 Hello and welcome to NewsClick. I am Poranjoy Gohar Thakurtha and with me here in the studio, I have a 40 year old political analyst, calm advertiser, advertising expert, calm data analyst. He is many things rolled into one. I am very happy to welcome Narendra Nag, co-founder of Anthro.ai, formerly a digital officer in a firm in Norway. In the 20 years that working, Narendra, you were half, you've been a journalist, the other half you've been communications and advertising. In journalism, you've worked with the Hindustan Times, CNN IBM, NewsX, Mint. And as a communications and advertising person, you worked in China, Japan, India and the US among other with the big, big publicists. All right. Narendra, the reason why a lot of people are very keen to know your views today is your organization has projected that the Bharata Janta Party could get as few as 21 out of the 80 Lok Sabha seats in the elections which are over the results of which will be declared on the 23rd of May. And you have said that the Mahagat Bandhan, if you call it, the Bhaujan Samaj party, Samajwadi party, Rasul Uqdal would win over 54 or 54 out of thereabouts. So Narendra, if you can, you will be a hero. But many people are saying that this exit poll is saying something else. And if you approve wrong, people will say that from where? This is our Prime Minister, Bhavishwani is talking about the biggest kingdom of Bharatvish. Yes, your views. Well, my response to that is very simple. We decided to do this to test out systems and methods that have traditionally not been applied in the political space as far as we are aware. And from the very beginning we've been fine with the idea of being wrong. We don't think we are, but we've been okay with that because whether we're right or wrong, there's a lot of learning in both for us. That learning will help us strengthen our models, strengthen our methods, and hopefully be able to analyze and understand human motivations and behavior better. That genuinely is the only reason why we wanted to do this. Because an exercise of this sort doesn't come along often where you have so many people who care so deeply about something. Everybody knows that Uttar Pradesh is the most populous state in India. One out of six Indians live in Uttar Pradesh. If Uttar Pradesh was an independent country, it would have been the sixth most populous country in the world after China, India, Indonesia, the United States and Brazil. And Uttar Pradesh has the diversity, the heterogeneity. Well, that was definitely part of the attraction of the state. Political Mahitva informed the decision, right? We wouldn't want to do something which didn't really matter. We were doing it in a state which didn't really matter. We wouldn't actually have enough signals out there to help test our models. Doing it in a state like UP gives us the diversity of a country. It gives us scale and size of a country and it gives us the importance of a country as well. So I think those were the reasons which sort of drove our decision more than anything else. If what you are saying is right, if the Indian People's Party and the Indian Union are closed, how far do we go? When will Modiji again become Prime Minister? It is very difficult for us to predict that. The reason for that is that if we are right in Uttar Pradesh and Madhya Pradesh, Rajasthan, Chhattisgarh, West Bengal, they sweep the entire state like no one has said, then they can become Prime Minister again. Our job is not to predict whether they will become Prime Minister or not. Our job is very simple. All the exit polls have come. Times now, VMR is saying 8000. ABPC Voter is saying 25000. India TV CNX 50000. Today's channel is 65000. Republic Television and C Voter. Only one C Voter. ABP and Republic are together. In one place 25000 and in one place 38000. I don't know if one person did it or the other person did it. Republic Bharat. Jan Ki Baat. 45 to 57000. News Nation 38 to 40000. India Today Access. Basat to 8000. And you have to say it alone. How far will it go? ABP and C Voter are closest. You said 25000. You said 21000. So either you will be proved. Either we will be very wrong or we will be very right. It's going to be one of the two. It can't be anything in between. We actually think that if we are wrong, then we are going to be wrong at so many levels that it's not going to be 21. It won't be like 21 to 30. It will be like the BJP is meaning 65 or 68 or 70 seats or something. So you are saying it will either be the BJP will get too very well close to what it got last time or it will get too very badly. That's exactly right. And we haven't seen anything to suggest that it will do very well. We could of course be wrong and we are happy being wrong but we haven't seen anything yet which would suggest that. You have shown many reasons. You have said that the vote share of the Indian People's Party is 3 to 6% possible. There is a tight bond between the Indian People's Party and the Indian People's Party. So how did you come across? Why did you arrive at these figures? So one of the things that we have been looking at is what happened in the BIPOLS. So in the BIPOLS it wasn't just when the SP and the BSP came together and won them. They didn't just win them by adding up their vote shares. They actually got more