707
ื”ืžืคืชื— ืœื›ืš ื”ื•ื ืœืžืฆื•ื ืžืืคื™ื™ื ื™ื ืฉืžืืคืฉืจื™ื ืœื‘ืฆืข ืืช ื”ื”ื‘ื—ื ื” ื”ื–ืืช.
03:50
So if I was going to teach it the difference
69
230807
2109
ืœื“ื•ื’ืžื”, ืื™ืœื• ืจืฆื™ืชื™ ืœืœืžื“ ืืช ื”ืžื›ื•ื ื” ืœื”ื‘ื“ื™ืœ ื‘ื™ืŸ ืชืคื•ื–ื™ื ืœืชืคื•ื—ื™ื,
03:52
between an apple and orange,
70
232940
1381
ื”ื™ื™ืชื™ ืžืฆื™ืข ืœื” "ืื•ืœื™ ืชื‘ื“ืงื™ ืฆื‘ืข ืื• ืฆื•ืจื”?"
03:54
I might say, "Why don't you use color or shape?"
71
234345
2375
03:56
And the problem that we're solving is, what are those things?
72
236744
2943
ื”ื‘ืขื™ื” ืฉืื ื—ื ื• ืžื ืกื™ื ืœืคืชื•ืจ ื”ื™ื, ืžื”ื ื”ืžืืคื™ื™ื ื™ื ื”ืืœื”?
03:59
What are the key features
73
239711
1247
ืžื”ื ืžืืคื™ื™ื ื™ ื”ืžืคืชื— ืฉืžืืคืฉืจื™ื ืœืžื—ืฉื‘ ืœื ื•ื•ื˜ ื‘ืขื•ืœื ื”ื ืงื•ื“ื•ืช ื”ื ืขื•ืช?
04:00
that let a computer navigate the world of moving dots?
74
240982
3499
04:04
So figuring out all these relationships with relative and absolute location,
75
244505
4823
ื”ื‘ื ืช ื”ื™ื—ืกื™ื ื‘ื™ืŸ ืžื™ืงื•ื, ืžืจื—ืง, ืชื–ืžื•ืŸ ื•ืžื”ื™ืจื•ื™ื•ืช ื™ื—ืกื™ื™ื ื•ืžื•ื—ืœื˜ื™ื,
04:09
distance, timing, velocities --
76
249352
1909
04:11
that's really the key to the science of moving dots, or as we like to call it,
77
251440
4928
ื”ื™ื ื”ืžืคืชื— ืœืžื“ืข ื”ื ืงื•ื“ื•ืช ื”ื ืขื•ืช,
ืื• ื›ืคื™ ืฉืื ื—ื ื• ืžื›ื ื™ื ื–ืืช ื‘ืฉืคื” ื”ืืงื“ืžื™ืช, ื–ื™ื”ื•ื™ ืชื‘ื ื™ื•ืช ืžืจื—ื‘-ื–ืžืŸ.
04:16
spatiotemporal pattern recognition, in academic vernacular.
78
256392
3344
04:19
Because the first thing is, you have to make it sound hard --
79
259925
2898
ื–ื” ื ืฉืžืข ืžืกื•ื‘ืš,
04:22
because it is.
80
262847
1278
ื›ื™ ื–ื” ื‘ืืžืช ืžืกื•ื‘ืš.
04:24
The key thing is, for NBA coaches, it's not that they want to know
81
264410
3141
ื—ืฉื•ื‘ ืœื”ื‘ื™ืŸ ืฉืœืžืืžื ื™ ืืŸ-ื‘ื™-ืื™ื™ ืœื ื‘ืืžืช ืื›ืคืช
04:27
whether a pick-and-roll happened or not.
82
267575
1922
ืื ื”ื™ื” ืื• ืœื ื”ื™ื” ืคื™ืง-ืื ื“-ืจื•ืœ.
04:29
It's that they want to know how it happened.
83
269521
2076
ืืœื ื”ื ืจื•ืฆื™ื ืœื”ื‘ื™ืŸ ืื™ืš ื”ืžื”ืœืš ื”ืชืจื—ืฉ.
04:31
And why is it so important to them? So here's a little insight.
84
271621
2986
ื•ืœืžื” ื–ื” ื›ืœ ื›ืš ื—ืฉื•ื‘ ืœื”ื?
04:34
It turns out in modern basketball,
85
274631
1771
ืžืชื‘ืจืจ ืฉื‘ื›ื“ื•ืจืกืœ ืžื•ื“ืจื ื™,
04:36
this pick-and-roll is perhaps the most important play.
86
276426
2539
ื”ืคื™ืง-ืื ื“-ืจื•ืœ ื ื—ืฉื‘ ืœืžื”ืœืš ื”ื—ืฉื•ื‘ ื‘ื™ื•ืชืจ ื‘ืžืฉื—ืง.
04:39
And knowing how to run it, and knowing how to defend it,
87
279065
2620
ื•ื”ื™ื“ืข ืื™ืš ืœื ื”ืœ ืื•ืชื•, ื•ื”ื™ื“ืข ืื™ืš ืœื”ืชื’ื•ื ืŸ ืžืคื ื™ื•,
04:41
is basically a key to winning and losing most games.
88
281709
2670
ื”ื ืœืžืขืฉื” ื”ืžืคืชื— ืœื ื™ืฆื—ื•ืŸ ืื• ืœื”ืคืกื“ ื‘ืจื•ื‘ ื”ืžืฉื—ืงื™ื.
