Polling is Dead: Joe Biden vs. Bernie Sanders Fiasco.

there was a time in the distant past
when I really worked in political science political social science research and I found myself highly motivated all the time to speak out write articles and do what I could to declare hey what's going on in political research is a scam there was a time when I was involved with Buddhist research the history of Buddhism philology translation of ancient texts that sort of thing and I found myself tempted very frequently to speak out write articles and you know frankly it started all kinds of conflicts in controversy pointing the finger and saying hey guys wake up something going on here and eat is a scam until quite recently I was involved in vegan activism activism for veganism is political cause and I was frequently tempted frequently drawn out of my slumber to write articles and make youtube videos and really scream as loud as I could to draw attention sit look guys there's something going on here that is this game beneath the veneer of respectable mainstream press reporting but she's the reassuring kind of headlines that the people who write articles of the newspaper know what they're talking about and that you as a reader can become well informed by reading these articles we have we have a very dangerous game of self-deception and misinformation going on an example that matters today and maybe forgotten a few months from now but an example that matters today and that reveals the depth of this problem which is going to be a profound problem in the next 10 years Joe Biden allegedly is supported by 48 percent of african-american voters we normally say in plain English mmm forty eight percent of black voters in the Democrat Party allegedly support Joe Biden now there was a time there was a time when the footnote or the warning I would insert here is know what that means is 48 percent of black people who are willing to answer a phone call from a stranger and then answer a survey but via the phone that takes I don't know 10 12 minutes 48 percent of those black people vote for Joe Biden right and that that is itself a powerful source of skew a powerful source of bias a powerful source of misrepresentation but what we're dealing with today is so far removed from that it's so bizarre it's so surreal the fraud that is being passed off that's legitimate polling data that if I travel back in time to let's just say 1985 if I travel back in time to 1985 my only youth I was alive then and I walked up to a professor of political science and I described the methodology being used now by Ipsos knowledge panel he would say to me that's not polling what are you talking about that's not a poll of the general public of at all probably a mainstream professor of political science as recently as 1985 if you described him the methodology that's now being taken for granted and treated as legitimate by the Washington Post New York Times by mainstream Americans if you if you describe this to a professor at 9:35 he'd say what do you mean that sounds like something the Soviet Union would do that sounds like something some kind of communist government would do is like a conspiracy to misrepresent public opinion to create an illusion of what public opinion is what is its source it sews knowledge panel if so's knowledge panel is a a club of about 55,000 members as membership they're not polling the general public they have 55,000 members on the website of Epps's calm itself they do not directly say these people are paid it doesn't say they're paid and say they're unpaid I I have heard sources in political science stating it as a fact that they're paid googling on the internet now because you know there are 55,000 people who participate some of them must chitchat about this right I've found people saying that you're paid with some kind of voucher or coupon or discount so it's something like you get a gift card where you can buy something they were suggesting that it's not really payment in cash but that there is payment of some kind so there were 55,000 people selected recruited and remunerated 55,000 people employed to be the knowledge panel again if soviet russia did this we wouldn't regard it as legitimate it right now today if communist china did this we'd all laugh it out of the newspapers right if venezuela did this if iran did this all right if the muslim theocratic mixed democracy system in iran iran does have elections never forget that very religious country but they do have democracy they do they have elections and they have newspapers to get the debate who's the better candidate that kind of thing um if iran did this if iran said well we're not going to publish polling data we're not gonna pull people about who's popular we're gonna employ a few thousand people on a knowledge panel and their their job is to is to fill out these questionnaires for us over the internet through internet questionnaire all right it would be laughed out of the room and people would say very simply this is not a poll this is not what polling meant in 1985 it's not what polling met in 1995 either right the most fundamental technical technological change here was actually not the internet it was cellular phones and people not wanting to answer or comply or participate on cellular phones and even worse it's not uniform so a lot of the polling data done by telephone under-represented the type of person who would vote for donald trump so this is now an established fact enough money was paid to research it it's common sense but it has also been proven in as much as the social sciences could prove anything it's not just that left-wing people and moderates and right-wing people all equally refuse to comply with telephone polling you can imagine there are certain types of left-wing social justice warrior type people who are happy to get that call say hey we want to do a poll with you and we want to talk to you for 12 minutes and ask you some questions and your answers could help make the world a better place there are certain types of optimistic happy people who will then participate that fall and disproportionately the type of people who vote for Donald Trump hang up the phone you can say they're impatient you can say they're anti-establishment you can see whatever that is this was one of the reasons not the only one why polls failed to report accurately how many people in the Rust Belt bridge in the United States were actually going to vote for Donald Trump in general polling is dead polling as we knew it polling as we came to rely on it is dead and I can accept that political science can move forward journalism can move forward but I'm not gonna lie to here I think there should be a law a federal law passed in the United States that you cannot refer to a a knowledge panel survey as a poll all right now there are other kinds of social science research that I think could rise up to challenge and replace what we thought of as polling in the past and the current scam fraudulent pseudo polls that are being published claiming the 48% of black people support Joe Biden and look I don't even know but a few days later there was another another poll published which again is probably no more reliable that instead found 27% of African Americans support you abide okay why do you think that is why do you say there were two poles with two radically different numbers and we're not even talking about a very large there aren't that many African Americans the United States who vote for Joe Biden we're not talking a phenomenon of the scale of billions of people this is not like trying to ascertain how many people in China are satisfied with their