Statistics? Or Lies and Damn Lies? You Choose

Now that we have 15 days of voting returns in the amalgamation polls, it is tempting to apply some statistical analysis to the patterns over time. The conventional wisdom is that postal voting returns taper off after the first week or so - and indeed, this pattern is evident from the returns so far - but not to the extent that might have been expected.

The picture is slightly complicated by the clear downward effect of weekends (the data points after the gaps show the cumulative effect of the two weekends so far, one of which had three days), so the picture will be clearer by, say, Tuesday of next week.

However, the 'line of best fit' seems to indicate that any tapering-off has either ended or, at least, greatly diminished. On the basis of these trend lines:

- Kwinana should exceed 50% on Monday 2nd February;
- East Fremantle should exceed 50% by Tuesday 3rd February; and
- South Perth should exceed 50% by Friday 6th February

Clearly Kwinana and East Fremantle can be quietly confident of getting the required votes for the poll to be binding - and I hope noisily celebrating when they do.

South Perth is going to be very tight - but I hope to be celebrating with them when the votes delivered to the South Perth Civic Centre tip them over the line.