04:44
So it turns out that this dance has a great many variations
89
284403
3801
ืžืชื‘ืจืจ ืฉืœืจื™ืงื•ื“ ื”ื–ื” ื™ืฉ ื”ืจื‘ื” ืžืื•ื“ ื•ืจื™ืืฆื™ื•ืช,
04:48
and identifying the variations is really the thing that matters,
90
288228
3648
ื•ื”ื–ื™ื”ื•ื™ ืฉืœื”ืŸ ื”ื•ื ืฉื•ืจืฉ ื”ืขื ื™ื™ืŸ,
04:51
and that's why we need this to be really, really good.
91
291900
2529
ื•ืœื›ืŸ ืื ื—ื ื• ืฆืจื™ื›ื™ื ืœืžืฆื•ื ื“ืจืš ืœื‘ืฆืข ืืช ื”ื–ื™ื”ื•ื™ ื‘ืฆื•ืจื” ืžื™ื˜ื‘ื™ืช.
04:55
So, here's an example.
92
295228
1176
ื”ื ื” ื“ื•ื’ืžื”.
ื™ืฉ ืฉื ื™ ืชื•ืงืคื™ื ื•ืฉื ื™ ืžื’ื™ื ื™ื
04:56
There are two offensive and two defensive players,
93
296428
2379
ืฉืžืชื›ื•ื ื ื™ื ืœื‘ืฆืข ืืช ืจื™ืงื•ื“ ื”ืคื™ืง-ืื ื“-ืจื•ืœ.
04:58
getting ready to do the pick-and-roll dance.
94
298831
2152
ื”ืฉื—ืงืŸ ืขื ื”ื›ื“ื•ืจ ื™ื›ื•ืœ ืœื‘ื—ื•ืจ ื‘ื™ืŸ Take ืœื‘ื™ืŸ Reject.
05:01
So the guy with ball can either take, or he can reject.
95
301007
2683
05:04
His teammate can either roll or pop.
96
304086
3001
ื”ืฉื•ืชืฃ ืฉืœื• ื™ื›ื•ืœ ืœื‘ื—ื•ืจ ื‘ื™ืŸ Roll ืœื‘ื™ืŸ Pop.
ื”ืฉื—ืงืŸ ืฉืฉื•ืžืจ ืขืœ ื”ืฉื—ืงืŸ ืขื ื”ื›ื“ื•ืจ ื™ื›ื•ืœ ืœื‘ื—ื•ืจ ื‘ื™ืŸ Over ืœ-Under.
05:07
The guy guarding the ball can either go over or under.
97
307111
2986
05:10
His teammate can either show or play up to touch, or play soft
98
310121
4565
ื”ืฉื•ืชืฃ ืฉืœื• ื™ื›ื•ืœ ืœื‘ื—ื•ืจ ื‘ื™ืŸ Show ืœ-Up to Touch ืื• ืœ-Soft.
05:14
and together they can either switch or blitz
99
314710
2618
ื•ื‘ื™ื—ื“ ื”ื ื™ื›ื•ืœื™ื ืœื‘ื—ื•ืจ ื‘ื™ืŸ Switch ืœ-Blitz.
05:17
and I didn't know most of these things when I started
100
317352
2659
ื‘ืชื—ื™ืœืช ื”ื“ืจืš ืœื ื”ื›ืจืชื™ ืืช ืจื•ื‘ ื”ืžื•ื ื—ื™ื ื”ืืœื”.
05:20
and it would be lovely if everybody moved according to those arrows.
101
320035
3920
ืื ื”ืฉื—ืงื ื™ื ื”ื™ื• ืชืžื™ื“ ื ืขื™ื ืœืคื™ ื”ื—ืฆื™ื, ื”ื—ื™ื™ื ืฉืœื ื• ื”ื™ื• ื”ืจื‘ื” ื™ื•ืชืจ ืงืœื™ื.
05:23
It would make our lives a lot easier, but it turns out movement is very messy.
102
323979
3905
ืื‘ืœ ืžืชื‘ืจืจ ืฉืชื ื•ืขื” ื”ื™ื ืขืกืง ืžืื•ื“ ืžื‘ื•ืœื’ืŸ.
ื”ืชื ื•ืขื” ืฉืœ ื”ืฉื—ืงื ื™ื ืžืคื•ืชืœืช ืžืื•ื“,
05:28
People wiggle a lot and getting these variations identified
103
328047
5484
ื•ืงืฉื” ืžืื•ื“ ืœื–ื”ื•ืช ืืช ื”ื•ื•ืจื™ืืฆื™ื•ืช ื”ืฉื•ื ื•ืช ื‘ืจืžืช ื ื›ื•ื ื•ืช ื’ื‘ื•ื”ื”,
05:33
with very high accuracy,
104
333555
1303
05:34
both in precision and recall, is tough
105
334882
1868
ื•ืœืฉืžื•ืจ ืขืœ ืจืžื•ืช ื’ื‘ื•ื”ื•ืช ืฉืœ ื“ื™ื•ืง ื•ืื—ื–ื•ืจ.
05:36
because that's what it takes to get a professional coach to believe in you.
106
336774
3618
ื›ื“ื™ ืฉืžืืžื ื™ื ืžืงืฆื•ืขื™ื™ื ื™ืืžื™ื ื• ื‘ืžืขืจื›ืช ืฉืœื ื• ื”ื™ื ื—ื™ื™ื‘ืช ืœืขืžื•ื“ ื‘ื“ืจื™ืฉื•ืช ื”ืืœื”.
