Whether forecasts...

[harry_rag]

...are a good bet or not, that is the question! Separate thread so as not to clutter up the L15 thread and so those who (understandably) don't want to wade through hypothetical waffle can look away now. Basically following on from a previous discussion with @Zilzalian, all figures based on Saturday's 1:15 at Ascot.

• The overround on the win market (mid morning) was 109.7% with 365 (at SP it was 113.3%). Nothing unusual there; we know that some people can make a profit from backing winners and you can reduce the disadvantage by shopping around for best price or betting at BFSP.
• On the forecasts it was 135.1% so 365 are taking a lot more margin out of that market than the straight win. Stating the obvious to say that makes it much harder to win betting on forecasts than it is betting to win. You could find 3 runners that you rate well clear of the field and get great value about them in the win market. Back them in the forecast market and you're either getting worse value or no value whatsoever.
• The three horses put up for forecast purposes were priced at 3.5, 34 and 71 to win. If you remove the bookies margin (making the simplistic assumption of it being applied equally to all runners) that implies 365 had the true odds at 3.84, 37.3 and 77.89.
• Compare that for the forecasts involving the 3.5 and 34 shot; the prices offered were 76.78 and 126.97 whereas the "true" prices with no margin were 103.75 and 171.57. If they applied the same lower margin to the forecasts as they do to the win market you'd have been getting a less mean 94.57 and 156.38.
• Out of curiosity I looked at the prices you'd get backing a horse to win by dutching all 7 relevant forecasts and also to finish second by doing the same. The 33/1 shot paid 32.35 win and 20.66 to come 2nd while the 5/2 shot paid 2.84 win and 2.92 second. Suggesting that more margin is taken out of the fancied runners than the outsiders and (perhaps) more is taken out of the 2nd place element than the win.

So much for the numbers themselves, I could do with a brew before I post my (relatively brief) conclusions. Hopefully no-one's breath is excessively bated!

[harry_rag]

Conclusions then, and going a bit deeper into this on a specific race has only confirmed the view I'd already formed, basically I don't like the bet and I think it's poor value. Why voluntarily take on a market with such massive margins added when there are better alternatives available? I accept I'm unlikely to change Zil's views based on years of actual experience and his own firmly held beliefs on the subject with a purely theoretical angle but the numbers are hard to argue with.

I doubt anyone could make a profit by playing this market. You'd need a huge sample to prove it either way but I'd be surprised if anyone could show 1000 races where they'd played the forecasts and were in front. If they were I'd still suspect they could have got a better return by taking a different approach. I can't get away from the fact that you're either getting worse value or non at all compared to just backing the win singles.

Bottom line is I believe anyone would be better off just backing the singles rather than splitting a small stake off to play the forecasts. If you want to maximise the work you've done in identifying 2 or 3 runners that are clear of the field (and have a tilt at a bigger return) then I'd say do doubles across 2 races rather than fc in each race. Doubles multiply your edge where you've got one whereas the forecasts reduce it. Yes you can hit a big win every now and again but the odds are almost certain to be poor and the cross-race doubles are the better way to pursue that.

I suppose you can make the "fun" argument and there's going to be satisfaction in calling the first two home correctly but, for me, what's the point if you believe the bet is demonstrably poor value.

That concludes the case for the prosecution. Defense views welcome!

[Zilzalian]

3 hours ago, harry_rag said:

Conclusions then, and going a bit deeper into this on a specific race has only confirmed the view I'd already formed, basically I don't like the bet and I think it's poor value. Why voluntarily take on a market with such massive margins added when there are better alternatives available? I accept I'm unlikely to change Zil's views based on years of actual experience and his own firmly held beliefs on the subject with a purely theoretical angle but the numbers are hard to argue with.

I doubt anyone could make a profit by playing this market. You'd need a huge sample to prove it either way but I'd be surprised if anyone could show 1000 races where they'd played the forecasts and were in front. If they were I'd still suspect they could have got a better return by taking a different approach. I can't get away from the fact that you're either getting worse value or non at all compared to just backing the win singles.