than what their vote shares were in 2014. And as we've studied the state over the last six months now, actively, we saw that when they were winning there was always an additive factor. That additive factor could come from Muslim voters who were moving away from the Congress and choosing to vote for them because they believed that this was a more winnable candidate. So there are other people who are joining together to vote for a candidate they think will win. So we didn't believe that they would just win by the addition of the SP and the BSP vote. We sort of saw that already happening that there would be other voters who would vote for them. Many of our politicians, such as Vislashak, are asking who is helping Sapa. Who is helping Sapa, the OBCs, other backward classes. In this, there are Yadav people. And you yourself said that this time people would vote for them. So if Yadav is not there, but the OBCs, why would they go together this time, along with Sapa, who didn't go in the name of Vidhan Sapa. After the vote. The other thing is that along with it, people are saying that it's okay, the vote of the BJP, the caste vote and all the other votes, okay, the vote will be transferred. They will go to Sapa's hope. Not the other way around. Actually, we found something very interesting. The Jata vote in the first few phases, it less of it transferred than in the last few phases. And the reason for it is that the initials, there was a seminal moment. The moment is when Mayavati and Mulayam shared a stage together. At that point, even the Jata voter who was unlikely to go and vote for the others. Correct. That guest house happened 25 years later. Correct. You know, for whatever reason, there was a certain section that said, let's just wait and watch. So, I don't think we saw as much Jata's turnout for the Garbandan in the first phase as we did in the seminal. You said that when when Mayavati and Mulayam Singh Yadav were in one stage. That was hugely impactful. Hugely impactful at multiple levels. The second hugely impactful thing was to see Dimple Yadav touch Mayavati's feet in another rally. The one that happened at Nimkan Nog. And both of these events, you know, you can sort of say these things don't matter or are normative within a sort of urban context, but these are not normative within the reality of the operation. In the run-up to 2014, Amit Shah moved quickly to capitalize on years of work by the BJP wooing non-Yadav OBCs, non-Jata Dalits in Uttar Pradesh by picking office bearers, candidates from these communities. The BJP offered itself as a first party of choice for the people who were left out, in a sense, who felt left out by the Samajwadi Party, the Bhuvjan Samajwadi and the Congress. And the people had expectations rooted in what they believed had happened and they were saying that on the ground, the Yadav and the Jatav are enemies. How can they vote for a candidate? How can they be united? So access to power that translates into a disproportionate, disproportionate share of benefits and increased respect in their neighborhoods. So how did this change? I actually think this started with the Bipoles. If the Bipoles had not happened and had not been pulled off successfully, then these communities, the existing sort of suspicions would have stayed. The second thing is that Akhilesh Yadav's role in the way that he has behaved with Mayavati all along, with great respect, by allowing her to be the senior partner or seem like the senior partner, by giving her more seats, there has been this constant sort of, you know, he's made sure that at the risk of sometimes perhaps alienating some of his own members of his own party, he has made sure that she has been accorded the respect which allows for people within communities to start to treat each other with less suspicion and he started to define what the behavior should be. Narendra, you have written that the people of UP might use these elections as an opportunity to treat their lesson. And you quote William Congry's famous lines of 1697, heaven hath no rage like love to hate return, no hell a fury like a woman scorned. And you also said that this is not hunger this is an undercurrent. It's not an undercurrent, I actually call it earthquake. So you are saying that this is hunger. Yeah, because you know when one of the things, we kept picking this up. We kept picking up anger against the BJP for having left Yogi Adityanath actually having sprung him as a surprise chief minister after 2017. So you know the old social coalition and I hesitate to call it a social coalition. I'll call it a political sort of coalition where they got the Kushwas on board and the Kormis on board and the other OBCs on board and many of the other non-Jata Dalits on board. They didn't expect a Thakur chief minister and they certainly didn't expect the sort of the semiotics of the Thakur chief minister to be as to be what Yogi Adityanath are. That's one thing which I think has played on some people. And this there is a lack of respect in that choice inherent in that choice. You are saying that there are a lot of issues around the issue that you have isolated, resource crunch and the education that is completely over the infrastructure and the deficit you mentioned about the iron internal migration in the rural