05:40
And despite all the difficulties with the right spatiotemporal features
107
340416
3380
ื•ืœืžืจื•ืช ื”ืงืฉื™ื™ื ื”ืžืจื•ื‘ื™ื ื‘ืžืฆื™ืืช ืžืืคื™ื™ื ื™ ื”ืžืจื—ื‘-ื–ืžืŸ ื”ืžืชืื™ืžื™ื
05:43
we have been able to do that.
108
343820
1474
ื”ืฆืœื—ื ื• ืœืขืฉื•ืช ืืช ื–ื”.
05:45
Coaches trust our ability of our machine to identify these variations.
109
345318
3927
ืžืืžื ื™ื ื‘ื•ื˜ื—ื™ื ื‘ื™ื›ื•ืœืช ืฉืœ ื”ืžื›ื•ื ื” ืฉืœื ื• ืœื–ื”ื•ืช ืืช ื”ื•ื•ืจื™ืืฆื™ื•ืช ื”ืืœื”.
05:49
We're at the point where almost every single contender
110
349478
3533
ื›ืžืขื˜ ื›ืœ ื”ืงื‘ื•ืฆื•ืช ืฉืžืชืžื•ื“ื“ื•ืช ื”ืฉื ื” ืขืœ ืืœื™ืคื•ืช ื”ืืŸ-ื‘ื™-ืื™ื™
05:53
for an NBA championship this year
111
353035
1623
05:54
is using our software, which is built on a machine that understands
112
354682
4408
ืžืฉืชืžืฉื•ืช ื‘ืชื•ื›ื ื” ืฉืœื ื•,
ืฉืžื‘ื•ืกืกืช ืขืœ ืžื›ื•ื ื” ืฉืžื‘ื™ื ื” ืืช ื”ื ืงื•ื“ื•ืช ื”ื ืขื•ืช ื‘ืžืฉื—ืง ื”ื›ื“ื•ืจืกืœ.
05:59
the moving dots of basketball.
113
359114
1634
06:01
So not only that, we have given advice that has changed strategies
114
361872
5153
ืžืขื‘ืจ ืœื–ื”, ื”ื™ื™ืขื•ืฅ ืฉืœื ื• ืฉื™ื ื” ืืกื˜ืจื˜ื’ื™ื•ืช ืฉืœ ืงื‘ื•ืฆื•ืช
06:07
that have helped teams win very important games,
115
367049
3352
ื•ืขื–ืจ ืœื”ืŸ ืœื ืฆื— ื‘ืžืฉื—ืงื™ ืžืคืชื—.
06:10
and it's very exciting because you have coaches who've been in the league
116
370425
3732
ื–ื” ืžืื•ื“ ืžืจื’ืฉ ืื•ืชื ื•.
ืžืืžื ื™ื ืขื ื•ืชืง ืฉืœ 30 ืฉื ื•ืช ืื™ืžื•ืŸ ืžื•ื›ื ื™ื ืœืงื‘ืœ ืขืฆื•ืช ืžืžื›ื•ื ื”.
06:14
for 30 years that are willing to take advice from a machine.
117
374181
3067
06:17
And it's very exciting, it's much more than the pick-and-roll.
118
377874
2906
ื•ื–ื” ืžืื•ื“ ืžืจื’ืฉ. ืœื ืžื“ื•ื‘ืจ ืจืง ื‘ืคื™ืง-ืื ื“-ืจื•ืœ.
06:20
Our computer started out with simple things
119
380804
2076
ื”ืžื—ืฉื‘ ืฉืœื ื• ื”ืชื—ื™ืœ ื‘ื“ื‘ืจื™ื ืคืฉื•ื˜ื™ื,
06:22
and learned more and more complex things
120
382904
2064
ื•ืœืžื“ ื“ื‘ืจื™ื ื™ื•ืชืจ ื•ื™ื•ืชืจ ืžื•ืจื›ื‘ื™ื.
06:24
and now it knows so many things.
121
384992
1561
ื•ื”ื™ื•ื ื”ื•ื ื™ื•ื“ืข ื”ืจื‘ื” ืžืื•ื“ ื“ื‘ืจื™ื.
06:26
Frankly, I don't understand much of what it does,
122
386577
2835
ืœืžืขืŸ ื”ืืžืช, ืื ื™ ื›ื‘ืจ ืœื ืžื‘ื™ืŸ ื”ืจื‘ื” ืžื”ื“ื‘ืจื™ื ืฉื”ื•ื ืขื•ืฉื”.
06:29
and while it's not that special to be smarter than me,
123
389436
3715
ืืžื ื ืœื ืงืฉื” ืœื”ื™ื•ืช ื™ื•ืชืจ ื—ื›ื ืžืžื ื™,
06:33
we were wondering, can a machine know more than a coach?
124
393175
3644
ืื‘ืœ ืชื”ื™ื ื• ืื ืžื›ื•ื ื” ื™ื›ื•ืœื” ืœื“ืขืช ื™ื•ืชืจ ืžืžืืžื ื™ื?
06:36
Can it know more than person could know?
125
396843
2055
ื”ืื ื”ืžื›ื•ื ื” ื™ื›ื•ืœื” ืœื“ืขืช ื™ื•ืชืจ ืžืื“ื?
06:38
And it turns out the answer is yes.
126
398922
1745
ื•ื”ืชืฉื•ื‘ื” ื”ื™ื ื›ืŸ.
06:40
The coaches want players to take good shots.