Bottom line is I believe anyone would be better off just backing the singles rather than splitting a small stake off to play the forecasts. If you want to maximise the work you've done in identifying 2 or 3 runners that are clear of the field (and have a tilt at a bigger return) then I'd say do doubles across 2 races rather than fc in each race. Doubles multiply your edge where you've got one whereas the forecasts reduce it. Yes you can hit a big win every now and again but the odds are almost certain to be poor and the cross-race doubles are the better way to pursue that.

I suppose you can make the "fun" argument and there's going to be satisfaction in calling the first two home correctly but, for me, what's the point if you believe the bet is demonstrably poor value.

That concludes the case for the prosecution. Defense views welcome!

Defense is going to be a long one but i will give it a go when i have a spare few days. 😁

[MCLARKE]

You've distracted me from my work on my shares portfolio, it is suffering from survivorship bias !

I agree that trying to overcome a 35% overround in the long term is almost impossible. I am happy to make a net 3% profit on my exchange win bets, this is probably equivalent to a 5% loss at traditional bookie odds (my lucky 15 bets are showing a 4% loss if backed as singles).

The only case for the defence is if there is a factor in a race that is likely to favour a certain section of the field and this isn't reflected in the odds.

As an example in the Balmoral Handicap on Saturday this was my comment

"The key stat is the draw with those drawn 1 - 11 producing 8 winners out of the last 9 with a profit of 81 points. Those drawn higher have 1 winner from 97 runners and a loss of 70 points." 

The first 7 home out of a field of 20 were drawn 4 5 2 7 3 9 8.

The odds of the first 3 home were 25/1 40/1 and 7/1.

The Trifecta (tote bet) paid £9,194, the tricast (bookies calculation) paid £4,250 .

I'm sure there is a way of calculating whether this was a good pay out but that is for another time or someone else ?

Obviously only 1 race but one conclusion maybe is that when the odds are high it is better to bet with the tote than the bookies.

It has piqued my interest and I may look at the big handicaps to see if there is a potential edge.

 

 

[Sporting Sam]

Great stuff and an interesting read.
My thoughts provoked by this read led to me getting a forecast up last night amid  a bit of lateral thinking based alongside the waste of backing two horses to win the same race. I back two in a race not because I think they’ve  got an equal chance of winning but because they both  have chances of winning for very different reasons. 
But in the event of one of them being the strong favourite, if you fancy something else then you could well be in forecast country. I fancied an outsider to win and he’d need to overhaul the top horse to do so. In that event a reverse forecast was insurance, rather having a saver on the favourite.
I’m committed to making a few tricast attempts soon as October has historically been the month I land them. I’ve left it a little late but the forecast last night was a confidence booster.  
keep these coming Harry and Mc  and Zilzalian , it’s a cracking read.

Edited October 24, 2023 by Sporting Sam

[Zilzalian]

First thing I will say I love data, I use data for my speed figures but I also recognise that using data in betting has many flaws and you can sometimes get bogged down in it or to be blunt Data can make you one dimensional, how often do we hear the data people say things like “long term” or “ the data shows” I would argue it is what the data doesn't show that is important. I will give an example, before the Lingfield all weather track was re-layed for some reason number 1 on the card won an inordinate amount of times. Yes you will argue that data would show you that but the thing is it wont unless your looking for it, in other words it is something you notice then you can either act on it or research it further.

 

The best data you can use is a notebook, that is jotted down spur of the moment stuff but even that is useless unless you use it and part of that use is your memory, another thing number crunching doesn't do is give you instant in the moment information where you don't have time to do research on what is happening, I will give you an example from a breeders cup meeting, my main bet was Invasor in the main race. After two races it dawned on me that the inside stall was favoured I made a fortune betting stall 1 (coincidental with the previous Lingfield example), Invasor in the classic was drawn 11 so a decision had to be made. I had already backed it to decent money with the intention of going in again on the night so that was dead money. So what do you do? Leave well alone or trust your judgement of new information? On this occasion I stuck to my guns but covered the 1 stall which was 20/1 if I remember correctly.

