Validity

One commentator, John Hyde Page, says about last Sunday night's 'debate'

Future generations will mock us - in the same way that we mock those who used to read the entrails of goats, or listen to the Delphic Oracle.

He is talking about the infamous worm the strange little gauge that had everyone's gander up, which fluctuated to allegedly reflect how 100 would-be voters reacted to the stuff that was being trotted out. Much has been written and I don't need to rehearse it again, but I rather agree with him.
I am worried about statistical validity, although my knowledge of such things is limited to having done 1st year Statistics at University. One thing I am clear about, however, is that Page is correct in suggesting that the worm is not to be taken seriously.
But what of the polls that we get once or twice a week in these pre-election days. We are told that some of the most serious ones are based on less than 2000 people being polled nationwide. We are assured this statistically valid. By my count there are 150 electorates each with about 58,000 enrolled electors. So if each electorate is polled then that is less than 20 people per electorate. 20 out of 58,000 doesn't seem to me to be really likely to be predictive.
Even if it a spread of the states then 300 people per State doesn't seem to me to be terribly likely to provide statistical validity.
Ratings
Equally well my previous question about who was watching what (here) was answered in the press today. It said that National Bingo Night was the big winner! The assertion in the Advertiser was that 1.6 million nationally, and 190,000 locally were watching Bingo. But only 1.4 million nationally and 135,000 locally watched channel 9's broadcast (incomplete as it was) though an additional 907,000 & 83,000 watched the debate through ABC. (see the Advertiser p.15 for these confusing figures).
Now forgive me but if I add these two lots of figures together doesn't it actually mean that both locally and nationally more people watched the debate (on 2 & 9) than watched Bingo...2.3 million v 1.6 nationally & 210, 000 v. 190,000 locally. But who cares.
And, as usual where do these ratings figures actually come from, and how do they declare them with such accuracy...and how do they account for people like me who were watching the debate and Australian Idol and SBS simultaneously! I am a true man!

Adelaide City Council elected a mayor and various councillors with less than 7,000 voters casting their lot. A flaw of voluntary voting.. But these people will make decisions which affect the whole fo the State who use the city of Adelaide.

These are musings, I guess, but perhaps we shouldn't believe everything we are told. Or even when we are told we should think about what it actually means!