127
400691
2557
ื”ืžืืžื ื™ื ืจื•ืฆื™ื ืฉื”ืฉื—ืงื ื™ื ื™ื–ืจืงื• ื–ืจื™ืงื•ืช ื˜ื•ื‘ื•ืช.
06:43
So if I'm standing near the basket
128
403272
1651
ืื ืื ื™ ืขื•ืžื“ ืœื™ื“ ื”ืกืœ ื•ืื™ืŸ ืืฃ ืื—ื“ ืื—ืจ ืœื™ื“ื™,
06:44
and there's nobody near me, it's a good shot.
129
404947
2166
06:47
If I'm standing far away surrounded by defenders, that's generally a bad shot.
130
407137
3940
ืื ืื ื™ ืจื—ื•ืง ืžื”ืกืœ ื•ืžื•ืงืฃ ื‘ืฉื—ืงื ื™ ื”ื’ื ื”, ื–ืืช ื‘ื“ืจืš ื›ืœืœ ื–ืจื™ืงื” ื’ืจื•ืขื”.
06:51
But we never knew how good "good" was, or how bad "bad" was quantitatively.
131
411101
4876
ืื‘ืœ ืœื ื™ื“ืขื ื• ืžื‘ื—ื™ื ืช ื›ืžื•ืชื™ืช ื›ืžื” "ื˜ื•ื‘" ื”ื•ื ื˜ื•ื‘ ืื• ื›ืžื” "ืจืข" ื”ื•ื ืจืข.
06:56
Until now.
132
416209
1150
ืขื“ ืขื›ืฉื™ื•.
06:57
So what we can do, again, using spatiotemporal features,
133
417771
3058
ื›ืขืช, ื‘ืขื–ืจืช ืžืืคื™ื™ื ื™ื ืฉืœ ืžืจื—ื‘-ื–ืžืŸ
07:00
we looked at every shot.
134
420853
1374
ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื‘ื—ื•ืŸ ื›ืœ ื–ืจื™ืงื”.
07:02
We can see: Where is the shot? What's the angle to the basket?
135
422251
3005
ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœืจืื•ืช: ืžืื™ืคื” ืžืชื‘ืฆืขืช ื”ื–ืจื™ืงื”? ืžื” ื”ื–ื•ื•ื™ืช ืœืกืœ?
07:05
Where are the defenders standing? What are their distances?
136
425280
2762
ืื™ืคื” ืขื•ืžื“ื™ื ื”ืžื’ื™ื ื™ื? ืžื” ื”ืžืจื—ืงื™ื ืžื”ื?
ืžื” ื”ื–ื•ื•ื™ื•ืช ืฉืœื”ื?
07:08
What are their angles?
137
428066
1331
ืื ื™ืฉ ื›ืžื” ืžื’ื™ื ื™ื,
07:09
For multiple defenders, we can look at how the player's moving
138
429421
2977
ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื‘ื“ื•ืง ืื™ืš ื”ืฉื—ืงืŸ ื ืข ื•ืœื—ื–ื•ืช ืืช ืกื•ื’ ื”ื–ืจื™ืงื” ืฉืœื•.
07:12
and predict the shot type.
139
432422
1433
07:13
We can look at all their velocities and we can build a model that predicts
140
433879
4074
ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื‘ื—ื•ืŸ ืืช ื”ืžื”ื™ืจื•ื™ื•ืช ืฉืœื”ื,
ื•ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื‘ื ื•ืช ืžื•ื“ืœ ืฉื—ื•ื–ื” ืžืจืืฉ
07:17
what is the likelihood that this shot would go in under these circumstances?
141
437977
4052
ืžื” ื”ืกื™ื›ื•ื™ ืฉื”ื–ืจื™ืงื” ื”ื–ืืช ืชื™ื›ื ืก ืœืกืœ ื‘ื ืกื™ื‘ื•ืช ื”ืืœื”.
07:22
So why is this important?
142
442188
1500
ืžื” ื”ื—ืฉื™ื‘ื•ืช ืฉืœ ื–ื”?
07:24
We can take something that was shooting,
143
444102
2803
ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœืงื—ืช ื–ืจื™ืงื”,
07:26
which was one thing before, and turn it into two things:
144
446929
2680
ื•ืœื—ืœืง ืื•ืชื” ืœืฉื ื™ ื—ืœืงื™ื:
07:29
the quality of the shot and the quality of the shooter.
145
449633
2651
ืื™ื›ื•ืช ื”ื–ืจื™ืงื” ื•ืื™ื›ื•ืช ื”ื–ื•ืจืง.
07:33
So here's a bubble chart, because what's TED without a bubble chart?
146
453680
3262
ื–ื” ืชืจืฉื™ื ื‘ื•ืขื•ืช, ื›ื™ ื‘-TED ืื™ ืืคืฉืจ ื‘ืœื™ ืชืจืฉื™ื ื‘ื•ืขื•ืช...
07:36
(Laughter)
147
456966
1014
(ืฆื—ื•ืง)
07:38
Those are NBA players.
148
458004
1311
ืืœื” ืฉื—ืงื ื™ ืืŸ-ื‘ื™-ืื™ื™.
07:39
The size is the size of the player and the color is the position.
149
459339
3120
ื’ื•ื“ืœ ื”ื‘ื•ืขื” ื”ื•ื ื’ื•ื‘ื” ื”ืฉื—ืงืŸ ื•ื”ืฆื‘ืข ืฉืœื” ื”ื•ื ื”ืžื™ืงื•ื ืฉืœื•.
07:42
On the x-axis, we have the shot probability.