So to make a long story even longer my point is this, data is very useful but has its limitations in betting on horses. So on to forecasts and tricasts, I have said elsewhere and it this approach has been contested by others that my speed figures produce an order of winning chances so I maximise the work I put in rating a race simply by betting win, F/C and T/C, now if I did this in every race yes I would soon be bankrupt so I have to be selective and proportionate, again I have said elsewhere that what I do is lets use £20 quid as an example, you can't put £20 quid on the win bit £20 on the f/c and £20 on the tricast that would be suicidal, now here is the main point, I set conditions based on “is it worth it?” Nothing is set in stone but if any of the 3 selections is less than 8/1 I rarely do the forecast but may do a small tricast this could be all ways or if the selection is a shortie I will go with that to win and the other 2 reversed i.e.

 

Top rated, 2^nd and 3^rd and top rated 3^rd and 2nd (2 bets) so the bet might look like this.

 

£18 win horse top rated A, £1 T/C ABC and a £1 tricast ACB = £20 quid.

 

Now if all three are good prices ie above 8/1 I would break the £20 quid down to thus.

 

£14.00 win, horse top rated 50p all ways forecast (£3) 50p all ways tricast (£3)

or if there are huge prices-

£15.50 win, horse top rated 50p all ways forecast (£3) 25p all ways tricast (£1.50)

[Zilzalian]

My response is all about why i back forecasts and tricasts. Micheal and Harry will tell you all the reasons why you shouldn't and for most punters they are correct so don't do the lottery because it will take you about 45 million years to win it.

[MCLARKE]

Exactly why I don't do the lottery (unless it's a rollover and then it could be a "value" bet)

To be fair I have given reasons both for and against tricasts and I may in fact try a forecast on the big handicaps

[harry_rag]

17 hours ago, MCLARKE said:

I am happy to make a net 3% profit on my exchange win bets

Always surprised when you say that as my instinctive reaction (flying in the face of the fact that profit is profit and hard to achieve at all) is surprise at how small a return you're getting. You'd need a very large sample size (which I suspect you've got) to prove that a 3% edge was down to "skill" rather than possibly luck. I tend to hanker for a double digit edge to satisfy myself that a betting approach is viable. Whilst that's certainly ambitious at least it takes a smaller sample size to prove it's sustainable!

17 hours ago, MCLARKE said:

The only case for the defence is if there is a factor in a race that is likely to favour a certain section of the field and this isn't reflected in the odds.

Yeah, two questions that you'd all have a better idea of than me. Has the draw bias been adequately factored into the win price and, potentially, do the bookies factor it further into the forecast returns as well (assuming they know which way people are likely to bet based on that factor). But with the benefit of a 135% book I'm not sure it really matters. 

Great explanation from @Zilzalian of his take on it and I think it just ratifies where we'd got to before; agreement up to a point but with a degree of disagreement beyond that. I'm never going to convince him with pure theory that he should stop doing those bets (and have no wish to do so) and he's never going to persuade me that it's the optimum way for him to back his selections without an unfeasible amount of data. At least anyone who's taken the time to read it all can take both sides on board and reach a conclusion in terms of their own betting.

And having got all that out of my system I can concentrate on trying to eke some profit out of tonight's Champions League games! 

[LEE-GRAYS]

I got 1 up top of league at moment 😂😂😂

[LEE-GRAYS]

2 minutes ago, LEE-GRAYS said:

I got 1 up top of league at moment 😂😂😂

Actually I made a f**k up did @Zilzalian nap was to was supposed to be tricast hit wrong button 😂😂

[harry_rag]

1 hour ago, LEE-GRAYS said:

I got 1 up top of league at moment

Only another 999 to go in order to be admitted into evidence! 

[Zilzalian]

5 hours ago, harry_rag said:

Always surprised when you say that as my instinctive reaction (flying in the face of the fact that profit is profit and hard to achieve at all) is surprise at how small a return you're getting. You'd need a very large sample size (which I suspect you've got) to prove that a 3% edge was down to "skill" rather than possibly luck. I tend to hanker for a double digit edge to satisfy myself that a betting approach is viable. Whilst that's certainly ambitious at least it takes a smaller sample size to prove it's sustainable!

