How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge
-
0:21 - 0:25Roy Price is a man that most of you
have probably never heard about, -
0:25 - 0:28even though he may have been responsible
-
0:28 - 0:34for 22 somewhat mediocre
minutes of your life on April 19, 2013. -
0:35 - 0:38He may have also been responsible
for 22 very entertaining minutes, -
0:38 - 0:40but not very many of you.
-
0:40 - 0:42And all of that goes back to a decision
-
0:42 - 0:44that Roy had to make
about three years ago. -
0:44 - 0:49So you see, Roy Price
is a senior executive with Amazon Studios. -
0:49 - 0:52That's the TV production
company of Amazon. -
0:52 - 0:55He's 47 years old, slim, spiky hair,
-
0:55 - 1:00describes himself on Twitter
as "movies, TV, technology, tacos." -
1:00 - 1:05And Roy Price has a very responsible job,
because it's his responsibility -
1:05 - 1:09to pick the shows, the original content
that Amazon is going to make. -
1:09 - 1:12And of course that's
a highly competitive space. -
1:12 - 1:14I mean, there are so many
TV shows already out there, -
1:14 - 1:17that Roy can't just choose any show.
-
1:17 - 1:21He has to find shows
that are really, really great. -
1:21 - 1:24So in other words, he has to find shows
-
1:24 - 1:26that are on the very right end
of this curve here. -
1:26 - 1:29So this curve here
is the rating distribution -
1:29 - 1:33of about 2,500 TV shows
on the website IMDB, -
1:33 - 1:36and the rating goes from one to 10,
-
1:36 - 1:39and the height here shows you
how many shows get that rating. -
1:39 - 1:44So if your show gets a rating
of nine points or higher, that's a winner. -
1:44 - 1:45Then you have a top two percent show.
-
1:46 - 1:49That's shows like "Breaking Bad,"
"Game of Thrones," "The Wire," -
1:49 - 1:52so all of these shows that are addictive,
-
1:52 - 1:55whereafter you've watched a season,
your brain is basically like, -
1:55 - 1:57"Where can I get more of these episodes?"
-
1:57 - 1:58That kind of show.
-
1:59 - 2:01On the left side, just for clarity,
here on that end, -
2:01 - 2:05you have a show called
"Toddlers and Tiaras" -- -
2:05 - 2:07(Laughter)
-
2:07 - 2:09-- which should tell you enough
-
2:09 - 2:11about what's going on
on that end of the curve. -
2:11 - 2:15Now, Roy Price is not worried about
getting on the left end of the curve, -
2:15 - 2:18because I think you would have to have
some serious brainpower -
2:18 - 2:20to undercut "Toddlers and Tiaras."
-
2:20 - 2:24So what he's worried about
is this middle bulge here, -
2:24 - 2:26the bulge of average TV,
-
2:26 - 2:29you know, those shows
that aren't really good or really bad, -
2:29 - 2:30they don't really get you excited.
-
2:30 - 2:35So he needs to make sure
that he's really on the right end of this. -
2:35 - 2:37So the pressure is on,
-
2:37 - 2:39and of course it's also the first time
-
2:39 - 2:41that Amazon is even
doing something like this, -
2:41 - 2:45so Roy Price does not want
to take any chances. -
2:45 - 2:47He wants to engineer success.
-
2:47 - 2:49He needs a guaranteed success,
-
2:49 - 2:51and so what he does is,
he holds a competition. -
2:51 - 2:55So he takes a bunch of ideas for TV shows,
-
2:55 - 2:57and from those ideas,
through an evaluation, -
2:57 - 3:01they select eight candidates for TV shows,
-
3:01 - 3:04and then he just makes the first episode
of each one of these shows -
3:04 - 3:07and puts them online for free
for everyone to watch. -
3:07 - 3:10And so when Amazon
is giving out free stuff, -
3:10 - 3:11you're going to take it, right?
-
3:11 - 3:16So millions of viewers
are watching those episodes. -
3:16 - 3:20What they don't realize is that,
while they're watching their shows, -
3:20 - 3:22actually, they are being watched.
-
3:22 - 3:24They are being watched
by Roy Price and his team, -
3:24 - 3:26who record everything.