150
462483
2132
ืฆื™ืจ ื”-x ืžืฆื™ื’ ืืช ืกื™ื›ื•ื™ื™ ื”ื–ืจื™ืงื”.
07:44
People on the left take difficult shots,
151
464639
1953
ื”ืฉื—ืงื ื™ื ืฉืžืฉืžืืœ ื–ื•ืจืงื™ื ื–ืจื™ืงื•ืช ืงืฉื•ืช,
07:46
on the right, they take easy shots.
152
466616
2229
ื•ืฉื—ืงื ื™ื ื‘ืฆื“ ื™ืžื™ืŸ ื–ื•ืจืงื™ื ื–ืจื™ืงื•ืช ืงืœื•ืช.
07:49
On the [y-axis] is their shooting ability.
153
469194
2057
ืฆื™ืจ ื”-y ืžืฆื™ื’ ืืช ื›ื™ืฉื•ืจื™ ื”ื–ืจื™ืงื” ืฉืœื”ื.
07:51
People who are good are at the top, bad at the bottom.
154
471275
2562
ื”ืฉื—ืงื ื™ื ื”ื˜ื•ื‘ื™ื ื ืžืฆืื™ื ืœืžืขืœื”, ื•ื”ื’ืจื•ืขื™ื ืœืžื˜ื”.
07:53
So for example, if there was a player
155
473861
1760
ื ื ื™ื— ืœื“ื•ื’ืžื” ืฉืฉื—ืงืŸ ืžื’ื™ืข ืœ-47 ืื—ื•ื–ื™ ื”ืฆืœื—ื” ื‘ื–ืจื™ืงื•ืช.
07:55
who generally made 47 percent of their shots,
156
475621
2097
07:57
that's all you knew before.
157
477718
1389
ื‘ืขื‘ืจ ื–ื” ื›ืœ ืžื” ืฉื™ื“ืขื ื• ืขืœื™ื•.
07:59
But today, I can tell you that player takes shots that an average NBA player
158
479345
4850
ืื‘ืœ ื”ื™ื•ื ืื ื™ ื™ื›ื•ืœ ืœืกืคืจ ืœื›ื ืฉื”ืฉื—ืงืŸ ื–ื•ืจืง ื–ืจื™ืงื•ืช
08:04
would make 49 percent of the time,
159
484219
1961
ืฉืฉื—ืงืŸ ืžืžื•ืฆืข ื‘ืืŸ-ื‘ื™-ืื™ื™ ืžืฆืœื™ื— ื‘-49 ืื—ื•ื–ื™ื ืžื”ืŸ.
08:06
and they are two percent worse.
160
486204
1684
ื›ืœื•ืžืจ, ื”ื•ื ื’ืจื•ืข ืžื”ืฉื—ืงืŸ ื”ืžืžื•ืฆืข ื‘ืฉื ื™ ืื—ื•ื–ื™ื.
08:08
And the reason that's important is that there are lots of 47s out there.
161
488266
4515
ื•ื”ื ืชื•ืŸ ื”ื–ื” ื—ืฉื•ื‘, ื›ื™ ื™ืฉ ื”ืจื‘ื” ืฉื—ืงื ื™ 47.
08:13
And so it's really important to know
162
493714
2549
ื•ื—ืฉื•ื‘ ืžืื•ื“ ืœื“ืขืช
08:16
if the 47 that you're considering giving 100 million dollars to
163
496287
3956
ืื ืฉื—ืงืŸ ื”-47 ืฉืฉื•ืงืœื™ื ืœืฉืœื ืœื• ืžืื” ืžื™ืœื™ื•ืŸ ื“ื•ืœืจ
08:20
is a good shooter who takes bad shots
164
500267
3055
ื”ื•ื ืงืœืข ื˜ื•ื‘ ืฉื–ื•ืจืง ื–ืจื™ืงื•ืช ื’ืจื•ืขื•ืช
08:23
or a bad shooter who takes good shots.
165
503346
2397
ืื• ืงืœืข ื’ืจื•ืข ืฉื–ื•ืจืง ื–ืจื™ืงื•ืช ื˜ื•ื‘ื•ืช.
08:27
Machine understanding doesn't just change how we look at players,
166
507130
3333
ื”ื”ื‘ื ื” ืฉืœ ื”ืžื›ื•ื ื” ืœื ืžืฉื ื” ืจืง ืืช ื”ื”ืกืชื›ืœื•ืช ืฉืœื ื• ืขืœ ืฉื—ืงื ื™ื,
08:30
it changes how we look at the game.
167
510487
1858
ื”ื™ื ืžืฉื ื” ื’ื ืืช ื”ื”ืกืชื›ืœื•ืช ืฉืœื ื• ืขืœ ื”ืžืฉื—ืง.
08:32
So there was this very exciting game a couple of years ago, in the NBA finals.
168
512369
3755
ืœืคื ื™ ื›ืžื” ืฉื ื™ื ื”ืชื ื”ืœ ื”ืžืฉื—ืง ื”ืžืจื’ืฉ ื”ื‘ื ื‘ืกื“ืจืช ื”ื’ืžืจ ืฉืœ ื”ืืŸ-ื‘ื™-ืื™ื™.
08:36
Miami was down by three, there was 20 seconds left.
169
516148
3207
ืžื™ืืžื™ ืคื™ื’ืจื” ื‘ืฉืœื•ืฉ ื ืงื•ื“ื•ืช, ื•ื ื•ืชืจื• 20 ืฉื ื™ื•ืช ืœืกื™ื•ื.
