Labor and the KAP have done a preference deal in Queensland, for both the House and the Senate. This is me thinking out loud.
Labor was always going to recommend preferencing KAP ahead of LNP, so I don't think that direction of the deal is news. More interesting is how many KAP voters will follow their how-to-vote cards. Ballpark figures plucked from my memory because I can't be bothered researching this: in ALP-GRN contests, a change in a Liberal HTV can see preference flows to ALP go from 75% to 35%. In ALP-LNP contests, a change in the Green HTV sees preference flows to ALP go from 80% to 76%. I vaguely recall Family First putting ALP ahead of LNP on one of the HTV's once and instead of a ~65% flow to the LNP it was ~35%.
So, there's a range of possibilities of how obedient KAP voters will be to their HTV cards. (I think in the Qld state election they recommended "just vote 1", but if someone happens to look this up and corrects me, that'll be interesting to know....) I think that a reasonable guess is that the "default" position is a 60% flow to LNP, but that with the preference deal that'll fall to 40%.
Now, the impression I get from the very occasional stories in the Courier-Mail on the topic is that the KAP is struggling at an organisational level, and so the last state election might be their high watermark. In 2013 they're also battling against the PUP, who'd be rather less likely to recommend preferences to Labor....
Still, if you map the state results onto federal electorates, the seats of Herbert and Dawson (in and around Townsville) _really_ stand out, with possible KAP votes of 20+% in seats that are marginal LNP. If they poll 10% federally, then the preference deal might be sending a couple of extra points of the 2PP vote to the ALP, bringing the effective LNP margins there to about zero. To a lesser extent, the deal might also shore up Labor in their seat of Capricornia, which is looking more vulnerable as Labor's polling generally seems to dip.
I don't think I'm willing to bet on Labor in Herbert/Dawson on the basis of what I've outlined above, but if they _do_ manage to pick up one or both of those seats by a slim margin, then I'll do my best to claim it as a prediction.
I have no idea how to even try projecting this. For a little while Centrebet had a market on KAP winning a Senate seat, but a) they pulled the market after the preference deal was announced, and b) would enough people who knew what they were doing be betting on it anyway? I certainly wouldn't trust myself....
In 2010, on primaries, the LNP won 2.90 quotas, ALP 2.06, GRN 0.89, and there was the usual array of little parties below 0.25. About the only thing I'm confident of in 2013 is that the GRN vote will fall, and that KAP/PUP will combine to at least 0.5 quotas. But whether the KAP/PUP voters will be taking votes of the Shooters and Fishers et al., or a taking chunk out of the majors, I have no idea. It seems plausible to me that the ALP vote will fall below 2 quotas, but I don't have a feel for how likely that is. If it falls substantially, then we could see the second ALP candidate fall behind KAP after the latter harvests the right-wing micros and PUP, and then the ALP preferences would elect KAP, and we'd probably end up with 3 LNP 2 ALP 1 KAP. (Edit: That makes no sense. Luckily no-one read this.)
If the ALP keep their noses above 2 quotas, then they won't have much surplus to transfer to KAP and the deal won't mean too much, but it still seems quite plausible to me that the combined left-wing (if you'll allow me to describe the ALP as such) vote will struggle to get to 3 quotas, and the ALP/KAP preference deal will make it just that bit harder for the GRN to be elected.
Probably I should wait for the Senate tickets to be finalised and then let everyone punch numbers into Antony's calculator.