-
3:26 - 3:29They record when somebody presses play,
when somebody presses pause, -
3:29 - 3:32what parts they skip,
what parts they watch again. -
3:32 - 3:34So they collect millions of data points,
-
3:34 - 3:36because they want
to have those data points -
3:36 - 3:39to then decide
which show they should make. -
3:39 - 3:41And sure enough,
so they collect all the data, -
3:41 - 3:44they do all the data crunching,
and an answer emerges, -
3:44 - 3:45and the answer is,
-
3:45 - 3:50"Amazon should do a sitcom
about four Republican US Senators." -
3:50 - 3:52They did that show.
-
3:52 - 3:54So does anyone know the name of the show?
-
3:55 - 3:56(Audience: "Alpha House.")
-
3:56 - 3:57Yes, "Alpha House,"
-
3:58 - 4:02but it seems like not too many of you here
remember that show, actually, -
4:02 - 4:03because it didn't turn out that great.
-
4:04 - 4:05It's actually just an average show,
-
4:05 - 4:10actually -- literally, in fact, because
the average of this curve here is at 7.4, -
4:10 - 4:12and "Alpha House" lands at 7.5,
-
4:12 - 4:14so a slightly above average show,
-
4:14 - 4:17but certainly not what Roy Price
and his team were aiming for. -
4:18 - 4:21Meanwhile, however,
at about the same time, -
4:21 - 4:23at another company,
-
4:23 - 4:27another executive did manage
to land a top show using data analysis, -
4:27 - 4:29and his name is Ted,
-
4:29 - 4:32Ted Sarandos, who is
the Chief Content Officer of Netflix, -
4:32 - 4:34and just like Roy,
he's on a constant mission -
4:34 - 4:36to find that great TV show,
-
4:36 - 4:38and he uses data as well to do that,
-
4:38 - 4:40except he does it
a little bit differently. -
4:40 - 4:44So instead of holding a competition,
what he did -- and his team of course -- -
4:44 - 4:47was they looked at all the data
they already had about Netflix viewers, -
4:47 - 4:49you know, the ratings
they give their shows, -
4:49 - 4:52the viewing histories,
what shows people like, and so on. -
4:52 - 4:54And then they use that data to discover
-
4:54 - 4:57all of these little bits and pieces
about the audience: -
4:57 - 4:58what kinds of shows they like,
-
4:58 - 5:00what kind of producers,
what kind of actors. -
5:00 - 5:03And once they had
all of these pieces together, -
5:03 - 5:04they took a leap of faith,
-
5:04 - 5:07and they decided to license
-
5:07 - 5:09not a sitcom about four Senators
-
5:09 - 5:12but a drama series about a single Senator.
-
5:13 - 5:14You guys know the show?
-
5:14 - 5:16(Laughter)
-
5:16 - 5:19Yes, "House of Cards," and Netflix
of course, nailed it with that show, -
5:20 - 5:22at least for the first two seasons.
-
5:22 - 5:26(Laughter) (Applause)
-
5:26 - 5:29"House of Cards" gets
a 9.1 rating on this curve, -
5:29 - 5:32so it's exactly
where they wanted it to be. -
5:32 - 5:34Now, the question of course is,
what happened here? -
5:35 - 5:37So you have two very competitive,
data-savvy companies. -
5:37 - 5:40They connect all of these
millions of data points, -
5:40 - 5:42and then it works
beautifully for one of them, -
5:42 - 5:44and it doesn't work for the other one.
-
5:44 - 5:46So why?
-
5:46 - 5:49Because logic kind of tells you
that this should be working all the time. -
5:49 - 5:52I mean, if you're collecting
millions of data points -
5:52 - 5:53on a decision you're going to make,
-
5:53 - 5:56then you should be able
to make a pretty good decision. -
5:56 - 5:58You have 200 years
of statistics to rely on. -
5:58 - 6:01You're amplifying it
with very powerful computers. -
6:01 - 6:05The least you could expect
is good TV, right? -
6:06 - 6:09And if data analysis
does not work that way, -
6:10 - 6:12then it actually gets a little scary,
-
6:12 - 6:15because we live in a time
where we're turning to data more and more -
6:15 - 6:20to make very serious decisions
that go far beyond TV. -
6:21 - 6:24Does anyone here know the company
Multi-Health Systems? -
6:25 - 6:27No one. OK, that's good actually.