08:39
They were about to lose the championship.
170
519379
2025
ื”ื ืขืžื“ื• ืœืื‘ื“ ืืช ื”ืืœื™ืคื•ืช.
08:41
A gentleman named LeBron James came up and he took a three to tie.
171
521428
3341
ื‘ื—ื•ืจ ื‘ืฉื ืœื‘ืจื•ืŸ ื’'ื™ื™ืžืก ื–ืจืง ืฉืœืฉื” ื›ื“ื™ ืœื”ืฉื•ื•ืช ืืช ื”ืชื•ืฆืื”.
08:44
He missed.
172
524793
1198
ื”ื•ื ื”ื—ื˜ื™ื.
08:46
His teammate Chris Bosh got a rebound,
173
526015
1837
ื—ื‘ืจื• ืœืงื‘ื•ืฆื” ื›ืจื™ืก ื‘ื•ืฉ ืชืคืก ืืช ื”ื›ื“ื•ืจ ื”ื—ื•ื–ืจ,
08:47
passed it to another teammate named Ray Allen.
174
527876
2159
ื•ืžืกืจ ืื•ืชื• ืœื—ื‘ืจื ืœืงื‘ื•ืฆื” ืจื™ื™ ืืœืŸ.
08:50
He sank a three. It went into overtime.
175
530059
1919
ื”ื•ื ืงืœืข ืฉืœืฉื” ื•ื”ืžืฉื—ืง ื”ื’ื™ืข ืœื”ืืจื›ื”.
08:52
They won the game. They won the championship.
176
532002
2096
ื”ื ื ื™ืฆื—ื• ื‘ืžืฉื—ืง. ื”ื ื–ื›ื• ื‘ืืœื™ืคื•ืช.
08:54
It was one of the most exciting games in basketball.
177
534122
2444
ื–ื” ื”ื™ื” ืื—ื“ ื”ืžืฉื—ืงื™ื ื”ืžืจื’ืฉื™ื ื‘ื™ื•ืชืจ ื‘ื›ื“ื•ืจืกืœ.
08:57
And our ability to know the shot probability for every player
178
537438
3429
ื”ื™ื›ื•ืœืช ืฉืœื ื• ืœื“ืขืช ืืช ืื—ื•ื–ื™ ื”ืงืœื™ืขื” ืฉืœ ื›ืœ ืฉื—ืงืŸ ื‘ื›ืœ ืฉื ื™ื™ื”
09:00
at every second,
179
540891
1188
09:02
and the likelihood of them getting a rebound at every second
180
542103
2956
ื•ืืช ื”ืกื™ื›ื•ื™ ืฉืœ ื›ืœ ืฉื—ืงืŸ ืœืชืคื•ืก ื›ื“ื•ืจ ื—ื•ื–ืจ ื‘ื›ืœ ืฉื ื™ื™ื”
09:05
can illuminate this moment in a way that we never could before.
181
545083
3443
ื™ื›ื•ืœื” ืœืฉืคื•ืš ืื•ืจ ืขืœ ื›ืœ ืจื’ืข ื‘ื“ืจืš ืฉืœื ื”ื™ืชื” ืงื™ื™ืžืช ื‘ืขื‘ืจ.
09:09
Now unfortunately, I can't show you that video.
182
549618
2668
ืœืจื•ืข ื”ืžื–ืœ ืื ื™ ืœื ื™ื›ื•ืœ ืœื”ืฆื™ื’ ืœื›ื ืืช ื”ืกืจื˜ื•ืŸ.
09:12
But for you, we recreated that moment
183
552310
4493
ืื‘ืœ ืฉื—ื–ืจื ื• ืขื‘ื•ืจื›ื ืืช ื”ืจื’ืข ื”ื”ื•ื
09:16
at our weekly basketball game about 3 weeks ago.
184
556827
2336
ื‘ืžืฉื—ืง ื”ื›ื“ื•ืจืกืœ ื”ืฉื‘ื•ืขื™ ืฉืœื ื• ืœืคื ื™ ืฉืœื•ืฉื” ืฉื‘ื•ืขื•ืช.
09:19
(Laughter)
185
559279
2167
(ืฆื—ื•ืง)
09:21
And we recreated the tracking that led to the insights.
186
561573
3410
ื•ืฉื—ื–ืจื ื• ื’ื ืืช ื›ืœ ื”ืžื”ืœื›ื™ื ืฉื”ื•ื‘ื™ืœื• ืœืชื•ื‘ื ื•ืช ืฉืœื ื•.
09:25
So, here is us. This is Chinatown in Los Angeles,
187
565199
4255
ื”ื ื” ืื ื—ื ื• ื‘ืฆ'ื™ื™ื ื˜ืื•ืŸ ืฉื‘ืœื•ืก ืื ื’'ืœืก
09:29
a park we play at every week,
188
569478
1564
ื‘ืคืืจืง ืฉื‘ื• ืื ื—ื ื• ืžืฉื—ืงื™ื ื›ืœ ืฉื‘ื•ืข.
09:31
and that's us recreating the Ray Allen moment
189
571066
2231
ื•ื›ืืŸ ืื ื—ื ื• ืžืฉื—ื–ืจื™ื ืืช ื”ืจื’ืข ืฉืœ ืจื™ื™ ืืœืŸ
09:33
and all the tracking that's associated with it.
190
573321
2229
ื•ืืช ื›ืœ ื”ืžื”ืœื›ื™ื ืฉื”ื•ื‘ื™ืœื• ืืœื™ื•.
