« Money and mental health | Main | A heriditary ruling class »

August 19, 2006



There's another good reason as well. It would promote a more rational approach to coercive retail practices as well. A short course in the logic of bookmaking would teach punters to avoid Yankees - or indeed all accumulative stakes. It would urge them to abandon bravado and favour each-way bets (bookies hate them). In fact, given the emergence of betting exchanges, such teaching would probably decimate the whole poxy industry of bookmaking (I'm not bitter!).

It would also teach kids that price determines the quality of information we receive. The 30p soaraway Sun gives you 'form' but not the really valuable info like the speed recording of particular horses. Indeed, they may learn that the competitive nature of the mass media results in the dissemination of misleading information.

And this understanding of how things are packaged would also mean we would all buy less tat from places like Ikea. Which would have deflationary consequences and would probably result in mass unemployment....

...oh dear...

Backword Dave

"You know I'm born to lose/And gambling's for fools/But that's the way I like it baby/I don't wanna live for ever"

My reading of Kahnemann was that even people who knew statistics (like post grad economists) still only used that ability in certain contexts. Gambling may be a fun way to learn maths (though surely it's going to seem arbitrary and unfair), but part of the reason we teach basic maths to younger children isn't so that they know pi or whatever, but so they learn basic congnitive skills. Young kids have a much more animist view of causality than adults; they tend to believe that certain people are lucky, it's a combination of brain development and experience that disabuses them of this, not teaching.

Anyway, kids do have rudimentary gambling: they play board games -- you may not stake anything on rolling a double six in Monopoly, but surely everyone picks up that's there's nothing involved by pure chance. And, of course, two di (two dice? I can never remember that one) yield a normal distribution curve of possible scores. Some things may be better learned by casual messing about rather than in a classroom.

Paulie does anyone give "the speed recording of particular horses"? And is that information valuable? Horse racing is on variable terrain, which seems to affect form a lot. Also, surely not all races are flat out all the way. Horses with better finishing speed, sprinters essentially, are unlikely to be the ones holding course records (from what I know of distance running and cycling). Also, the point of the Sun is that it may be cheap to the consumer but this is offset by bulk sales. I thought the Sun had the best form guides (compared to the, ah, broadsheets) because it has deep pockets. And, final point, I promise, surely its information is (partly) market led? That's what the punters want.


There's a good reason why people wrongly use yankees. They fail to see that probabilities quickly multiply to give a very low probability - 4 5-1 shots have only a 7 in 1000 chance of all winning.
They commit the anchoring heuristic. A similar thing might explain explain the failure to complete big construction projects on time. Each aspect of the project might have a very high chance of completion, but the product of high probabilities leaves a low probability.

james higham

Yet again you have it right. As a former educator, I was all for this but got into trouble over it. But it's a hell of a lot of fun.


"we are merely preparing them to work for the Guardian": well, if it's good enough for the KGB....
Taking your excellent suggestion seriously -when I was at school, it would certainly have made sense to teach us less Euclid and more probability. But one reservation: the elementary accounts of probability must have been the worst-written maths I've ever read. I think the problem is that probability is applied maths: you have to keep translating back and forward from the reality to the ideal model. This calls for clear thinking and expression about both, and about their relation to each other. That's a tester that's too hard for many mathematicians.


" does anyone give "the speed recording of particular horses"? And is that information valuable?"

You'd be surprised how much data is compiled. And plenty on speeds over particular goings. Remember, on the exchanges, value over a part of the race is worth knowing.

A few punters I know who are more serious than I am have pointed out to me that speed is so obviously the most important data that they can't understand why it is so widely ignored - and whan you think about it, it should be.

There's a fair bit of detailed intelligence that is for sale - and some of it more useful than the form guides in the papers. I looked at them for a while, but it's a full time job just processing it. Info from the course on the day is the most valuable to exchange punters, I'm told, particularly in early indications on 'steamers' - the way odds are shortening.


You might want to look at 'Racetrack Betting' Asch and Quandt. Although it is USA based it does offer some very interesting insights specifically on how "little" most people gamble on very short odds (e.g.1 to 20) & yet how effective a winning strategy that is. As they ruefully note however there are few bragging moments with winning on those odds. And that in turn is a hat tip to bahavioural finance


"...how "little" most people gamble on very short odds (e.g.1 to 20)..."

Too true. I saw a study a few years ago (someone at Cork Uni I think?) in which it was shown that £1 on every horse that left on one particular very short s/p over a whole season would have yeilded a profit - something you'd think would be highly improbable given the usual 17% advantage that bookies build in to their calculations. The study pointed to relatively small losses on very short odds in general.

Horses that are very short odds with the bookies would seem to be very good value on the exchanges. I've yet to test this theory though....


Getting back to the point about teaching probability through gambling,

1. I don't think kids would actually find it fun. You certainly don't find any kids who gravitate towards it on their own.
2. Gambling is only like science if it's done rigorously, with an understanding of the principles involved, otherwise it's like a lot of junk academic science that's currently spewed out, done only to waste time.
3. Gambling would teach about probabilities and all evidence, in all the sciences, is ultimately probabilistic. The problem is that you have so much of human activity based on probability (The gambling known as business looks with austere disfavor upon the business known as gambling. - Ambrose Bierce), why choose this particularly worthless and possibly addictive activity to highlight? I'm not saying it should be banned but it's role should be small (Would you have kids play 3D video games to learn about physics?). Great point about students learning that truth comes from authority starting from school, btw.

4. No kids are going to learn the more advanced stuff in school. I didn't have my first class in probability till I was in my second year of my engineering degree.

Of course, the fact that probability and statistics are not taught in high school is a perfect example of the complete failure of the education system, and public education in particular.


I hate you

The comments to this entry are closed.

blogs I like

Blog powered by Typepad