I read an article about Wisdom of Crowds and EMH and betting on horses

i don't agree with nick

EMH?

Think Leslie Crowther.

Efficient Market Hypothesis.

The notion is that just before the off, all the info that will be available is out there, the market has absorbed the all money from all the competitors and is betting to around 103% and therefore is very accurate.

And?

if you backed every horse of a certain fixed price in every race(13,000 of them p.a.) then you would lose. 9-1 shots at BSP win 1 in 10 races(max). But you don't have to back every one, that's your edge.

Int terr esting

if you backed every horse of a certain fixed price in every race(13,000 of them p.a.) then you would lose. 9-1 shots at BSP win 1 in 10 races(max). But you don't have to back every one, that's your edge.


That's saying that in the races you choose to back in, your opinion is better than that of the combined market.  Maybe it is.  That's the debate.

long term winners' views (the 2% or whatever) must have a better opinion than the 98% of long term losers. Unless as you say they always beat the price.

The gist was that you cannot win in the long run by backing at BSP (or ISP obv).  The only way is to take advantage of early overnight or morning prices.


can you tell me which bookmaker will take a bet the night before.

but I see your point nick, if you want to win long term don't back at BSP. But then that's always been the case with betfair and with conventional bookmakers before that.

luckyme, and that's now the problem, night before prices are not really 'in existence'

OP -- The gist was that you cannot win in the long run by backing at BSP (or ISP obv).  The only way is to take advantage of early overnight or morning prices.

This must be rubbish given there are pro-punters in Tote-only jurisdictions who cannot, by definition, take overnight, morning or any other sort of prices.

Not all BSPs are equally efficient. In the very large sample of BSPs there will be "accurate" prices, reflecting true probability, "over compensated" prices, lower than true probability and "under compensated" prices, higher than true probablity. Over the large sample you can't win, because the outcome will regress to the mean, but you can if you're good at finding find the favourable outliers.
The key, (as it always is), is having the skill to know the true price and, (again as ever), you won't know how right you are till the day of reckoning.
I still have the instinct to look for early prices, liquidity allowing. Once the firms stopped being viable, I found it extremely difficult coming to terms with the different dynamic on here. Early price value doesn't present itself in the same way as an old fashioned bookie's rick. All too often that market "wisdom" is in play in way before post time and predicting which way prices will go is so much harder. I'm still stuggling even now as I can't shake off the "yips" of wanting to get on, only to find I'm staring at a horrible slide come post time, and of course, that usually (not always), means I was wrong to begin with, and far more often than I ever used to be.

EMH in all its forms discredited many years ago by David Dreman, et al. It's a pile of nonsense.

Look at some of the people that have won elections, that would make me dispute the wisdom of crowds.

There are 3 forms within the EMH.  Horse racing markets would reflect the weak form, hence allowing the possibility of beating the market.

Not an endorsement, in any way, of the EMH.

If that where true then it would be saying that  only horses that get gambled inwards win horse races.
And if it stayed the same or drifted it cant win, Both these statments are nonsense ofcourse.
I would guess that most of the 2% of winners or thats what they say is the precentage could be rubbish too maybe dont even bet to win they could bet to place or dutch betting.
But the fact is no one knows what will win the race before its run in general terms so we have to guess no one can argue with that.
I would say being very selective with the races you look in is far more important  first of all and only bet in ones you have a very strong feeling of winning then and only then have you got to worry is the price right.
Then once you have done this for while and see how your doing percentage of winners wise can you possibly even think of value prices.
  which will all be differant and you still will get value wrongmore than right for sure.So to think value is every thing and only thing is absolute bollocks.
If you walk in a shop and ask for 20 **** and he hands you two packets and tells you were giving two for one today, You dont walk out thinking what a deal you walk out your brain saying something isnt right should i even take one.
Its the same with value betting you think what a price then instantly your brain goes there something wrong here and usually is too.

I have a record of all my bets where I record the price I got and the BSP of the horse.

If I had backed all my 9328 bets so far at BSP, I would have an expected an £19,000 loss, -1.38% on turnover.

If we take commission out of it, then even with the expected 2475.72 winners,  I have an expected profit of £0.00.

So by backing at BSP and having commission to pay means you are on for an expected loss.

Looking at my bets where I beat the BSP, a proft is expected and so it proves, my value bets bring in the money.
Looks at my bets where I don't beat the BSP, a loss is expected and again a heavy loss is suffered.

At the moment my value bets outweight my bad value bets and I am in profit.

The best thing to do is to record the price you get (which you think is value), your stake and aso record the BSP.
From these 3 variables you can work out a whole host of things, expected winners, exepected profit/loss, losing runs/winning runs etc.

If your expected profit from the value bets (where you beat the BSP) outweights the expected loss (where yo don't beat the BSP) then in theory you should be able to make a profit.

My records prove that the BSP is very accurate indeed.

Don't think there's such a thing as "wisdom of crowds", most crowds can turn into an ugly mob, look at the riots last year, the wisdom is in going against the crowd imo.

Of course, one person who completely disproves the theory is George B whose threads make a profit at BSP.

In fact, for 2023 and 2024 combined, he won more at BSP than at his advised prices.

I should add that was for a level £1 stake win only. (The vast majority of his selections are ew.)

Here are the figures for 2023 and 2024.

£1 win adv price +£79.80
£1 win BSP +£126.34
£1 place BSP +£133.31

usually most profitable is to go against the crowd
aye,gets it reet sometimes/most but always overbet
always value not follow the sheep