However, yesterday, with the additional counting included, my model showed a sub 10% chance of this unlikely event occurring. In a head-scratching moment, I had to investigate what had happened.
I assume my readers are familiar with the online ABC senate results. I also assume you played around with the senate calculator - this is still online. So I will walk you through a scenario in which the Democrats win, and you can judge for yourself how likely it is.
Count 30: Motoring Enthusiasts vs Family First
At this count, AMEP lead FF by 480 votes. Once the remaining declaration votes are counted, this lead will reduce. Luckily, these are all ATL votes, simplifying our analysis. But it is unknown how BTL votes for the other previously excluded candidates will flow. I estimate there to be approx 120,000 votes of other previously excluded candidates, and approx 10,000 of these will be BTL. But will there be 4% of these such votes putting Family First as the very next party? Unlikely.
In order to see what happens after this point, one will need to enter the primary vote percentages into the pre-election Senate calculator. To make it easy for you, here are the group percentages in the order you'll need to input them into the pre-election calculator:
Next manually adjust FF to 0.40% and AMEP to 0.39% (+0.01% and -0.01% respectively) and press calculate.
Looking at the end of Count 40, we can now see just a 439 vote difference between the Sex Party and the Democrats. Although the Democrats are slightly vulnerable to ATL leakage as they have more preference votes than primaries, I would contend the Democrats are more likely to pick up BTL leakage from other parties. So there is a strong chance that if the Democrats make it to count 40, they will just beat the Sex Party to 2nd last.
Again, we can model this by changing DEM to 0.22% (+0.01%) and SXP to 1.00% (-0.01%). Click calculate.
Voila we have a Democrat elected to the Senate.
Likelihood: Low. But not out of the question. Arthur Sinodinos is not yet safe.