Yeah, two questions that you'd all have a better idea of than me. Has the draw bias been adequately factored into the win price and, potentially, do the bookies factor it further into the forecast returns as well (assuming they know which way people are likely to bet based on that factor). But with the benefit of a 135% book I'm not sure it really matters. 

Great explanation from @Zilzalian of his take on it and I think it just ratifies where we'd got to before; agreement up to a point but with a degree of disagreement beyond that. I'm never going to convince him with pure theory that he should stop doing those bets (and have no wish to do so) and he's never going to persuade me that it's the optimum way for him to back his selections without an unfeasible amount of data. At least anyone who's taken the time to read it all can take both sides on board and reach a conclusion in terms of their own betting.

And having got all that out of my system I can concentrate on trying to eke some profit out of tonight's Champions League games! 

Yes the bookies do factor in the draw Horses and dogs for f/c and t/c and obviously singles. The unfeasible amount of data are probably the wrong words because all my selections are based on the data i have accumulated through study put simply the selections are not random but an accumulation of everything i know and experience, what i am saying is that you can't win money on forecasts just using available data that is available to all who care to mine it. the 135% book is irrelevant because it assumes you are going to cover all or most of the options. What are the percentages if you do x or y or z or any other criteria (how often does something happen in reality as opposed to theoretically? bottom line is, do i make money from singles? the answer to that is yes do i make money from forecasts? the answer to that is yes, do i make money from tricasts? the answer to that is no. but i like to have many fingers in many pies especially away from the serious betting as in singles bets. 

Advice for @LEE-GRAYS if you think the fav will win don't bother with the forecast, pick races with dodgy favs n second favs and back decent prices to small stakes for a bit of fun. 

[MCLARKE]

5 hours ago, harry_rag said:

Always surprised when you say that as my instinctive reaction (flying in the face of the fact that profit is profit and hard to achieve at all) is surprise at how small a return you're getting. You'd need a very large sample size (which I suspect you've got) to prove that a 3% edge was down to "skill" rather than possibly luck. I tend to hanker for a double digit edge to satisfy myself that a betting approach is viable. Whilst that's certainly ambitious at least it takes a smaller sample size to prove it's sustainable!

The advantage of betting on the exchanges is that you only need to have an edge of 2% or more to make a profit.

The disadvantage is that this is where all the serious punters bet because they can not bet with the traditional bookmakers. The competition is of a much higher quality than the typical mug punter. It is the premier league of horse race betting. I am competing with the Man Citys of this world.

Those betting with with the traditional bookies are competing with the equivalent of Sutton United. 

League 3 is where I began many years ago and I eventually clawed my way up to the premier league !

[harry_rag]

15 minutes ago, Zilzalian said:

the 135% book is irrelevant because it assumes you are going to cover all or most of the options.

It’s not irrelevant because it means that in the vast majority of races every forecast price will be way less than the true odds for that pair of horses to finish first and second. I’ll accept that it’s not necessarily unbeatable but it’s like setting the bar really low for a limbo dance or really high for a pole vault. Most people aren’t going to be able to get over/under it. Let’s agree it’s a bet for specialists who know what they’re doing. You’ve got to have the right sort of edge to begin with if you’re going to give it a go.

Obviously I’d have no chance even allowing for my spectacular form in the “bit of fun” comps! My edges lie elsewhere. But as I don’t get enough chance to talk technical on my specialist subjects I end up forcing my way in here as a guest lecturer on the mechanics of various racing bets! What do we fancy covering next? 

[harry_rag]

@MCLARKE I’m guessing there are very different dynamics at play in your BF markets than in mine (anytime goalscorer). Most of my profit is coming from the bookies for now but it feels like I’d struggle to get much more money on to take full advantage of any edge. I’ve been thinking about scaling up the stakes more by betting on the exchange even if I have to take slightly shorter odds (or same odds but minus commission). It does look like there are people who simply copy the best bookies prices off Oddschecker and lay them on the exchange.

If I’ve got any confidence in the edge I’m currently getting it feels like I should either be increasing my stakes or the volume of bets, probably a bit of both.

