Exclusive - Nate Silver Extended Interview Pt. 2

  • Aired:  03/27/14
  •  | Views: 18,303

Nate Silver discusses the data analysis-driven reporting model of FiveThirtyEight.com and shares the challenges of achieving objectivity. (6:15)

[CHEERS AND APPLAUSE]WE'RE BACK HERE WE'RE DISCUSSING

SOME THINGS WITH NATE SILVER.

HE HAS A NEW WEB SITEFIVETHIRTYEIGHT.COM IT'S THE

SAME NAME AS YOUR OLD WEBSITEBUT IT HAS EXPANDED CONTENT.

>> THAT'S RIGHT. SO WE HAVE FIVEMAJOR VERTICALS.

SPORTS, POLITICS, ECONOMICS,SCIENCE AND LIFESTYLE

OF COURSE VERYTHING IN THEWORLD IS AT LEAST ONE OF THOSE

THINGS IF NOT MORE.

BUT THE IDEA IS THAT WE WANT TOTRY AND FIND THINGS WHERE OUR

SKILL SET MEANING STATISTICS,COMPUTER PROGRAMMING WE STILL DO

SOME REAL REPORTING.

WE'RE NOT AGAINST REPORTING.

WE'RE AGAINST PUNDITRY.

SO IT'S A MIX OF SKILLS WE'RETRYING TO FIGURE OUT HOW BEST

TO -->> Jon: PEOPLE HAVE GONE AFTER

YOU AS THOUGH YOU HAVE SAIDYOURSELF THIS IS THE HOLY GRAIL

OF ANSWERS.

AND I HAVE NEVER HEARD YOU --YOU HAVE ALWAYS BEEN VERY HUMBLE

ABOUT THE LIMITATIONS OFDATA ANALYSIS.

>> THAT'S KIND OF POINT OF MYBOOK.

ACTUALLY THE STUFF IS REAL HARDBECAUSE IT'S HARD TO KNOW WHAT

YOUR OWN PERSPECTIVE OR BIASESMIGHT BE.

OFTEN TIMES THE DATA ITSELF IS AVERY, YOU KNOW, POOR QUALITY.

A LOT OF ECONOMIC STATS WHENTHEY FIRST COME OUT AREN'T VERY

RELIABLE.

THEY GET REVISED YEARS LATERAFTER YOU'VE ALREADY MADE THE

WRONG POLICY CHANGE,FOR EXAMPLE.

>> Jon: HOW DO YOU FINDTRUSTED SOURCES OF NUMERICAL

INFORMATION?

DO YOU HAVE GO-TOS?

HOW DO YOU FIND THEM THATHAVEN'T BEEN POLLUTED BY

IDEOLOGICAL BIAS OR BY SAMPLEBIAS OR ANY OF THOSE TYPES OF

THINGS?

>> IT'S REALLY HARD.ONE THING WE'VE CERTAINLY

LEARNED IS THAT ALL THECHALLENGES THAT ARE

REGULAR, NEWSROOM FACES WE'RENOT ABOVE AT ALL.

THEY ARE VERY, VERY CHALLENGING.

HOW DO YOU VET STORIES.

HOW DO YOU DEAL WITH A CASEWHERE SOMEONE DISAGREES?

I THINK OUR VIEW IS THAT THEMORE YOU SHOW THE READER INSTEAD

OF TELL THEM.

SAY HERE IS MY HOMEWORK.WE PUT THINGS UP ON GITHUB AND

OTHER FILE SHARING PLACES WHERE YOU CAN SAY GO CHECK OUR WORK

>> Jon: SURE.

>> TODAY SOMEONE FOUND, FOREXAMPLE, A MINOR ERROR IN THE

NCAA TOURNAMENTPROJECTIONS RIGHT.

IF WE HADN'T PUBLISHED THAT THEERROR WOULD HAVE EXISTED OUT

THERE AND NO ONEWOULD HAVE KNOWN.

SO THAT'S WHATWE'RE MOVING TOWARD, I GUESS THE

IDEA BEING TRANSPARENCY ISTHE NEW OBJECTIVITY.

>> Jon: I'VE FOUND THERE'SNOTHING AND I THINK THE AUDIENCE

WILL VET THIS AND THERE'SNOTHING THAT MY AUDIENCE LIKES

BETTER THAN WATCHING SOMEONE DOTHEIR HOMEWORK.

[ LAUGHTER ]>> IT'S PEOPLE LIKE THAT KIND OF

ARTICLE ON GITHUB.

>> Jon: NOW WHAT IS GITHUB?

I FEEL BADLY THAT I'M THAT OUTOF TOUCH THAT I DON'T KNOW

GITHUB BECAUSE KNOW I FEEL LIKEIT'S SOMETHING I SHOULD KNOW.

>> IT'S LIKE FACEBOOK FORUBER-NERDS BASICALLY.

IT'S A PLACE WHERE PEOPLE SHARECODE AND DATA.

ESPECIALLY PEOPLE WHO ARE INDATA JOURNALISM AND COMPUTERASSISTED

REPORTING.

>> Jon: SO IT'S LIKE TINDER FORNEO. IT WOULD BE,

YOU WOULD LOOK AND SEE 0111SWIPE TO THE RIGHT.

>> THAT'S THE RIGHT S.A.T.

ANALOGY.

>> Jon: 01011 SWIPE TO THELEFT.

DO YOU FIND THAT IN GENERAL THEPEOPLE THAT USE THIS FOR

POLITICAL PURPOSES WANT TOCO-OPT WHAT YOU DO AND

THEY'RE NOT UNDERSTANDING THATTHE METHODOLOGY YOU APPLY

IS CONSISTENT.

