As we went into elections yesterday, BJP had enjoyed a clear and distinct lead in terms of seats on MAPi – Micro Analytics Projections (INTELLIGENCE) – our indigenously developed data-tech platform to analyze elections bottom-up from the polling booths. But as the voting progressed, BJP’s advantage began to wean away, bit by bit. We believe there are five primary reasons for this trend reversal:


  • The rural-to-urban voter turnout was not in favour of BJP. Especially Bangalore City saw extremely low voter turnout indicating urban Hindu voters’ apathy for BJP candidates. This could have a negative impact on many marginal seats in urban areas. What this also meant was that the silent poor, rural voter came out in larger number showing some sort of a gratitude to Siddaramaiah’s welfare schemes.
  • JDS underperformance in at least a dozen seats of their stronghold areas due to counter-consolidation of non-Vokkaliga voters meant that Congress kept surging. Possibly, the BJP strategy of having a virtual alliance with JDS was not so productive after all.
  • Extremely high turnout in minority areas (possibly due to a massive campaign by PM Modi and Yogi) created a skew unfavourable for the BJP. Take for example the case of Basavakalyan in Bidar district; BJP was ahead going into elections yesterday, but the Muslim turnout yesterday was above 86% against an overall turnout of 64.56% which put BJP behind INC. Similarly, in Gulbarga North, the overall turnout was merely 53.6% but Muslim areas saw a turnout of over 74%. Comparable patterns were reported from the coastal belt too. Our predictive models had not accounted for such a large difference.
  • BJP’s famed booth outreach program of Panna Pramukhs was not as effective as in UP and Gujarat. Take for example, K.R. Pura and Hebbal seats in Bangalore where Congress workers were more active at the polling booth level in distribution of money and goodies whereas BJP Karyakartas were mostly dependent on middle-class voters turning out on their own due to the Modi factor.
  • The BSY faction and the “original BJP” faction did come out to play despite all efforts by Modi-Shah to paper over the differences. For instance, in Gadag and Haveri districts, those leaders and workers who were in KJP earlier formed a separate grouping known as “the Brigade” and were working against the official candidates in many seats on polling day.


Despite all these negatives, MAPi –Election tracker is still giving a slight overall edge to the BJP. Let me make it absolutely clear that we have not conducted any pre-poll or exit poll in the state, but have simply tracked all the 224 assembly segments to arrive at these projections. Therefore, we are not giving out any vote-share percentages to any political party. This is a new way of tracking elections which we are planning to deploy across India in all the 543 parliamentary seats in the run-up to the next Lok Sabha elections of 2019.


If you see these numbers, you will notice that after the elections yesterday, Congress has gained leads in some 21 seats while BJP has lost leads in some 9 seats as compared to the trends before the election. JDS too has lost its lead positions in 4 seats. The other 8 leads for Congress have come from the swing seats. Based on these trends the following scenarios can unfold on the counting day.


  1. BJP’s optimal performance would be in the 90’s, so a simple majority can be virtually ruled out now (although elections can surprise even the best placed mathematical models by stretching the means)
  2. Congress  party’s weakness is that it has too many (57) low-margin leads and if BJP can snatch even a few of those, the party can cross the 100-seat mark and easily form the government.
  3. BJP’s big worry continues to be Bangalore where neither the party put up strong desirable faces nor were the voting populace willing to forget this and vote in the name of Modi. MAPi shows that BJP is ahead only in 11 seats out of 28 and 2 of those are vulnerable too.
  4. The most likely scenario as per MAPi Election Tracker is that both the national parties could end up with around 90 seats, thereby needing the regional party, JDS support to form a government.
  5. 94.2% of all elections since 2000 have produced nearly clear mandates for one or the other party. Therefore, Congress party’s chance of crossing the 100-seat mark is much higher because of the election day trends.

You can track all the 224 assembly seats of Karnataka here