Sunday 19 September 2010

MediaWatch on iPredict

Radio NZ's MediaWatch programme highlighted iPredict's numbers on the Christchurch mayoral race (hit the 4:37 mark.)

The host there is a bit skeptical of iPredict, wondering on what basis traders chose to reverse the odds on Parker post-quake given the absence of polling, and noting that iPredict's numbers aren't a poll but are rather the views of a bunch of betters on whether Anderton or Parker will win.

First off, iPredict's track record is pretty good (hit the iPredict tab, below).

Second, prediction markets are generally better than polls, not worse than them. Folks updated their views on the likely election outcome four hours after the quake. Not immediately after the quake, but rather about the time Parker started demonstrating competence in response to the quake. And that's when the price moved. And the price kept moving as Parker continued demonstrating competence, with traders expecting that voters would be entirely reassessing the candidates' relative merits.

Days after iPredict moved, Twitter started moving. I've been watching the #eqnz hashtag reasonably closely over the last couple weeks. For the first few days, there wasn't much politics. Then folks started talking about how pleased they've been with Parker and how they'd be worried about a mayor who spends half his time in Wellington as a sitting MP. For every tweet along the lines of "I hope people don't forget how terrible Parker was about X", there are a half dozen or more saying things along the lines of "I was going to vote Anderton but now I'm supporting Parker." That's now started to turn a little, with news that Parker's polling well also for the Auckland election, and Christchurch has to put out at least nominal opposition to anything that Auckland likes; some on Twitter also see Parker as being in cahoots with National on the emergency powers front and Twitter's heavily anti the emergency powers legislation. But what I could draw out of Twitter, and out of the letters pages from the local papers, showed a decent shift towards Parker. iPredict's numbers aren't inconsistent with what other signals we've got.

MediaWatch goes further, wondering how it is that Parker's seen such an increase in support; the host there blames coverage that's been equating Parker with Giuliani. I'll call BS. Parker demonstrated no particular skills during his prior run as mayor. Nothing screwed up too badly, no huge scandals, but unease about several smallish issues which incensed a few groups - like bailing out developer David Henderson, the proposed addition of a music conservatory to the Arts Centre, and the expensive new city hall. Then Anderton formally announced his candidacy and started running a really good campaign. I'm no fan of Anderton, but I'd probably have pushed the button for him over Parker given Parker's push for an expensive light rail system for the city. And Parker seemed determined to destroy his own campaign. The feature Press interview with the Mayoress would have put off a fair few Christchurch voters. Everything changed when Parker started post-quake management. Why?

For starters, Anderton's always dodgy position that he could be both Mayor and MP for Wigram was thrust to the fore. Before it wasn't salient - it didn't much matter if he'd let a few balls drop for the year before the Parliamentary elections. Suddenly it mattered a lot to have a full time mayor.

Second, it became a lot more important that the local guy be on good terms with the government. Key does better on that front than Anderton.

Third, it gave Parker a new shot at demonstrating competence; he passed very well.

iPredict may have overshot a bit - there's plenty of chance for Parker to start screwing up again in the coming fortnight. But that's seeming a fair bit less likely. I don't see any reason to go heavily for one or the other at current prices. I'll curse my Anderton-heavy portfolio, but backing out at current prices doesn't seem likely to improve my position.

If MediaWatch's Colin Peacock thinks the Christchurch numbers are out of line, there's a lot of money he can make by going in and buying Anderton at $0.25. I paid a lot more than that for my Anderton contracts (grumble...)

No comments:

Post a Comment