>> SOME PEOPLE THOUGHT WE WEREKINDA BLUFFING LIKE OH HE'S

REALLY A DEMOCRAT IN DISGUISEAND NOT BECAUSE THAT'S WHAT THE

POLLS POINT TOWARD SO PEOPLE ARELEARNING THIS YEAR LIKE IN

2010 IT'S LIABLE TO BE ATOUGH YEAR FOR DEMOCRATS.

BUT I DO THINK IT'S HARD INPOLITICS IN PARTICULAR BECAUSE

PEOPLE, I ADMIRE POLITICALPASSION, I ADMIRE PEOPLE

WHO ARE WORKING IN THEIR MIND TOMAKE THE WORLD A BETTER PLACE.

IT'S VERY HARD TO PUT YOURBELIEFS AHEAD OF WHERE THE

EVIDENCE COME IN -- I MEANOBJECTIVITY IS A REALLY HARDTHING

I THINK A LOT OF NEWS ROOMS HAVEKINDA SAID WE'RE JUST

GOING TO REPORT WHAT THEY SAYAND WHAT THEY SAY BECAUSE IT'S

TOO HARD TO KNOW WHAT THEWORLD LOOKS LIKE.

>> Jon: IT ALSO, YOU KNOW, ITBEGINS TO RESEMBLE AN IDEA

THAT THE NEWS ROOM HASABDICATED ANY SENSE OF

EDITORIAL AUTHORITY.

WHEN YOU LOOK AT -- I READ ANARTICLE TODAY.

THERE WERE 10600 PUBLISHEDCLIMATE STUDIES, 10598 OF THEM

AGREED THAT GLOBAL WARMINGIS REAL AND HUMANS HAVE A

ROLE TO PLAY IN THAT.

TWO DIDN'T.

>> RIGHT.

>> Jon: STATISTICALLYIT'S INSIGNIFICANT BUT

IF YOU LOOK AT THE WAY THATARGUMENT

PLAYS OUT IN THE NEWS MEDIAWOULD YOU THINK IT'S A MUCH

CLOSER BATTLE.

>> WHAT IS DIFFICULT I THINK ISPEOPLE HAVE THE FRAME WHERE IT'S

EITHER 50-50 OR 100-0.

A LOT OF TRUTHS MIGHT BE 75-25OR 98-2 OR 83-17.

YOU KNOW, TO DESCRIBE THAT ASVERY CHALLENGING TO DESCRIBE

PROBABLISTIC THINGS IS VERY HARDTO SAY THEY ARE MOSTLY RIGHT BUT

HERE ARE SOME CAVEATS THAT FORTHE INTEREST OF COMPLETENESS WE

OUGHT TO SAY AND FIGURING OUTHOW TO TOE THAT LINE BUT ALSO

HAVING A WEB SITE THATAPPEALS TO REGULAR PEOPLE. THAT

IS WHERE WE'RE GOING TO NEEDMORE REPS AND MORE PRACTICE.

>> Jon: THE MORE YOU WATCH THENEWS MEDIA AND THE MORE THEY

BREAK THINGS DOWN IS THAT GOODOR IS THAT BAD.

TRYING TO BRING TO YOU WELL ITCAN BE PERHAPS MORE THEMAJORITY

GOOD BUT THERE'SSTILL A DOWNSIDE IS A DIFFICULTPLACE TO BE.

>> I THINK IT'S ALSO WHY WHEN WEHAVE A MORE COMPLETE BODY OF

WORK FOR PEOPLE TO SEE, RIGHT.

>> Jon: IT'S BEEN A WEEK AND AHALF, NATE.

I THINK WE KNOW.

>> THE FIRST ARTICLE ON CLIMATECHANGE IS NOT REPRESENTATIVE OF

EVERY ARTICLE OF WHAT WE HOPEWILL BE HUNDREDS OF ARTICLES YOU

WOULD SEE ABOUT CLIMATE CHANGEOVERALL ON THE SITE.

THERE'S AN IRONY.

I HAVE GIVEN A LOT OF SHIT TOPEOPLE OVER THE YEARS.

SO WE HAVE A LOT OF THIS COMING.

THERE'S A IRONY IN WHATWE SAY ABOUT PUNDITS AS WELL,THEY

KINDA CHERRY PICK FACTS AND RUSHTO JUDGMENT AND YOU SEE A

LITTLE OF THAT AS WELL BUTTHERE'S TERRIFIC STUFF ON THESITE.

>> Jon: I'M ENJOYING IT.

IT THINK IT LOOKS GREAT.

THE ONE THING I WOULD SAY YOURARTICLE ON THE DUKE LOSS TO

MERCER BEING WORSE THAN THE LOSSTO LEHIGH LET'S FACE FACTS

THAT RIDICULOUS. YOU'RE NOTTAKING INTO ACCOUNT THE LOSS OF

THE PLUMLEES. THERE'SA HEART THERE.

YOU UNDERSTAND WHAT I'M SAYING.

YOU REALLY THINK THE DUKE LOSSTO MERCER WAS WORSE THAN THE ONE

TO LEHIGH.

>> OUR MANAGING EDITOR HAS ASTUDENT AT DUKE RIGHT NOW.

WE'RE HAVING A VERY IMPORTANTMEETING AND HE BROKE DOWN.

MIKE IS BACKSTAGE.

>> YEAH IT'S NOT TRUE.

THE WON TO LEHIGH IT WAS A LOTWORSE.

[LAUGHTER]NATE SILVER.

[ APPLAUSE ]

Loading...