Double Blind

The current odds at 4:50pm Pacific time indicate that it is now a race between Bob Rae and Stephane Dion. Will be interesting to see what happens – Ignatieff and Rae’s relations are beyond repair, which suggests that if either Ignatieff or Rae is forced to drop out that they will be supporting one of Kennedy or Dion. We’ll find out in a couple days what happens.

Ignatieff: 24.4%
Rae: 37.7%
Dion: 35.3%
Kennedy: 13.3%
All other candidates: 1.3%

As usual, divide by 1.1 to get the “true” odds.

Wow. All I can say is that Sacha Baron Cohen is one of the most bravest actors, period. He’s had more than half a decade to practise his technique and it really shows in this movie. I’m just curious how they skirted around most of the legal aspects of filming the movie – i.e. getting those release forms signed by the participants he interacts with.

While I can’t claim that this movie was the funniest of all times (or even close to it), it did have quite a few moments. Setting up the situations with the large groups of people and finding enough suckers to say “funny” things on camera must have been a heck of a logistical challenge.

I’m glad I didn’t pay to see the movie in the theatres though – it would have been cash wasted.

I finally got around to watching AI (the movie) . It was the last movie that Stanley Kubrick helped to form (he died before filming began) and I recall when the movie was released in 2001 it was touted to be a huge blockbuster that will be remembered just as ET.

The movie lasted nearly two and a half hours, which can be clearly divided into three phases. The first third was the most “realistic” of the movie, where the viewer tries to immerse into the new world that the director is trying to present. It also sets the stage for the rest of the movie, which descends into a rather dystopian “Escape from New York” tone. I thought when the middle of the movie began (the transition is perfectly clear even as you watch the movie) that my sense of ‘immersion’ was dramatically reduced. The final third of the movie deals with how the main character (the guy that said “I see dead people”) deals with his own existence.

A few things that stood out was the acting – all the robots (“mechas”) did not blink in the movie which must have taken a bit of training for those scenes that lasted more than 15 seconds. Also, the movie asks basic questions about ethics and “what is life” when you have man-made machines that are nearly life-like, although obviously not quite human. In a rather eerie way, it ties into what Rodney Brooks was trying to say in his book Flesh and Machines – what exactly is it “to be”? If you create a perfect machine that passes the Turing test and can’t be distinguished from normal people, does that machine get granted the rights normally associated with humans? Or will it always be in servitude to humans?

These are questions which have been asked many times before in literature, including Do Androids Dream of Electric Sheep?, which the movie Blade Runner was derived from.

The final scene of the movie deals with the main character getting most of what he wanted, but this was all synthetic, generated by the “aliens”. Despite the fact that all of humanity gets wiped out and the only successor race are the machines created by humans, the machines still try to excavate old cities on the earth for more information on who created them. Do they have their own identity crisis as well?

It’s tough to say. AI asks some interesting questions, but there is something difficult about the delivery of the movie that would have turned it into a 2001: A Space Odyssey classic that I’m sure Spielberg was aiming for.

Here is my risk profile in the Liberal Leadership race. I do not intend to hedge the position (this is done by selling futures on Tradesports since bets on Pinnacle are buy-side only):

Essentially what this graphic means is that I have Dion at +925 (9.8%) and Kennedy at +2000 (4.8%) that will pay out (profits) of $3286/$3360 if Dion/Kennedy win. If they do not win, then I am down $491 ($32 of that was due to a 10% deposit bonus). I received incredibly good odds and I think the bookmakers across the Atlantic Ocean (in the UK) were not doing their homework on this one.

If I win, one can always say “why didn’t you bet more?”. The reason is because one’s wagering should always be constrained by the amount of available capital and your expectations of being correct. Even if you had a high degree of probability being correct, you should never ever wager your entire stack of cash on such an event since you would be wiped out in an adverse outcome.

This is governed by a simple edict of game theory – assuming you have a biased coin (landing heads 0.5+x of the time) that pays off 1:1 on any bet, how much of your capital should you wager on any bet? The answer is 2x. So if the coin is biased 55% to land on heads, your optimal wager is to bet 10% of your capital. If the coin is biased 75% of the time, your optimal wager is to bet 50% of your capital. Intuitively this should make sense – if the coin is not very biased then you lose more capital than gain if you bet too high a fraction. If you bet too little of your capital, your variance will be lower but your maximum expectation will be lowered as a result.

If I lose, I still think I made a great wager. It’s much easier to say it wasn’t a good wager in hindsight, but that’s what predicting the future is all about – you don’t have the power of hindsight. I received great value. I’ll still have plenty of capital left to continue making great wagers and eventually some of them will hit.

Fortunately from what has remained a hobby has materialized into a situation where one can make substantial sums of money and still remain within a proper risk profile. As a result, I’ve turned a three-digit figure into a mid four-digit over the process of a couple years. Not bad for recreational work.

As of 1:00pm today, Bob Rae is actually the favourite to win the race. There has been a lot of rampant distribution of Ignatieff. Kennedy has also increased significantly over the past couple days, presumably due to the Quebec nation issue. Dion and Rae have also seen an increase, although not as dramatic as Kennedy.

