FAWS comments on Open Thread: March 2010 - Less Wrong

5 Post author: AdeleneDawner 01 March 2010 09:25AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (658)

You are viewing a single comment's thread. Show more comments above.

Comment author: FAWS 09 March 2010 02:23:39PM *  1 point [-]

Is this supposed to be obvious to people unfamiliar with college basketball in general and that tournament in particular? Gur bqqf (vs V haqrefgnaq gurz pbeerpgyl RQVG: V qvq abg) vzcyl oernx rira cebonovyvgvrf gung nqq hc gb nobhg 0.94, juvpu vzcyvrf gung n obbxznxre bssrevat gubfr bqqf jbhyq ba nirentr ybfr zbarl, ohg gung'f pybfr rabhtu gb abg or erznexnoyl fghcvq sbe n wbheanyvfg.

If the tournament is single elimination knockout, and the figures in brackets are win-loss record against roughly comparable opponents the odds for the sleepers and long-shots seem insanely good. South Florida in particular.

Comment author: Jack 09 March 2010 02:33:45PM *  2 points [-]

Is this supposed to be obvious to people unfamiliar with college basketball in general and that tournament in particular?

Yes

The odds (if I understand the correctly) imply break even probabilities that add up to about 0.94, which implies that a bookmaker offering those odds would on average lose money, but that's close enough to not be remarkably stupid for a journalist.

Rot13: Gel gur zngu ntnva, guvf gvzr pbairegvat sebz bqqf gb senpgvbaf, svefg. Vg nqqf hc gb nobhg .8... V qba'g xabj ubj ybj gung lbhe fgnaqneqf ner sbe wbheanyvfgf gubhtu.

If the tournament is single elimination knockout, and the figures in brackets are win-loss record against roughly comparable opponents the odds for the sleepers and long-shots seem insanely good. South Florida in particular.

This is also true. But the mistake I was thinking of was the first one.

Comment author: FAWS 09 March 2010 02:55:11PM 1 point [-]

Try the math again, this time converting from odds to fractions, first. It adds up to about .8... I don't know how low that your standards are for journalists though.

So betting 1$ at 3-1 means that winning means you get 4$ total, your original bet + your winnings? I had assumend you'd get 3$.

Comment author: rhollerith_dot_com 09 March 2010 06:47:50PM *  1 point [-]

So betting 1$ at 3-1 means that winning means you get 4$ total, your original bet + your winnings? I had assumend you'd get 3$.

To which Robin Z replies, "Yes, you get $4."

This confused me, too, for a while, so let me share with you the fruits of my puzzling.

You do get 3$ over the course of the whole transaction since at the time of the bet, you gave the bookmaker what you would owe him if you lose the bet (namely $1).

In other words, your 1$ bought you both a wager (the expected value of which is 0$ if 3-1 reflects the probability of the bet-upon outcome) and an IOU (whose expected value is 1$ if the bookmaker is perfectly honest and nothing happens to prevent you from redeeming the IOU).

The reason it is traditional for you to pay the bookmaker money when making the bet (the reason, that is, for the IOU) is that you cannot be trusted to pay up if you lose the bet as much as the bookmaker can be trusted to pay up (and simultaneously to redeem the IOU) if you win. Well, also, that way there is no need for you and the bookmaker to get together after the bet-upon event if you lose, which reduces transaction costs.

Comment author: RobinZ 09 March 2010 03:20:50PM 0 points [-]

Yes, you get $4.

Comment author: Jack 09 March 2010 03:33:34PM 0 points [-]

You should Rot13 your second sentence.