Monday 20 January 2014


There not yet being iPredict markets on Kim DotCom's yet-to-be-registered political party, I've been reduced to bilateral bets.

I figure there's an 80% chance, or thereabouts, that there will be some new political party with which Kim DotCom is affiliated. 

I also put even odds on that there will be substantial Snowdon GCSB revelations during the election campaign. Our election is coming up this year. The Snowdon releases thus far haven't said much about New Zealand, but he and Greenwald seem to be timing things for best media effect. So why wouldn't they have saved any of the good NZ stuff for the election? 

Conditional on this party's being established, and conditional on there being substantial GCSB stuff coming out during the election, I expect that Kim DotCom will be exceptionally well placed to capitalise on it. And so, in that state of the world, I put even odds on that DotCom gets 5% of the party vote.

So 0.8 * 0.5 * 0.5 = 20% chance that DotCom gets 5%. One chance in five.

@BKDrinkwater and @kiwipollguy reckoned that overstated DotCom's chances. So they each put $80 up against my $20. If some DotCom party gets in at 5%, they each owe me $80. If some DotCom party doesn't get in at 5%, I owe each of them $20. @ClintVSmith did a bit of yapyapyap about how my numbers were wrong, but wouldn't put any money on it

That both Drinkwater and Heffernan took the other side of the bet suggests that I've likely overestimated things somewhere. I'm not sure which parts of the combinatorial they figured I'd overestimated. I suppose iPredict will let me know sooner or later. 

I've reckoned that there's space in the NZ political market for a party that's strong on civil liberties but is also friendly to free markets. The Greens own the "civil liberties, except around smoking tobacco and eating, plus heavy state control of the economy" space. ACT has the free market spot but isn't credible on civil liberties when it comes to surveillance issues: ACT's enthusiastic support of the GCSB power extension means they've no chance on that issue, regardless of any leadership change. No new party, or at least no successor to ACT, can come up in the traditional liberal "free minds, free markets" space while ACT is still there. 

But a civil liberties party that stays basically neutral on economic issues other than favouring whatever makes internet better could do really well. 

Chris Keall over at Ars Technica provides a nice summary of last week's cancelled DotCom party. Kim DotCom planned an album release party and everyone was invited. The Electoral Commission thought it might count as "Treating" - providing of gifts in hopes that those so-gifted would reciprocate with votes. This application of the law doesn't make any darned sense to me. Suppose that, at your party's annual convention, a band played. Would that be treating? How about if you have a hospitality suite with free drinks? What if the party's meeting is held in an exclusive spot that's otherwise hard to get into, and the fee for attending is way less than the venue would normally charge out at? And if the distinction here is whether the event is members-only, surely DotCom could have signed up each and every attendee as a $1 member of his new party and called it members-only too. 

At its core there is an adolescent mindset here: uncritical fandom with all the depth and maturity as a bunch of screaming teenagers at a Justin Bieber concert on the positive side, and on the negative side, a pubescent screech of "you're not the boss of me!" directed at both the current New Zealand government and the US authorities.
I'm not sure that changes much of the reckoning above. Sure, he doesn't have a slate of policies. But, at least for now, that ambiguity is a major advantage for him. Without policies, every fan of DotCom, adolescent or not, can project their idealised policies onto him. Any policy he announces will disappoint some faction. Staying quiet on everything other than surveillance and internet policy gives him flexibility for any post-election negotiations. 

And a "You're not the boss of me" party could crack 5% if we get Greenwald/Snowdon revelations that, hypothetically, the NSA's Waihopi spy base is being used to scrape down and store each and every text message sent within New Zealand, monitor all our phone calls, and that NSA has taps on the cables running into New Zealand. 


  1. I say it's almost a given that the Waihopi base and/or some other method is being used to intercept all text messages and phone calls. That's what a spy base does, and if the US are prepared to do that to their own citizens, then it's reasonably good odds they're prepared to do it to us. I guess someone writing it down as a newspaper article might briefly make it exciting for the general population though.

    I think the bit you're underestimating is that you're calculating the odds for a theoretical internet party with particular policies, but you haven't included a conditional on this being a close election (tending to push people towards the major parties a bit more), nor a conditional on the usual newspaper polls that will proclaim that a vote for an internet party is a wasted vote (that hurt ACT's party vote a lot last time they did that with dodgy polls in Epsom), nor finally have you factored in dotcom's involvement himself - people who like the internet and freedom might not like a convicted criminal who actively facilitates pirating of content being the "owner" of the party. I think those factors would push it down far enough that then the "wasted vote" meme takes over.

  2. I'm not really calculating odds - I'm guessing at odds I'm prepared to put a bit of money on. Agree that DotCom is polarising. And note that my betting on him was at odds. One chance in five he gets there means four in five not. I'd likely have taken the other side of the bet at one in four.

  3. True. But I suspect your guessed odds are perhaps higher than they should be.

    Interesting methodology question. You note that you'll eventually be proved right or wrong by iPredict. Clearly the election isn't a valid proof point - it doesn't give any odds, just a result.

    But, how do you measure success against iPredict? Do you measure the probability today against the probability once (if) iPredict offer a contract? Or would it be correct to say there's more information later, so it's not the same probability measurement (i.e. do you measure a poll 6 months out from the election against the election result, or do you measure only against the last poll before the election - in Australia they do the latter, and all the large polling companies double their sample on the last poll so as to make it as accurate as possible).

    I can see two arguments:
    1. You're predicting the election now based on your guesses at all relevant information, so the correct iPredict number to compare against is the last position before the election
    2. You're predicting based on currently known information only, so the first available iPredict number would be the best comparison point, so long as nothing major and material changes between now and then

    Given that you're explicitly calculating in a likely release of GCSB related information between now and the election, it seems that item 1 would be the measurement point, not item 2, as item 2 could become available before that release of information.

    Anyway, random thoughts for the day.

  4. What you may be overlooking is his policies will be identical to the Greens on these issues. So if you don't want the GCSB in Five Eyes, why would you vote for him not Greens?

    Also the chance of a future Govt led by National or Labour pulling out of Five Eyes is 0.000%, so what will voting for him actually achieve? It will achieve a high profile protest, but I'm not sure 120,000 will want to cats a vote just to protest.

  5. Even if his policies were identical to the Greens' on surveillance, you could expect his party would push for stronger privacy protections and perhaps ground-up review of the GCSB/TICS legislation with meaningful input from tech stakeholders, where the Greens might get distracted by any of whale sanctuaries, banning soft drinks, light rail, pay equity, the environment, chemicals, GMOs, parental leave....

  6. hide your bets dude, stuff coming ,can you keep away from poilitic,at least several months

  7. paulL is in Australia. this is New Zealand