-
6:27 - 6:30OK, so Multi-Health Systems
is a software company, -
6:30 - 6:33and I hope that nobody here in this room
-
6:33 - 6:36ever comes into contact
with that software, -
6:36 - 6:38because if you do,
it means you're in prison. -
6:38 - 6:39(Laughter)
-
6:39 - 6:43If someone here in the US is in prison,
and they apply for parole, -
6:43 - 6:47then it's very likely that
data analysis software from that company -
6:47 - 6:51will be used in determining
whether to grant that parole. -
6:51 - 6:53So it's the same principle
as Amazon and Netflix, -
6:53 - 6:58but now instead of deciding whether
a TV show is going to be good or bad, -
6:58 - 7:01you're deciding whether a person
is going to be good or bad. -
7:01 - 7:07And mediocre TV, 22 minutes,
that can be pretty bad, -
7:07 - 7:09but more years in prison,
I guess, even worse. -
7:10 - 7:14And unfortunately, there is actually
some evidence that this data analysis, -
7:15 - 7:19despite having lots of data,
does not always produce optimum results. -
7:19 - 7:21And that's not because a company
like Multi-Health Systems -
7:22 - 7:23doesn't know what to do with data.
-
7:23 - 7:25Even the most data-savvy
companies get it wrong. -
7:25 - 7:28Yes, even Google gets it wrong sometimes.
-
7:29 - 7:33In 2009, Google announced
that they were able, with data analysis, -
7:33 - 7:37to predict outbreaks of influenza,
the nasty kind of flu, -
7:37 - 7:41by doing data analysis
on their Google searches. -
7:41 - 7:45And it worked beautifully,
and it made a big splash in the news, -
7:45 - 7:47including the pinnacle
of scientific success: -
7:47 - 7:50a publication in the journal "Nature."
-
7:50 - 7:53It worked beautifully
for year after year after year, -
7:53 - 7:55until one year it failed.
-
7:55 - 7:57And nobody could even tell exactly why.
-
7:57 - 7:59It just didn't work that year,
-
7:59 - 8:01and of course that again made big news,
-
8:01 - 8:03including now a retraction
-
8:03 - 8:05of a publication
from the journal "Nature." -
8:06 - 8:10So even the most data-savvy companies,
Amazon and Google, -
8:10 - 8:12they sometimes get it wrong.
-
8:12 - 8:15And despite all those failures,
-
8:15 - 8:19data is moving rapidly
into real-life decision-making -- -
8:19 - 8:21into the workplace,
-
8:21 - 8:22law enforcement,
-
8:23 - 8:24medicine.
-
8:24 - 8:28So we should better make sure
that data is helping. -
8:28 - 8:31Now, personally I've seen
a lot of this struggle with data myself, -
8:31 - 8:33because I work in computational genetics,
-
8:33 - 8:35which is also a field
where lots of very smart people -
8:35 - 8:39are using unimaginable amounts of data
to make pretty serious decisions -
8:39 - 8:43like deciding on a cancer therapy
or developing a drug. -
8:44 - 8:46And over the years,
I've noticed a sort of pattern -
8:46 - 8:48or kind of rule, if you will,
about the difference -
8:48 - 8:51between successful
decision-making with data -
8:51 - 8:53and unsuccessful decision-making,
-
8:53 - 8:57and I find this a pattern worth sharing,
and it goes something like this. -
8:59 - 9:01So whenever you're
solving a complex problem, -
9:01 - 9:02you're doing essentially two things.
-
9:02 - 9:06The first one is, you take that problem
apart into its bits and pieces -
9:06 - 9:08so that you can deeply analyze
those bits and pieces, -
9:08 - 9:10and then of course
you do the second part. -
9:10 - 9:13You put all of these bits and pieces
back together again -
9:13 - 9:14to come to your conclusion.
-
9:14 - 9:17And sometimes you
have to do it over again, -
9:17 - 9:18but it's always those two things:
-
9:18 - 9:21taking apart and putting
back together again. -
9:22 - 9:24And now the crucial thing is
-
9:24 - 9:27that data and data analysis
-
9:27 - 9:29is only good for the first part.
-
9:29 - 9:32Data and data analysis,
no matter how powerful, -
9:32 - 9:36can only help you taking a problem apart
and understanding its pieces. -
9:36 - 9:40It's not suited to put those pieces
back together again -
9:40 - 9:41and then to come to a conclusion.
-
9:42 - 9:44There's another tool that can do that,
and we all have it, -
9:44 - 9:46and that tool is the brain.
-
9:46 - 9:48If there's one thing a brain is good at,
-
9:48 - 9:50it's taking bits and pieces
back together again, -
9:50 - 9:52even when you have incomplete information,
-
9:52 - 9:53and coming to a good conclusion,
-
9:53 - 9:56especially if it's the brain of an expert.