09:36
So, here's the shot.
191
576772
1517
ื–ืืช ื”ื–ืจื™ืงื”.
09:38
I'm going to show you that moment
192
578313
2516
ืขื›ืฉื™ื• ื ืจืื” ืืช ื”ืจื’ืข ื”ื”ื•ื
09:40
and all the insights of that moment.
193
580853
2587
ื•ืืช ื›ืœ ื”ืชื•ื‘ื ื•ืช ืฉืœื ื• ืขืœ ื”ืจื’ืข ื”ื”ื•ื.
09:43
The only difference is, instead of the professional players, it's us,
194
583464
3730
ื”ื”ื‘ื“ืœ ื”ื™ื—ื™ื“ ื”ื•ื ืฉืื ื—ื ื• ืžืžืœืื™ื ืืช ืžืงื•ื ื”ืฉื—ืงื ื™ื ื”ืžืงืฆื•ืขื™ื™ื,
09:47
and instead of a professional announcer, it's me.
195
587218
2618
ื•ืื ื™ ืžืžืœื ืืช ืžืงื•ืžื• ืฉืœ ื”ื›ืจื•ื– ื”ืžืงืฆื•ืขื™.
09:49
So, bear with me.
196
589860
1477
ืื– ืชื”ื™ื• ืกื‘ืœื ื™ื™ื ื›ืœืคื™ื™.
09:53
Miami.
197
593153
1150
ืžื™ืืžื™.
09:54
Down three.
198
594671
1150
ืคื™ื’ื•ืจ ืฉืœ ืฉืœื•ืฉ ื ืงื•ื“ื•ืช.
09:56
Twenty seconds left.
199
596107
1150
ืขืฉืจื™ื ืฉื ื™ื•ืช ืœืกื™ื•ื.
09:59
Jeff brings up the ball.
200
599385
1198
ื’'ืฃ ืžืงื“ื ืืช ื”ื›ื“ื•ืจ.
10:02
Josh catches, puts up a three!
201
602656
1535
ื’'ื•ืฉ ืชื•ืคืก ืืช ื”ื›ื“ื•ืจ ื•ื–ื•ืจืง ืฉืœืฉื”!
10:04
[Calculating shot probability]
202
604631
1849
[ื”ืกื™ื›ื•ื™ ืœืงืœื™ืขื” - 33%]
10:07
[Shot quality]
203
607278
1150
[ืื™ื›ื•ืช ื”ื–ืจื™ืงื” - ื’'ื•ืฉ - 33%]
10:09
[Rebound probability]
204
609048
1785
[ื”ืกื™ื›ื•ื™ ืœืจื™ื‘ืื•ื ื“ - ื ื•ืืœ - 12%]
10:12
Won't go!
205
612373
1173
ืœื ื ื›ื ืก!
10:13
[Rebound probability]
206
613570
1446
[ื”ืกื™ื›ื•ื™ ืœืจื™ื‘ืื•ื ื“ - ื ื•ืืœ - 32%]
10:15
Rebound, Noel.
207
615777
1256
ืจื™ื‘ืื•ื ื“, ื ื•ืืœ.
10:17
Back to Daria.
208
617057
1150
ื‘ื—ื–ืจื” ืœื“ืจื™ื”.
10:18
[Shot quality]
209
618509
3365
[ืื™ื›ื•ืช ื”ื–ืจื™ืงื” - ื“ืจื™ื” - 37%]
10:22
Her three-pointer -- bang!
210
622676
1620
ืฉืœื•ืฉ ื ืงื•ื“ื•ืช - ื‘ื ื’!
10:24
Tie game with five seconds left.
211
624320
2197
ื ื•ืชืจื• ื—ืžืฉ ืฉื ื™ื•ืช ืœืกื™ื•ื.
10:26
The crowd goes wild.
212
626880
1618
ื”ืงื”ืœ ืžืฉืชื•ืœืœ.
10:28
(Laughter)
213
628522
1659
(ืฆื—ื•ืง)
10:30
That's roughly how it happened.
214
630205
1547
ื›ื›ื” ื‘ืขืจืš ื–ื” ืงืจื”.
10:31
(Applause)
215
631776
1151
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
10:32
Roughly.
216
632951
1175
ื‘ืขืจืš.
10:34
(Applause)
217
634150
1531
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
10:36
That moment had about a nine percent chance of happening in the NBA
218
636121
5484
ืœืจื’ืข ื”ื–ื” ื”ื™ื” ืกื™ื›ื•ื™ ืฉืœ ื›ืชืฉืขื” ืื—ื•ื–ื™ื ืœื”ืชืจื—ืฉ ื‘ืืŸ-ื‘ื™-ืื™ื™.
10:41
and we know that and a great many other things.
219
641629
2261
ื•ืื ื—ื ื• ื™ื•ื“ืขื™ื ืืช ื–ื” ื•ื”ืจื‘ื” ื“ื‘ืจื™ื ื ื•ืกืคื™ื.
10:43
I'm not going to tell you how many times it took us to make that happen.
220
643914
3491
ืœื ืื’ืœื” ืœื›ื ื›ืžื” ื ื™ืกื™ื•ื ื•ืช ืขืฉื™ื ื• ืขื“ ืฉื”ืฆืœื—ื ื•.
10:47
(Laughter)
221
647429
1747
(ืฆื—ื•ืง)
10:49
Okay, I will! It was four.