Ignatieff: 33.3%
Rae: 33.6%
Dion: 28.8%
Kennedy: 13.0%
All others: 1.3%

As usual, these are market odds and don’t add to 100%; the remainder is the bookmaker’s spread, so divide by 1.1 if you wish to reflect “real” odds.

Update: What a difference four hours make. There must be some backroom dealings that are going on which is fuelling speculation on the betting markets. Current odds at 5:38pm:

Ignatieff: 29.1%
Rae: 35.8%
Dion: 30.8%
Kennedy: 15.0%
All others: 1.3%

Kennedy has picked up some odds at the expense of Ignatieff:

Odds, 4:00pm Eastern Time on November 28:
Ignatieff: 50%
Rae: 27.1%
Dion: 22.8%
Kennedy: 9.1%
All others: 1.3%

Again, to find “real” market odds, divide the above figures by 1.1 – the remainder is the bookmaker’s spread. What probably happened is somebody put $500 on Kennedy (at a profit of $10,000 if he wins) and the corresponding odds of other candidates change. The market isn’t very liquid.

OK, there is some serious egg on my face. The prediction I made in the previous article (Greens to get 7-8% of the vote) did not materialize at all; my prediction of an easy (greater than 5%) Liberal victory, however, was correct. Here were the results and the 2006 general election results in brackets:

Liberal – 34.9% [40.1%; -5.2%]
Green – 25.9% [5.5%; +20.4%]
Conservative – 24.4% [29.9%; -5.5%]
NDP – 14.1% [23.8%; -9.7%]

This performance by the Green party was much better than I was ever expecting – even better than the most optimistic of scenarios. It’s tough to extrapolate how a by-election result will convert into a general election result, but it will be very tough to argue that this 25.9% result was nothing more than spectacular.

The Green candidate (party leader Elizabeth May) appeared to suck support away from all the other main parties – the Liberals and Conservatives had their support decline 5-6%, while the NDP had their support decline 10%. This performance is really bad news for the NDP.

This by-election will have very curious implications for the next federal one, especially in British Columbia where the NDP are hanging onto their federal seats with the skin of their teeth – if they have two or three percent of their vote split by the Greens, it can only translate into more Liberal or Conservative seats.

Keep in mind, however, that a by-election is just a by-election.

Right now as I’m writing, the members are “debating” the motion without notice to adjourn the legislature (presumably until Spring). The NDP members are stalling and their speeches are quite imaginative – right now Leonard Krog is giving a dissertation on the history of parliament.

Each member has 30 minutes to discuss this motion. In theory, the Legislature closes at 9pm, so they will be going beyond this limit. My knowledge of parliamentary procedure is not strong enough, so I’m not sure what will happen at the 9pm date – do they resume debate on this motion tomorrow or do they keep going until the wee hours of the night?

You can see the standing orders of the legislature and judge for yourself. Section 3 is pretty clear.

Update: The procedural trick that the BC Liberal house leader (Mike de Jong) used to ram the motion through is to adjourn the legislature until 9:05am on Tuesday. The reason why 9:05am is because they calculated that if each NDP member used their full allotment of time, they would run out of speakers before the 12:00pm recess. If the NDP makes it to the 12:00 recess then they will likely be able to get one more question period in. Right now the NDP members are going through a very agonizing effort in explaining why this motion should be rejected.

It doesn’t look like the NDP will get an afternoon question period unless if they can filibuster this motion for at least 12 hours. They can technically do this by each member speaking for half an hour (they have 33 members), but it doesn’t look like that they were prepared to do this.

Now I know why the speaker of the house gets paid extra – he has to listen to this. I wonder if he has a good seat cushion. He’ll need it.

The vote passed 266 to 16 in Parliament.

Who’s the winner in all of this? Gilles Duceppe. He’s now completely ready for the next election.

Just look at the carnage he’s left behind – Conservatives divided, Liberals divided. Despite the fact that you had most of the house voting for the motion (including the NDP, who surprisingly had the most to gain by voting no) there were many causalities in the process.

Apparently Gerald Kennedy will be opposing the Quebec nation motion. It’s easy enough for him since he doesn’t even have a seat in the House of Commons, but I think this is a powerful move that will differentiate him from the other three candidates. It will win a lot of mind-share from the majority of Canadians that think the “nation” business is garbage (which is shared between people of all political parties, something I find odd).

That said, this announcement is enough of a catalyst for me to take a financial position. I have placed US$150 on him at the odds of +2000, or 20:1 winning the leadership race. This means that my total risk profile is evenly split between Dion and Kennedy – if either win, I’ll stand to gain about US$3300. If not, I will be down US$493. I am receiving about 7:1 odds that either Kennedy or Dion will win the contest which I think is really, really, really good value for my money. Even if neither win the contest, if the true odds of them winning were 4:1, it still means I have a statistical expectation of +US$450 which is really good.

My thesis is that Rae will be gone after the second ballot. Dion will either support Kennedy or Kennedy will support Dion. Either way, it’s going to be Ignatieff vs. Dion/Kennedy and all that will matter is how the Rae delegates vote. Since the Rae delegates are more left of centre, you can be sure either Dion or Kennedy (especially Kennedy) will receive the majority of former Rae voters.

I put my money where my mouth is. Will any other Liberals do the same?