Thoughts On Exit Polls and Results

Exit polls, one would expect, are a little bit easier than opinion polls. The only question you ask is "Whom did you vote for?" and you expect one answer, the name of a party. You then add up the numbers. So what can go wrong?

Plenty, it seems, because the last few times, the margin of error has been fairly big. For one, it all depends on the sampling method, and representativeness of the sample. With experts doing the sampling, usually they do try and do it right.

Even if you get the percentage of votes right, sometimes it is tricky to predict who will win the seat. Indian contests usually involve at least three to four contenders (serious ones). The votes get divided, and the winner (almost) never gets 50% of the vote. Therefore, if one party is at 31% and another at 29%, it's hard to predict (given unavoidable errors) who will actually win the seat. This can happen over many seats that are closely fought.

But if there is a massive swing against a ruling party or FOR a new party, it is probably easier to catch that in the air, apart from the poll data. Like many expect this time around.

Interesting to see what happens on the 16th. Swing, wave, et.al.- yes, or no?


1 comment:

Diamond Head said...

regardless of the seat an entry pole is same as the exit pole .. maybe I am confused

XLRI Doctoral Colloquium 2024

 Pics from a well-organised event.  With TAS Vijayaraghavan, my classmate and XLRI prof., and Dr. Prashant, host. My talk on Building Academ...

These Were Liked a Lot