-
9:56 - 9:59And that's why I believe
that Netflix was so successful, -
9:59 - 10:03because they used data and brains
where they belong in the process. -
10:03 - 10:06They use data to first understand
lots of pieces about their audience -
10:06 - 10:10that they otherwise wouldn't have
been able to understand at that depth, -
10:10 - 10:12but then the decision
to take all these bits and pieces -
10:12 - 10:16and put them back together again
and make a show like "House of Cards," -
10:16 - 10:17that was nowhere in the data.
-
10:17 - 10:21Ted Sarandos and his team
made that decision to license that show, -
10:21 - 10:24which also meant, by the way,
that they were taking -
10:24 - 10:26a pretty big personal risk
with that decision. -
10:26 - 10:29And Amazon, on the other hand,
they did it the wrong way around. -
10:29 - 10:32They used data all the way
to drive their decision-making, -
10:32 - 10:35first when they held
their competition of TV ideas, -
10:35 - 10:38then when they selected "Alpha House"
to make as a show. -
10:38 - 10:41Which of course was
a very safe decision for them, -
10:41 - 10:43because they could always
point at the data, saying, -
10:43 - 10:45"This is what the data tells us."
-
10:45 - 10:49But it didn't lead to the exceptional
results that they were hoping for. -
10:50 - 10:55So data is of course a massively
useful tool to make better decisions, -
10:55 - 10:57but I believe that things go wrong
-
10:58 - 11:00when data is starting
to drive those decisions. -
11:00 - 11:04No matter how powerful,
data is just a tool, -
11:04 - 11:07and to keep that in mind,
I find this device here quite useful. -
11:07 - 11:08Many of you will ...
-
11:09 - 11:10(Laughter)
-
11:10 - 11:11Before there was data,
-
11:11 - 11:14this was the decision-making
device to use. -
11:14 - 11:15(Laughter)
-
11:15 - 11:16Many of you will know this.
-
11:17 - 11:18This toy here is called the Magic 8 Ball,
-
11:18 - 11:20and it's really amazing,
-
11:20 - 11:23because if you have a decision to make,
a yes or no question, -
11:23 - 11:26all you have to do is you shake the ball,
and then you get an answer -- -
11:26 - 11:29"Most Likely" -- right here
in this window in real time. -
11:29 - 11:31I'll have it out later for tech demos.
-
11:31 - 11:33(Laughter)
-
11:33 - 11:36Now, the thing is, of course --
so I've made some decisions in my life -
11:36 - 11:39where, in hindsight,
I should have just listened to the ball. -
11:39 - 11:42But, you know, of course,
if you have the data available, -
11:42 - 11:46you want to replace this with something
much more sophisticated, -
11:46 - 11:49like data analysis
to come to a better decision. -
11:49 - 11:52But that does not change the basic setup.
-
11:52 - 11:55So the ball may get smarter
and smarter and smarter, -
11:55 - 11:58but I believe it's still on us
to make the decisions -
11:58 - 12:01if we want to achieve
something extraordinary, -
12:01 - 12:03on the right end of the curve.
-
12:03 - 12:07And I find that a very encouraging
message, in fact, -
12:07 - 12:11that even in the face
of huge amounts of data, -
12:11 - 12:15it still pays off to make decisions,
-
12:16 - 12:18to be an expert in what you're doing
-
12:18 - 12:20and take risks.
-
12:20 - 12:23Because in the end, it's not data,
-
12:23 - 12:27it's risks that will land you
on the right end of the curve. -
12:28 - 12:29Thank you.
-
12:29 - 12:32(Applause)
- Title:
- How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge
- Description:
-
Does collecting more data lead to better decision-making? Competitive, data-savvy companies like Amazon, Google and Netflix have learned that data analysis alone doesn't always produce optimum results. In this talk, data scientist Sebastian Wernicke breaks down what goes wrong when we make decisions based purely on data -- and suggests a brainier way to use it.
This talk was given at a TEDx event using the TED conference format but independently organized by a local community. Learn more at http://ted.com/tedx
- Video Language:
- English
- Team:
- closed TED
- Project:
- TEDxTalks
- Duration:
- 12:40
Ivana Korom approved English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
Ivana Korom accepted English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
TED Translators admin edited English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
TED Translators admin edited English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
TED Translators admin edited English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
TED Translators admin edited English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
TED Translators admin edited English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge | ||
TED Translators admin edited English subtitles for How to use data to make a hit TV show | Sebastian Wernicke | TEDxCambridge |