Lay the field strategy

[JollyGreen]

13:50 Uttoxeter

If you needed proof that some people betting in play do not have a ScoobyDoo about horses then look at this screen shot. My Nosy Rosy was backed into 1.78 admittedly for small stakes but was officially beaten 15 lengths with the winner strolling home! This horse travelled well with about 3 hurdles left but never made the lead!

Nosy Rosy.jpg

[stevequal]

Jolly, very interesting response. I haven't really watched many races, as I am mainly a pre-race trader, but I'll try and look out for some of the characteristics you've highlighted.

My Nosy Rosy was backed into 1.78 admittedly for small stakes but was officially beaten 15 lengths with the winner strolling home! This horse travelled well with about 3 hurdles left but never made the lead!

I think this a good example of where if I had personally been watching the race then I wouldn't have known if 1.78 was good or bad value. I would have to look at similar races and see what prices horses are usually available on the exchange for at that point in the race. There are two ways I could look at it. I could apply the characteristics you highlighted (and more as I am sure there are many) and make a instinctive judgement or I could look at previous race data and judge from that whether the horse has value. The ultimate goal would be both combined. It's probably a good exercise to separate distinctive and statistical judgements.

Here is something for you to try if you want to. Look at the lower grade races like 5 6 on the flat and jumps... Have a look at the last price of the favourite pre-race and then look at how it performs in running. You will find a lot where the price drifts a lot in running which can prove profitable.

So the idea here is that if a horse prefers soft ground, but is favourite on a heavy ground race lay it at the start and look for a potential drift?

[Zenyatta]

Jolly, for people that don't 'have a ScoobyDoo about horses', the In-Play odds on offer are on average astonishingly accurate.

As I said, I've tried booking, laying multiple runners at very low odds (or conversely, dutching and backing multiple runners at high odds), and let the spread-sheets run on for weeks or even months. They never do better than break-even. I'm sure that many many people have tried this and could tell you the same thing. That shows that the In-Play odds are, on average, incredibly accurate.

As I understand it from the comments, the people that are profitable In-play (e.g, yourself and Peter Le), don't just set-it-and-forget-it, they have figured out a way to respond to what is happening in-play - and that is probably the hard part.

[Iron]

Jolly, for people that don't 'have a ScoobyDoo about horses', the In-Play odds on offer are on average astonishingly accurate.

A while back, I obtained some data that included BSPs, maximum in-play odds and minimum in-play odds for thousands of horses.

I looked at what the outcome would have been if I had blindly backed every horse in-play at 1.0 above its BSP odds (where possible), and was slightly surprised to discover that the method wouldn't have been profitable. I'd thought that market over-reactions might have made the approach profitable.

As an aside, I looked at what would have happened if I'd laid every horse at BSP, and the market turned out to be frighteningly accurate - I'd have made a profit of about 0.000001%!

Jeff

[JollyGreen]

As I understand it from the comments, the people that are profitable In-play (e.g, yourself and Peter Le), don't just set-it-and-forget-it, they have figured out a way to respond to what is happening in-play - and that is probably the hard part.

I can only speak for myself when I say this requires a manual approach and you must be able to both read the race and understand the horses in the race. I know it can be performed using automation because PeterLe has proved that to me.

Define accuracy?

[steven1976]

My definition of accurate in regards to BF horses in play would mean the book % is close to 100%. Which actually means little in terms of a horses price is correct.

[JollyGreen]

My definition of accurate in regards to BF horses in play would mean the book % is close to 100%. Which actually means little in terms of a horses price is correct.

Exactly Steve, spot on my friend!

So the latest chart showing My Rosy Nosy shows a massive error in the market but as you point out the book may well be sitting within the normal boundaries of an in running market and will be seen as accurate.

I find a lot of people make a similar mistake when they relate price movement pre-race with the result of the race. You hear TV pundits saying "The market got that one right!" when a steamer wins or "The market got that wrong" when a drifter wins. There is no reliable correlation between pre-race moves and the result of the race, if that was true then we would simply lay the drifters, back the steamers and just wait to collect our profits. As you all know, that isn't happening any time soon!

[Iron]

Hi JG

As you are probably aware, a few years ago Betfair published some research which showed that, over several thousand races, betting drifters and laying steamers (defined in terms of an x percent move in odds between y hours before the off and the start of the race) resulted in a very healthy profit.

That may no longer be the case, although I doubt the market has stopped creating value bets as a result of people blindly following the crowd and getting over-excited about a 'sure thing'.

I'm also suspicious when I see certain patterns of price action. See viewtopic.php?f=5&t=6690&p=68729#p68729 for example. Red Lion Rock drifted and drifted and then crashed in. That made me wonder whether someone knew something, and there was some manipulation was afoot (and I wasn't at all surprised when I checked the result and found that Red Lion Rock had won).

I'm not sure it's possible to bet and lay with any confidence based on what happens in the market, and it may be that my perceptions are swayed by confirmation bias, but sometimes it feels like there is almost an inevitability about a horse winning or running a very good race...

Jeff

There is no reliable correlation between pre-race moves and the result of the race, if that was true then we would simply lay the drifters, back the steamers and just wait to collect our profits. As you all know, that isn't happening any time soon!

[JollyGreen]

I'm sure if you took enough data, analysed it to the nth degree and then extrapolated the bits that suit you could do the same thing now. Could you predict it in real time and make it pay? I doubt it.

[Iron]

Fair point.

We don't know Betfair arrived at those stats - whether they took the first x thousand races that their database gave them, or if they looked at various random samples until they found one that suited their PR needs...

Would you agree with me that there is such a thing as market behaviour that hints at what connections know? For example, if a horse at 15.0 BSP has nearly as much money matched against it as a horse at 5.0 BSP, and has been steaming in steadily all day, then would you agree that that's likely to be due to people in the know rather than the mugs who back favourites all day long?

Jeff

I'm sure if you took enough data, analysed it to the nth degree and then extrapolated the bits that suit you could do the same thing now. Could you predict it in real time and make it pay? I doubt it.

[LinusP]

Fair point.


Would you agree with me that there is such a thing as market behaviour that hints at what connections know? For example, if a horse at 15.0 BSP has nearly as much money matched against it as a horse at 5.0 BSP, and has been steaming in steadily all day, then would you agree that that's likely to be due to people in the know rather than the mugs who back favourites all day long?

This is normally due to a horse being tipped by one of the big guys (Hugh Taylor,Pricewise) The increase in volume is due to people like me arbing the tip when it is first released.

[steven1976]

When there is as much traded on a 15/1 shot at the off as a 5/1 shot i would say it is just a little battle between a couple of traders.

[LinusP]

14:15 Wolves, "Here now and why" was tipped by Hugh this morning, its currently trading at 5.1 so should have about 20% of the money matched on it but as its a fav that could be +30% (Dark lane in the next wolves race has 30%) but it actually has 48% of the money matched.