Why TN election opinion polls go wrong

Have you ever noticed how CNN ‘calls’ for a candidate or party in US elections, and it does come true?  Have you ever wondered why (practically) every opinion or exit poll in India generally, and in Tamil Nadu specifically, are so wrong?  No one seems to get it right.  We saw it in the TN Assembly Elections 2011, where opinion polls conducted within days of each other, gave diametrically opposite projections.

So, what exactly is going on?  As I had blogged almost 2 moths ago (has it been so long?), there are four factors affecting the Opinion / Exit polls in India, namely:
  1. Apprehensive Voters – reluctant to ‘truthfully’ disclose their voting intention;
  2. Lack of decent database – vote share of various parties;
  3. Inability to gauge Swing -  voter’s mood or intention;
  4. Inability to gauge ‘anti-incumbency’ factor – amount of ‘Swing’
  5. Apart from above there are other 'minor' considerations:
  • adequate sample size, frequent sampling, concise data collection, and 'bias'.
Of the 4, we cannot do anything about the first, except wait for the prospective voters to change their mind set.  Hence, let us take a look at the second point – lack of decent data base, that is a ‘truthful’ vote share of various parties.

This factor is not a problem in say, USA, where you have 72% of people are registered (to vote) as Democrats or Republicans,  Even in remaining 28%, people identify as Democrats, Republicans or lean-Dem/GOPs.

In Tamil Nadu, it is frequent for a person to hold a membership card in two or more parties!  The most glaring example is the Tenkasi (222) candidate and actor Sarath Kumar who is a leader of his own party, but pulled out an ADMK membership card when challenged during filing of nomination!

But, you may ask why a correct, decent vote share of various parties matter?  Because, the initial dataset, or ‘initialization’ to IT people, is the most crucial factor in deciding the accuracy of an opinion poll projection – just like in many real life situations.  Let us see 2 examples of how wrong initialization can lead to errors, even to the extent of face palm.

The first face palm is mine, and it occurred just about 2 years ago, during the All India General Elections 2014.  While I correctly predicted that the voter would flip the 2009 vote share omelette around – BJP (19%) versus Congress (29%) – the tally of seats was stunning - BJP surged from 116 to 282, and the Congress slumped from 206 to 44.  This occurred because my initial data set was wrong.  I did not load the correct data, which was the BJP’s votes are concentrated in pockets, whereas Congress voters are spread all over India.  Hence for a similar percentage of votes, BJP would gain more seats.

The second example is even more empirical.  The 2 major climate weather prediction models in the World are the USA’s NOAA-NCEP Global Forecast System (GFS), and the European Centre for Medium-Range Weather Forecasts (ECMWF).  Surprisingly the ECMWF model has been more accurate than the GFS model.  After monkeying around, they finally found what was probably wrong with the GFS accuracy – the initialization data.  When GFS was run with ECMWF initialization data, the forecasts became more accurate.

So initial data matters, and is in fact crucial for accurate projections, whether in weather or in elections.

Here is my estimate of the Basic Vote Share (also called vote bank) of the various parties in TN Assembly Elections 2016.

Read More

Google Doodle on Mario Miranda's 90 birthday

One of the joys of the mid to late 70s was the Wednesdays, when my father used to bring the weekend issue of The Illustrated Weekly of India from his office library.  There were two half pages for immediate consumption, and one more for the next day – after returning from school, and no one used to be about.

The last shall remain undisclosed – but if you are really interested just Google for ‘lehigh university illustrated weekly of India’.  The first was the comic strip The Phantom by Lee Falk.  The second was the cartoon by Mario Miranda.

Mario João Carlos do Rosario de Brito Miranda (2 May 1926 – 11 December 2011) born in the then Portuguese Goa was the quintessential foil to the other great cartoonist R.K. Laxman who was also with The Times of India then.

Mario Miranda’s cartoons were full of interactions between various characters.  It was almost like spot-six-differences, with the added pleasure of figuring out the hidden stories.  Seldom would you find a Mario Miranda cartoon without an animal – the ever reliable spotted dog, the black cat or the crow.

So it was real pleasure to once again view a Mario Miranda like cartoon as a Google Doodle on what would have been his 90th birthday.  Here is the doodle, depicting a rainy day in Mumbai.

Have a smile, for Mario Miranda must be busy doodling the Saints – He once said, ‘My earliest victims, the priests, used to hate me. I replaced them with politicians. You will find much in common between them - their size, attitudes, even their sermonising!’