[Zilzalian]

33 minutes ago, harry_rag said:

It’s not irrelevant because it means that in the vast majority of races every forecast price will be way less than the true odds for that pair of horses to finish first and second. I’ll accept that it’s not necessarily unbeatable but it’s like setting the bar really low for a limbo dance or really high for a pole vault. Most people aren’t going to be able to get over/under it. Let’s agree it’s a bet for specialists who know what they’re doing. You’ve got to have the right sort of edge to begin with if you’re going to give it a go.

Obviously I’d have no chance even allowing for my spectacular form in the “bit of fun” comps! My edges lie elsewhere. But as I don’t get enough chance to talk technical on my specialist subjects I end up forcing my way in here as a guest lecturer on the mechanics of various racing bets! What do we fancy covering next? 

Force your way in all you want it is always good to have a bit of banter its what i like about this forum, exchanging ideas is good for the pocket much of what i have learnt over the years has been through listening to the oldies when i was young. The best tip i ever got was from an 80 year old, he said "follow Mr Charlton" and he was dead right i made a fortune off just his 1 piece of advice. Sadly Charlton retires at season end which is good for me because he hasn't been operating at a decent level for a couple of years now.

[harry_rag]

Yes it was a wise and venerable old soul who put me onto the notion of the Bet Builder which I’d previously dismissed as being purely mug bet fodder. Can’t remember his name!

[Bang on]

It's an interesting thread, and I can see both points of view. 

It's not for me though as has been pointed out already, the 'value' lies with the bookmaker (ask anyone high up the bookmaker food chain). Just like buying a lottery ticket (although with factors/influences that can help solve the equation), it is the dream of the big win that draws people in. No doubt, being selective about which races have some candidates rated clear of the rest will improve strike rate and potential returns. Not something I would get deeply involved with, good luck to those who do.

[MCLARKE]

The SP overround for the Balmoral handicap was 131%. 

I have calculated what I believe are the true odds for selecting the first 3 in the correct order and it is 17,698 to 1 compared to the tricast of 4,250 to 1.

I think I'll go back to my original conclusion, it's a terrible bet !

[Zilzalian]

2 hours ago, MCLARKE said:

The SP overround for the Balmoral handicap was 131%. 

I have calculated what I believe are the true odds for selecting the first 3 in the correct order and it is 17,698 to 1 compared to the tricast of 4,250 to 1.

I think I'll go back to my original conclusion, it's a terrible bet !

Yes it is a terrible bet at the odds but you have to work with what is available and so if you get the tricast it is a good bet to small stakes, I think it was you that said any winner is a good bet. To me every fav 2nd fav 3rd fav up to 8/1 is a bad bet especially in a say 12 runner handicap where all runners technically have and equal chance so is 2/1f a good bet in a 12 runner handicap? should it not be 12/1 give or take a bit of margin for bookies?

[Torque]

20 minutes ago, Zilzalian said:

I think it was you that said any winner is a good bet.

I'm struggling to imagine anybody on this thread saying the above. Some of my best bets have been losers and some of my worst have been winners. A winner at 10 to 1 on that should have been 5 to 1 on is a bad bet and that will show if bets like that keep being taken. Conversely a loser at 10 to 1 that should have been 5 to 1 is a good bet and that will show if bets like that keep being taken.

[MCLARKE]

I think I may have quoted that a 1/10 winner is better than a 100/1 loser. 

[Zilzalian]

13 hours ago, MCLARKE said:

I think I may have quoted that a 1/10 winner is better than a 100/1 loser. 

My apologies, yes that was the correct quote. not sure i agree with Torque tho, bottom line is if it wins it is a good bet, if it loses it doesn't matter how good you think your bet was, it won you nothing. I put what i thought was a great bet on yesterday it drifted from 20/1 out to 50's made it an even better bet, ran a great race to finish second but i won nothing. What i would concede is that there are good bets and bad bets to have value wise if they win. The bottom line is quite simple, do the good bets out weigh the bad ones and make a profit over the long term?

[MCLARKE]

You are right, value is key but very difficult to quantify