dentists or something okay this is a pretty specialized area of talking um not and not all african-americans even care about the the Democrat primary so you're polling a minority within a minority with a minority we should have consistency if there's any legitimacy to the polls at all if these are indeed polls at all okay look there are alternatives but we can't misrepresent these alternate methods as polling data we can't eat I don't even think you should use the term survey you had a selected public they have 55,000 people on the so-called knowledge panel and they're being paid to do this as far as I know they're being given vouchers and coupons they're giving some kind of payment to reward them for their time and put into this this is not polling okay and who are they who chose will they reassure you oh we used rigorous methodology and you know we select them so they can be representative guess what you probably chose the type of elderly especially interested in politics african-americans who support you Biden this very poll shows the problem sir this pseudo poll the results comparing this forty eight percent this shows exactly the kind of skew exactly the kind of bias exactly the kind of misrepresentation that's built into that not despite your rigorous methods but because of your rigorous methods what about the people who don't care about politics what about the people who vote but who do not want to sign up for your stupid knowledge panel who don't want to become a member who don't want to receive the junk email who don't want a 50 cent voucher or whatever the never the payment is people who don't want to play that game what about the people who hang up the phone what about the people who don't go in for this it's like that's the ultimate source of bias here is it's like the people who like participating in political bias I guess it's the same thing if you have one of those one of those one of those websites on the internet we people like click a button to show that they want social change you know and those people are so naive they're like oh we got 30 million clicks when we ask this question so you know therefore such-and-such percentage of the American people want this wonderful socially no no I mean you know you're measuring look this is not even getting into the question of intentional fraud if you are paying people to be on this stupid panel I don't know if they get paid $5.00 I don't know if they get paid $50 what I found the internet suggested the payment is very modest let me tell you something $5 is not much money if you live in Los Angeles if you're a retired african-american left-wing person living in Los Angeles $5 is not a lot of money five US dollars is a lot of money in Cambodia you could have people who sit down as a full-time enterprise in Cambodia sign up for multiple accounts with knowledge panel and they can do it all man because if they have multiple accounts they're getting five dollars from each account they might have 20 accounts Cambodia that's real money they can get themselves a VPN they can get some phony paperwork they can download some stuff from the internet say oh yeah this is my address they have one address for one email in California one that's in Maine and one that's in Michigan they can get set up and if you think that only scammers in Cambodia might be highly motivated to do this there might be some highly professional people in the service of foreign governments who want to do it one of the very few articles I read about this was written by a political science professor I'm sorry Fred he was either politics or history but a professor who said that he had a friend who was intentionally perverting the system a friend who was out of his own academic interest this was another university professor who had signed up so he could kind of spy on this system and CEO was working and he was intentionally putting in like really wacky and extreme results to see if if the system would check him or what would happen like what happens if you give really contradictory crazy answers like you know are they gonna detect that there's something wrong or you know what's gonna happen so I can imagine that some some bored academic at a university with a PhD signing up and joining one of these one of these survey groups because they want to know where do the poles come from this is like going inside the sausage factory and seeing how sausage is made okay obviously this set of cultural assumptions developed in response to the demise of telephone based polling the increasing public reluctance to answer such a call what have you and it evolved in reaction to the opportunities presented by the Internet where it now seems so easy to send out a spam email to 500 people and say hey would you like to make five dollars you got I know if it's five dollars or fifty dollars answering some questions on on polls and a certain type of person will say yes but it's not the average person it's not a representative person and the more of an effort this company makes to make it representative through rigorous methodology the more unnatural that the more phony the more fraudulent the whole procedure is because yes sure they have a preconceived notion of what percentage of black people support you abide and so if you just use the the data raw without waiting what do we mean by waiting the data in political science right that means they look at and go like oh well these numbers don't look right you know I think we should give more weight to the people in Indiana and less weight to the people in California for this reason they're you're thinking it through I mean the more of this you do the more nuanced you add the more interpretation the more rating the less legitimacy it as a poet was like the more methodological rigor they apply to this in trying to make it representative the phony Rand phony or gets and this claim this claim that 48% of African Americans vote for Joe Biden it was based on a survey of only 769 American blacks and they're not chosen at random they're not walking out of a subway station like the injury interviewed as they walk out of a subway station they're not chosen by walking down the street and picking every seventh house as you walk down the street and those are old-fashioned survey methods where you walked on foot door-to-door and every seventh number whatever whatever the methodology was okay no these are 769 people who are selected recruited and paid to participate in this poll okay polling is dead the good news is there's a new role like never before for documentary filmmaking to step in what we can do now more than ever before is get qualitative data where you go in and make a deal with a shoe factory owner say ok we're gonna pay you a little bit of money to interrupt your work at this shoe factory a little bit for one day you don't have to stop the whole factory we're gonna come in in one day we're gonna interview people in groups of 30 so 30 people come in and do an interview 30 people come in we're gonna work through all the staff we're gonna film it it's going to be good advertising for your shoe factory we're gonna put the film on the internet it's gonna be here on YouTube say hey guess what we decided to talk to all the employees of the shoe factory in depth about who they're voting for and why you what you're going to get you're going to get really interesting nuanced qualitative data and you're gonna get an open question we can all participate in of to what extent are the people who work in that shoe factory representative or not representative of Americans as a whole because in the 21st century when polling is dead you just can't know