areas that are going on farm distress us versus them dented by demonetization water woes, ballooning job crisis and you have given so much in English holy cow to holy terrors. We have had a little fun when you are doing work like this and you are primarily doing it for yourself but you cannot have not allow yourself to at least enjoy the work that you are doing would be a crying shame. So yeah we have had a little fun with those headlines all along. So you are saying that all these issues matter and a lot of people are saying that there is a nationality between the Hindu-Muslims who oppose the Indian party and say that it is possible but you are saying the issues of religion the issues of Ram Mandir the issues of nationalism post pulwama and Balakot will not matter and these issues will matter? No I think they do matter but they matter to the BJP voter and it is very, we have also written about the formidable BJP voting block that they have managed to create people who want the country to be who want a muscular national narrative people who believe that it is possible to sort of escape the circumstances of their birth and most of these people will be urban we have noticed that but there is a larger section which aspires to these things and we think that those people are voting for the BJP and perhaps for good reason because there is a world view that they subscribe to which makes sense to them it would be dangerous to dismiss them using negative terms because I would not dismiss the voter in that fashion I actually like whenever we were studying them we had a lot of empathy for them because what they wanted was a better life they were different So what technique did you use? You are saying that the technique you used it is unique you said Anthro.ai is an experiment in alchemy between experienced communication specialists data scientists anthropologists, mathematicians and between the human mind heart and our instincts and immersive data streams we have ingested and analyzed nearly all publicly indexable data in or about UP news stories, job postings, classifieds about farm equipment, questions, answers you have tried to match this data to a one square kilometer grid of Uttar Pradesh Sir, this is all gobbledy this is all gobbledy goop to many people I mean it's very simple if you look at the network penetration and the amount of things that we are saying and you choose not to filter it by popularity I will not look at things which everybody else is looking at I will look at things perhaps that have just been said that nobody might be looking at what you start to see then is a map of human expression that human expression is personal but it's also publicly posted which means it's indexable so like a search engine can go out and index all information that's posted online that information then becomes available to you now as human beings we can't necessarily consume that volume of information but in the last few years it's been possible to computationally process that information using multiple methods there is no one single monolithic way of doing this what those things do is give you patterns that's when you bring in a trained anthropologist or you bring in somebody who understands these things and allow them to look at these patterns to understand what's going on what are people's fears or most behaviors you are being scolded you are being scolded by the Congress you are being scolded by the National Herald and you said they are questioning your motives they've also called me a commie coder which I actually quite like I'm going to get a T-shirt printed but we don't actually have a problem with that because we surface that stuff before they find it many of those right-wing portals have helped us a lot to help us understand the BJP water or the right-wing sort of Indian and especially on places like Bharatraksha canal you'll find fairly cogent arguments being made by these guys I wouldn't dismiss them ever and I'm okay with being called names because it doesn't really matter to me and some of these names I actually like as well but we don't have an issue with being called names I think our data net allows us to study people in an empathetic way and then arrive at the conclusions that we arrive at which is a fairly multi-layered process I might add I mean I don't want to make it we'll have to wait and watch we'll know whether you're right or not you've stuck your neck out or if we are right we will double down on what we've learned and continue to build our systems if we are wrong we will learn from our mistakes fix our systems and be stronger for it I'll tell you some time thank you once again for giving us your time and thank you very much for being with NewsClick and we'll know really after the 23rd of May whether what Narendra Nag and his team at Anthro.ai and I must mention some of the members of his team they include Rahil Khursheed Aruna Handig, Zoya Wahi and others whether their prediction that the Bharti Janta party's number and the NDA's number would come crashing down from 73 out of 80 to 21 or they're about this and they're not very sure about 8 or 9 seats but even if it's like 30 that would be a very, very significant loss for the BJP we'll know only after the 23rd of May thank you for being with us keep watching NewsClick