Polls

5Forty3 Polls have extremely high success rate as compared to all other polls in India. For instance, since 2013, we have managed to project almost every election accurately including the 2014 national elections, the 2015 Bihar elections and last year's Assam elections. As of today there are no other psephological organizations in India who have such precision polling methodologies to understand the political trajectories with such depth.

The fundamental reason for our high success rate in an industry marred by deeply inherent biases and wrongfully deployed methodologies is our six dimensional approach to polling. These 6 dimensions include indigenously developed polling tools like VWISM and RSSI that have given us unprecedented success in analyzing Indian elections.
ADc 6 Sigma VWISM Near 100% Record RSSI Frames Data hostory Survey Design

Current Polls

Our final round of pre-polls in Uttar Pradesh and Punjab are underway currently with a targeted sample size of 6150 and 3426 respectively. As of now we are constantly updating the progress of our polls on a regular basis to give a clear idea about how the exercise is being conducted meticulously. In the near term 5forty3 aims to create platforms that would enable real-time data mapping of our poll numbers as and when the survey progresses till then we present a brief summary of the raw data.

Punjab Final Pre-Poll
Target Districts Achieved till now Target AC Segments Achieved till now Target poll booths Achieved till now Target sample size Achieved till now
13 13 27 27 133 133 3426 3229

Updated till 31st January 6 PM

100%
Uttar Pradesh Final Pre-Poll
Target Districts Achieved till now Target AC Segments Achieved till now Target poll booths Achieved till now Target sample size Achieved till now
26 26 63 63 252 252 6150 6007

Updated till 5th February 6 PM

100%

Summary:

At this point of time these are merely raw, unmapped, unweighted numbers that have not yet been regressed through our statistical models and are therefore prone to the usual noise minus the signals, but they do provide some early peek into the way a specific geography may be shaping up electorally. In clear layman’s terms, these raw numbers have the potential to show the direction of the “hawa” which is a crucial factor in Indian elections. Strictly speaking, statistically, raw unweighted numbers generally have an error margin of 6-7% whereas our final set of numbers that we announce after rigorous modelling have an error margin of 2-3% – that makes all the difference in the world.

Punjab:

Our Punjab poll survey concluded today in record time of just over 8 days. We now have the full data and now begins the actual work of adding weightages using VWISM (Voter Weightage Index Sampling Methodology) templates for different sub-regions of Punjab and data parsing using our statistical regressive models which would give us the final derivatives. Raw numbers are still mostly inconclusive so we are hoping that we would be in a position to give a definitive analysis of Punjab in the next 48 to 72 hours.
Bottom-line:: This is still quite a tough state to predict and we need to be extra cautious before actual voting on Saturday

[Updated: Jan-31-2017, 6 PM]

Uttar Pradesh:

After full pre-poll survey of UP which was slightly delayed due to a 2 day break in between for data auditing that enabled us to cover all the swing areas, BJP continues to show strong performance. Although this is just raw data, it is giving us enough indication of a strong "saffron hawa" in Uttar Pradesh. Of course, a deeper analysis of the raw data based on our regressive models will show us the extent of BJP's strength and vulnerabilities. What we are also studying is the extent of BSP's meltdown along with the impact of the SP-Congress alliance. Our full analysis should be out on or before Thursday, 9th February.

[Updated: Feb-05-2017, 6 PM]

 

Be an Angel and support India’s best pollsters.

Punjab:

After almost 3/4th of Punjab having been polled, it still looks like a 3 horse race as far as the raw numbers go. There are all tidings of a fight to finish political battle in the land of the 5 rivers. These kind of raw numbers are a nightmare scenario for a pollster to take a clear position, but we do hope there would be more clarity once all the weightages are accounted for and a thorough regression cycle is adhered to. It is indeed quite a task now for us to discover the electoral signal from all the noise. In fact, we believe in the need to be cautious about Punjab’s trajectory at least till post-poll numbers confirm the trends. Punjab is undoubtedly the trickiest part of the 2017 election season.

[Updated: Jan-29-2017, 10 AM]

Uttar Pradesh:

With 2/3rd of our survey having been completed in Uttar Pradesh and 3/4th of the geographies having been covered, we are now entering the crucial last part of our ground polling activity wherein we also make-up for any statistical anomalies till now (like over sampling of particular demographics etc.). Raw numbers still show a pretty strong tiding towards BJP while most opposition parties seem to be struggling, especially BSP. As of now raw numbers seem to be pretty consistent across all regions and all days with expected statistical deviations. Really, the only crucial aspect now is to model these numbers before we can pronounce complete clarity.
Bottom-line: UP looks like a far clearer state at least at the outset based on raw numbers, unlike Punjab.

[Updated: Jan-31-2017, 6 PM]

Punjab:

Close to midpoint of our survey activity, three aspects are becoming apparent; A] AAP seems to have somewhat halted its meltdown (as was the case in our first 2 surveys because the party was seemingly in a free fall) and maybe once again coming back into the reckoning, B] SAD-BJP is losing steam and becoming increasingly listless or at least that is what the initial raw numbers are showing and C] Congress seems to maintain its pole position and maybe even improving its leads

[Updated: Jan-26-2017, 7 PM]

Uttar Pradesh:

BJP is consolidating its lead... no that would be an understatement, Saffron leads have opened up big gaps in many places. In fact, what we are seeing is reminiscent of April 2014 in terms of raw numbers. The alliance seems to be falling flat and if these numbers hold (a big if?), all the media hoopla about Akhilesh Yadav will meet the same fate as the "planted" stories against the Modi wave of 2014. Maya's collapse also seems to continue with the same pace as 2014 and may be the biggest story of this election. There are of course three caveats on these numbers - 1] Almost 3 quarters of our survey is yet to be completed, 2] These are raw, unmapped and unweighted numbers that have not undergone the rigours of our statistical models as yet and 3] In recent times we have seen that many last moment changes do happen just a week before the actual election, so exit poll numbers will give a much clearer picture. Having added that caveat, at the present moment one simply cannot rule out a massive saffron sweep of Uttar Pradesh beyond the 90's RJB peak!

[Updated: Jan-29-2017, 10 AM]

Along with all the caveats of unweighted, raw numbers, comes the fact that these are very early days in our survey with less than 10% of the activity being recorded till now. As of now, BJP seems to be still in the pole position although with only a slight lead as compared to runaway leads of our past 2 pre-poll surveys. Not surprisingly, BSP is no longer the main challenger and it is the SP-Congress alliance that has now taken the leads over Maya’s party (again unlike our past surveys). CM Akhilesh Yadav seems to be ahead in personal popularity ratings too, but it is not a runaway lead as portrayed by the media, at least until now. The only question that we need to ask for now is, are we classically underestimating Maya and BSP? Historically, poll surveys manage to under-capture BSP support and is that being repeated in 2017? Although to be fair to our own surveys, we have always managed to give a fair estimate of Mayawati.

[Updated: Jan-26-2017, 4 PM]

Ready to go?

Select "Logout" below if you are ready
to end your current session.