National Predictions

Our predictions, based on simulations, have yielded the following results. We will continue to run these simulations till election day:


The numbers our models came up with are very different than the NDTV poll numbers due to several factors:

• The alliances in many states are still evolving or will evolve after the elections
• Jaya from TN will align with the BJP sooner than later
• TDP and Chiru's party in AP together will create a small ruffle in AP, of which Chiru's PRP will gain 10 seats and TDP will gain only about 3 seats.
• In UP, the mess will continue and the third front will continue to gain steam. BSP is also in this third front bucket.
• Mayawati will play kingmaker this time. She will have a large contingent of seats.
• Despite having less seats in their kitty, NDA will be joined by Mayawati to keep out Congress out of power.
• LK Advani will be the next PM of India (42% chance).

We will continue to refine our simulation models as time goes, so watch this space.

How we predict

What methods are used to predict the outcomes of the Indian Lok Sabha Elections 2009?

We are using several sophisticated analytical techniques to come up with the predictions. Some of these techniques have been used and proved to be very accurate in the US Presidential elections in 2004 and 2008. Here are the techniques we use:

• Triangulation from several polls- both online and offline
• Efficient markets hypothesis
• Trend Adjustments
• Regression
• Simulation methods (Monte Carlo simulations)
• Events Impact Analysis (e.g. impact of Varun's speech or shoe throwing at P Chidambaram)
• Projections from other polls including NDTV, IBN and even the astrologers

We use any and all data and use it in our simulation models, which we run over and over.

जय हिंद!!