The high confidence level just means you’re confident that the sample (your measurement) is very close to true distribution. In this case it just means that lots of voters really don’t wish to reveal their voting choice, and lots of voters are really undecided.

The one crucial element that’s missing is a model of the distribution among the undecideds and no-comments. I’ve never seen any pollsters doing it in Malta. For a start, it would require collecting data about the voters’ profiles, not just their voting choices.

At best, Maltese pollsters collect data on location (district) and age, but nothing more. Not that I know of.

I am simply putting forth my points of view on a blog post that highlights the fact that it is very hard to exactly predict the outcome of this particular election due to the silence of a large chunk of the electorate.

If you do not respect my input as I respect other points of views, but would rather resort to mocking me, what can I say except that if it turns you on, I simply feel sorry for you.

Even in 2003, I was feeling ill with tension, because so much was at stake for my sons and for their generation.

Thank you. That was extremely interesting.

http://www.cnbc.com/2016/07/28/political-polls-vs-betting-markets-heres-why-they-conflict.html

Coincidentally I was reading it regarding betting odds in politics. Malta is too small to adopt the same model used abroad. There are too few visible legal bets to be able to asses the accuracy of those odds. 100K loss for these companies just means a smaller Christmas party and besides they always hedge bets anyway.

As for your observation re the silent part? Interestingly Malta is not the only country where we have more people shunning calls. If it happens in bigger countries imagine how much bigger of an impact in Malta.

People may not answer for a number of reasons but I have actually handled surveys and I have kept track of the tone and reaction. Most of those who do not voice their opinion or do not answer are quite aggressive.

I’d say 70% of those who decline to give an answer are like that. They will be curt, rude and practically slam the phone on you.

The ones you speak of try to be nice on the phone. Clearly you do not have first hand experience of this. Just as they may be ashamed of saying they are going to vote for Muscat in Sliema those from the south may be ashamed to say they are going to vote PN.

So basically this is the problem when polls fail us. We end up putting our personal bias in their interpretation.

I therefore disagree with your assessment and while I do think the PN have it very tough I think this time around polls are not very helpful.

The difference between PN and PL is well within the margin of error and the “undecided” are being counted within the sample. That effectively increases the margin of error.

Nobody is considering confidence level that’s being assumed to be 95%. With such a high ‘no comment’ that skews everything. It is going to be close but I’m not giving up yet.

I do believe them because few people know or believe the pollsters do not know who they are. Either that or Malta is screwed.

To give them false hope probably.