222
649200
1872
ื˜ื•ื‘, ื‘ื›ืœ ื–ืืช ืื’ืœื” ืœื›ื. ื ื™ืกื™ื ื• ืืจื‘ืข ืคืขืžื™ื.
10:51
(Laughter)
223
651096
1001
(ืฆื—ื•ืง)
10:52
Way to go, Daria.
224
652121
1165
ื›ืœ ื”ื›ื‘ื•ื“, ื“ืจื™ื”.
10:53
But the important thing about that video
225
653647
4263
ืื‘ืœ ื”ืกืจื˜ื•ืŸ ื”ื–ื” ื•ื”ืชื•ื‘ื ื•ืช ืฉืœื ื• ืœื’ื‘ื™ ื›ืœ ืฉื ื™ื™ื” ืฉืœ ื›ืœ ืžืฉื—ืง ืืŸ-ื‘ื™-ืื™ื™
10:57
and the insights we have for every second of every NBA game -- it's not that.
226
657934
4568
ื—ืฉื•ื‘ื™ื ืžืกื™ื‘ื” ืื—ืจืช.
11:02
It's the fact you don't have to be a professional team to track movement.
227
662639
3929
ื•ื”ืกื™ื‘ื” ื”ื™ื ืฉืœื ืฆืจื™ืš ืœื”ื™ื•ืช ืงื‘ื•ืฆื” ืžืงืฆื•ืขื ื™ืช ื›ื“ื™ ืœืขืงื•ื‘ ืื—ืจื™ ืชื ื•ืขื”.
11:07
You do not have to be a professional player to get insights about movement.
228
667083
3657
ืœื ืฆืจื™ืš ืœื”ื™ื•ืช ืฉื—ืงืŸ ืžืงืฆื•ืขื™ ื›ื“ื™ ืœืงื‘ืœ ืชื•ื‘ื ื•ืช ืขืœ ืชื ื•ืขื”.
11:10
In fact, it doesn't even have to be about sports because we're moving everywhere.
229
670764
3858
ืœืžืขืฉื”, ื”ื ื•ืฉื ืื™ื ื• ืžื•ื’ื‘ืœ ืจืง ืœืกืคื•ืจื˜, ื›ื™ ืื ื—ื ื• ื ืขื™ื ื‘ื›ืœ ืžืงื•ื.
11:15
We're moving in our homes,
230
675654
2369
ืื ื—ื ื• ื ืขื™ื ื‘ื‘ืชื™ื ืฉืœื ื•,
11:21
in our offices,
231
681428
1205
ื‘ืžืฉืจื“ื™ื ืฉืœื ื•,
11:24
as we shop and we travel
232
684238
2690
ื›ืฉืื ื—ื ื• ื™ื•ืฆืื™ื ืœืงื ื™ื•ืช, ื›ืฉืื ื—ื ื• ืžื˜ื™ื™ืœื™ื
11:29
throughout our cities
233
689318
1253
ื‘ืชื•ืš ื”ืขืจื™ื
11:32
and around our world.
234
692065
1618
ื•ืžืกื‘ื™ื‘ ืœืขื•ืœื.
11:35
What will we know? What will we learn?
235
695270
2295
ืžื” ื ื“ืข? ืžื” ื ืœืžื“?
11:37
Perhaps, instead of identifying pick-and-rolls,
236
697589
2305
ืื•ืœื™, ื‘ืžืงื•ื ืœื–ื”ื•ืช ืคื™ืง-ืื ื“-ืจื•ืœ,
11:39
a machine can identify the moment and let me know
237
699918
3010
ื”ืžื›ื•ื ื” ืชื•ื›ืœ ืœื–ื”ื•ืช ืืช ื”ืจื’ืข
11:42
when my daughter takes her first steps.
238
702952
2059
ื•ืœื”ื•ื“ื™ืข ืœื™ ืžืชื™ ื”ื‘ืช ืฉืœื™ ืฆื•ืขื“ืช ืืช ืฆืขื“ื™ื” ื”ืจืืฉื•ื ื™ื.
11:45
Which could literally be happening any second now.
239
705035
2536
ื“ื‘ืจ ืฉืฆืคื•ื™ ืœื”ืชืจื—ืฉ ื‘ื›ืœ ืจื’ืข.
11:48
Perhaps we can learn to better use our buildings, better plan our cities.
240
708140
3697
ืื•ืœื™ ื ื•ื›ืœ ืœืœืžื•ื“ ืœื ืฆืœ ื‘ืฆื•ืจื” ื˜ื•ื‘ื” ื™ื•ืชืจ ืืช ื”ื‘ื ื™ื™ื ื™ื ืฉืœื ื•
ื•ืœืชื›ื ืŸ ื˜ื•ื‘ ื™ื•ืชืจ ืืช ื”ืขืจื™ื ืฉืœื ื•.
11:52
I believe that with the development of the science of moving dots,
241
712362
4173
ืื ื™ ืžืืžื™ืŸ ืฉื›ืฉืžื“ืข ื”ื ืงื•ื“ื•ืช ื”ื ืขื•ืช ื™ืชืคืชื—,
11:56
we will move better, we will move smarter, we will move forward.
242
716559
3643
ื ื ื•ืข ื˜ื•ื‘ ื™ื•ืชืจ, ื ื ื•ืข ื‘ืฆื•ืจื” ื—ื›ืžื” ื™ื•ืชืจ, ื ื ื•ืข ืงื“ื™ืžื”.
12:00
Thank you very much.
243
720607
1189
ืชื•ื“ื” ืจื‘ื”.
12:01
(Applause)
244
721820
5045
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
ื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