Great dissection, barry. But it applies to an awful lot more “libertarians” than Reynolds.

I can deal with crusty conservatives, with lunar leftoids, even with the odd bovver-boy fascist, but I find it really hard to tolerate these hypocrites. They just can’t admit to themselves that they’ve enabled a bunch of inept criminals who’ve gravely damaged their country and who should be facing a war crimes tribunal.

Advantages: proven lack of ethics, morality and honesty; 90% reliable cheerleader for the American GOP interests, with that 10% unreliability that the sophomoric think of as honesty; the sort of libertarian who understands the need for torture, secrecy and continual war for the strength of the state; the sort of ‘libertarian’ who understands that cronyism shouldn’t be discussed too much, lest it wither in the light; and finally, he’s the sort of ‘libertarian’ who will reliably revert back to ‘government-power/secrecy/torture/continual war is bad’ the moment that a Democrat is elected president of the USA.

In fact, once that happens I could just see him putting a banner on the Economist website, listing the number of days since OBL killed 3,000 Americans, and hasn’t been killed or brought to justice. It has that ‘literally true, but in service to the lie’ attitude that’s made him the premier blogger that he is today.

total expected value = (+£200×2/3 + (-£100×1/3)) = 0

I of course mean

total expected value = (+£200×1/3 + (-£100×2/3)) = 0

It is calculation errors like these which have cemented my reputation today, unfortunately. But the central point is that bookies like to maintain the polite fiction that they are just holding your money in escrow and the transaction is really a wager betwee two gentlemen that is settled after the race ends. In fact once your money goes over that counter, it’s gone.

As a probabilist, you think of it as:

paying £100 with certainty for

{a 1/3 chance of receiving £300 and a 2/3 chance of receiving £0}

total expected value = (-£100×1 + £300×1/3) = 0

But as a bookie, Paddy P thinks of it as having

{a 1/3 chance of winning £200 and a 2/3 chance of losing £100}

total expected value = (+£200×2/3 + (-£100×1/3)) = 0

You’re thinking of it as paying the expected value of a risky payoff, Paddy’s thinking of it as a zero-expected-value contract. This of course never fails to cause confusion.

The general conversion formula is that bookies’ odds of X/Y is equivalent to a probability of Y/(X+Y).

So the book above is equal to 0.56 + 0.33 + 0.14 + 0.13 + 0.09 + 0.04 + 0.01 + 0.005 = 1.175 which is a more reasonable (although still somewhat rich by the standards of horse racing) 17.5% vigorish, assuming negligible probability of someone outside that field winning.

Microsoft redesigns the iPod packaging (Via.)…

Some – but given Micklethwait’s previous form as US editor (and his book with Woodridge), I wouldn’t be confident if he gets the nod for the big job, as I’ve written in the past.