Read More

Google Doodle for Tamil New Year and Vishu

India is really Incredible.  So many subcultures, so many different customs and practices.  Just a week ago, we had Ugadhi and Gudi Padwa – the Telugu, Kannada & Marathi New Year celebrations.  Today, we celebrate the New Year of the Tamils, the Malayalis of Kerala, the Bengalis and others.

So what if Google did not decorate its India home page with a Doodle?  Let us make up our own.  Google had put up doodles for Nowruz or the Parsi New Year on 21 March, and for Songkran or the Thai New Year Festival yesterday.  All it took was half an hour to patch both up and make a doodle for the Tamil New Year and Vishu.

Wish everyone a Happy Vishu and Greetings on the Tamil New Year (தமிழ் புத்தாண்டு வாழ்த்துக்கள்)

Here is the patched up doodle.

Read More

A Boom Boom Ox at Chennai

One of the charming sights used to be the Boom Boom Ox (பூம் பூம் மாடு)) or the nodding bullock in the streets of Madras, now Chennai.  These bullocks are the traditional way of life of a particular nomadic tribal people, who are slowly being encouraged by government and private initiatives to settle down for a better way of life.

Within a couple of decades, the Boom Boom Ox would turn into another urban legend.  So it was with a childish delight I spied one such in my street, today.  A group of us were chatting about – what else, but the hot weather – when the ox trotted by.

I pulled out my mobile to grab a couple of photos, and the bullock owner acquiesced, as he had already taken some money from a passer by to entertain his grandchild, and he posed the Boom Boom Ox for a couple of snaps.  Then came the nodding routines.

First he told the bullock to raise its right hand – and up went the right front leg, and then the left hand too.  Next he asked the ox whether the old man gave him money or not, and bullock nodded, signalling an yes.  Then he asked whether the ox, would like to give the child a kiss.  It nodded yes, came near and kissed the frightened child’s toes..

By this time appeared the neighbourhood killjoy.  You know the type - if he sees a sleeping stray, he will chase it.  If he spies a feline trying to scoot across, he will demonstrate giving her an anxious ten minutes.  He will snaggle a thread to a nest, and tug whenever the crows’ sit down to incubate.

Anyway this pest, just to needle us, started commenting that performing animals are a no-no as per latest Animal Protection laws, and one complaint will land both the bull and its owner in jail.  That’s when the Boom Boom Ox showed its true mettle.

Perhaps it was a discreet tug of the restraining rope, some other unseen signal, or the animal really understood Tamil, especially the 'jail threat' to its master. Anyhow the bull charged, and would have gored the pestering fool, but for the suddenly taut restraint.  To wrap the story up, the fool scooted away, leaving us to enjoy one more chuckle*.


Tags: ,

*the last paragraph was amended on 14 Apr 16 for clarity
Read More

Durmukhi New Year Tharpana Sankalpams

The Tamil New Year of Durmukhi (துர்முகி) occurs on Thursday, 14 Apr 2016, that is the day after tomorrow.  However the actual timing of its ‘birth’ has been calculated as 18:30 IST tomorrow, 13 Apr 2016, and the Tharpanam has to be performed on Wednesday.  The Tharpana Sankalpam Tool has been updated to version 0.7.7 beta and can be used for free to generate a personalized Tharpana Sankalpam for the Durmukhi Year.

What is new in Tharpana Sankalpam Tool 0.7.7 beta for the Durmukhi year?  The new feature is the differentiation between the Druk Panchang and the Vakya Panchang whenever they occur.  For example, the Tamil month of Aani (ஆனி) occurs on 14 Jun 2016 at 23:18 as per Druk Panchang, whereas it is calculated as 02:24 IST on 15 Jun 2016 in Vakya Panchang.  So in this new version, we have differentiated them with the (D) or (V) as applicable along with the dates.

Advanced wishes on the Tamil New Year and Vishu.

Read More

Greetings on Ugadhi, Gudi Padwa 2016

2016 must be the closest Ugadhi (and Gudi Padwa), the New Year’s day for the people of Andhra Pradesh, Karnataka, Maharashtra, Konkan, Manipuri and others, occurs so close to the Tamil New Year (along with the New Year for the other peoples of India).

So, today, 08 Apr 2016 we celebrate Ugadhi, while the Tamil New Year falls on next Thursday, 14 Apr 2016.  As usual, there is no Google Doodle for Ugadhi.  Has never been one.  But Google this year, pushed out a Doodle for the Nuvroz, the Parsi New Year, on 24 Mar 2016, though it appeared only in the Kazakhstan home page.

But still a doodle for a new year is good enough for any other celebrating another.  And as the explanation for the doodle wishes, ‘hope today's freshly-painted Doodle by Alyssa Winans helps start your Nowruz with the beauty of the season to come’.

Here is the Doodle.

Read More