Riddle me this

[adhyde]

Seems quite easy to develop an automation strategy that is loss making.

So methinks, let’s just try the complete opposite strategy, that must be profit making, surely?

No.

Is this just the challenge of trading or the sample size is too small or just bad luck?

Is it akin to backing all favourites you lose and laying all favourites you will still lose?

Frustrating !!

[jimibt]

the problem is that converting a losing back based strategy to being lay based could end up worse ironically. this is due to the fact that as a pure STRIKE RATE, it may appear that the lays are going to buck the trend for you. however, due to the odds/liability, it usually ends up presenting an equal or worse outcome.

many folks on here here have had this *leftfield* eureka thought before, only to discover that variance and liability scupper the party. however - one way to look at this as a true leftfield experiment is to exclude the selections that were cherry picked and review the remainder to identify if some of those fall into a sweet spot for the intended scenario.

i'm not saying it NEVER happens, but as you've found, i doubt very much if flipping the bet orientation on a strategy will produce a positive outcome.

[sa7med]

Don't forget that there's a spread, so i would say the true opposite of a back all favs at the current back price strategy would not be to lay all favourites at the current lay price but to lay all favs at the back price. That would be your profitable strategy given that backing all favs is a loser. Of course the issue with this is adverse selection, meaning when you are more likely to get matched on the losers rather than the winners. So that's a whole other thing to consider, actually getting your prices taken.

https://en.wikipedia.org/wiki/Adverse_selection

[rik]

commission and if your taking prices spread??

likely your regular strategy just losing on that long term so reversing wont help

[ShaunWhite]

Seems quite easy to develop an automation strategy that is loss making.
So methinks, let’s just try the complete opposite strategy, that must be profit making, surely?

You won't be the first to think that and definately not the last.

Usually when faced with these "surely" questions I just remember that millions of people have spent 1000s of years studying gambling, from fools to Nobel Prize winners. If making money was as simple as doing the opposite as when losing money...believe me you'd know about it. I'm not saying anyone shouldn't try to have original thoughts, but it's going to take more than simply betting on red instead of betting on black.

Always worth remembering too that a market needs an equal amount won and lost, it's a competition. So it's not that there isn't a magic way to win, it's that there can't be. That doesn't need any maths, it's just logic.

[Derek27]

There's also a timing issue, which is similar to the spread issue. If you back at 3.5 and the price goes through the roof making a heavy loss, to do the opposite you need to lay at 3.5 but get the order in a few seconds earlier, requiring hindsight. Otherwise your lay would just sit at 3.5 while your would-be profit flies away from you.

[Euler]

If your risk and money management is correct you should break even before slippage. If that's not the case then you may have some of those elements wrong.

[jamesg46]

Always worth remembering too that a market needs an equal amount won and lost, it's a competition. So it's not that there isn't a magic way to win, it's that there can't be. That doesn't need any maths, it's just logic.
[/quote]

Love this part of the reply, paradoxical logic... gonna swirl around in my mind for a while is that.

[burntheory]

Seems quite easy to develop an automation strategy that is loss making.
So methinks, let’s just try the complete opposite strategy, that must be profit making, surely?

You won't be the first to think that and definately not the last.

Usually when faced with these "surely" questions I just remember that millions of people have spent 1000s of years studying gambling, from fools to Nobel Prize winners. If making money was as simple as doing the opposite as when losing money...believe me you'd know about it. I'm not saying anyone shouldn't try to have original thoughts, but it's going to take more than simply betting on red instead of betting on black.

Always worth remembering too that a market needs an equal amount won and lost, it's a competition. So it's not that there isn't a magic way to win, it's that there can't be. That doesn't need any maths, it's just logic.

It doesn't follow, though, that both sides of the book - back and lay - are going to perform equally in the long run. I've just run an analysis of 32,000 races from a few years ago to compare P/L figures for the back side of the book with those of the lay. Taking into account BF's practice of doubling the back amount on each runner, the total amount wagered per race was compared with the total amount returned in 'winnings', based in the weighted average price of the winner/s, producing an average loss per race of £8,900 for backers, i.e. the average amount wagered per race was £321.800 while the average amount returned in winnings was £312.900. The lay bets averaged £8,900 profit, obviously. The total loss to backers over this period was over £283 million.

About a year ago I collected another 6 months of data, taking it up to May this year. The sample size was only 3,800 races. This time the average loss per race for the backing book was a fraction over £11,000. The total backing loss was over £42 million. All of the figures quoted above exclude commission from the calculations.

Not sure if this has any relevance to the OP, but I thought it was kind of interesting.

[ShaunWhite]

the average amount returned in winnings was £312.900.

Let me clarify what I mean....

If the average amount returned in winnings was £312,900 then the average amount in loses must also be £312,900. If not then where did the difference come from or go to?

Take a simple market with just two people,. Person A can only win what person B loses. If I win £10 then you can't lose 5 otherwise where did my extra fiver come from? And you can't lose 15 otherwise where did the extra 5 go?

[jamesg46]

Seems quite easy to develop an automation strategy that is loss making.
So methinks, let’s just try the complete opposite strategy, that must be profit making, surely?

You won't be the first to think that and definately not the last.

Usually when faced with these "surely" questions I just remember that millions of people have spent 1000s of years studying gambling, from fools to Nobel Prize winners. If making money was as simple as doing the opposite as when losing money...believe me you'd know about it. I'm not saying anyone shouldn't try to have original thoughts, but it's going to take more than simply betting on red instead of betting on black.

Always worth remembering too that a market needs an equal amount won and lost, it's a competition. So it's not that there isn't a magic way to win, it's that there can't be. That doesn't need any maths, it's just logic.

It doesn't follow, though, that both sides of the book - back and lay - are going to perform equally in the long run. I've just run an analysis of 32,000 races from a few years ago to compare P/L figures for the back side of the book with those of the lay. Taking into account BF's practice of doubling the back amount on each runner, the total amount wagered per race was compared with the total amount returned in 'winnings', based in the weighted average price of the winner/s, producing an average loss per race of £8,900 for backers, i.e. the average amount wagered per race was £321.800 while the average amount returned in winnings was £312.900. The lay bets averaged £8,900 profit, obviously. The total loss to backers over this period was over £283 million.

About a year ago I collected another 6 months of data, taking it up to May this year. The sample size was only 3,800 races. This time the average loss per race for the backing book was a fraction over £11,000. The total backing loss was over £42 million. All of the figures quoted above exclude commission from the calculations.

Not sure if this has any relevance to the OP, but I thought it was kind of interesting.

I'm not sure this has any relevance at all, I mean for this information to be of use wouldn't people need to predict the future? Maybe I'm missing the point but surely you would need to know the outcome beforehand, that goes for the original comment too, how can you do the opposite of an outcome that hasn't yet happend.

Just to add, I'm not trying to be a twat with the above comment, I'm just confused by it.

[burntheory]

the average amount returned in winnings was £312.900.

Let me clarify what I mean....

If the average amount returned in winnings was £312,900 then the average amount in loses must also be £312,900. If not then where did the difference come from or go to?

Take a simple market with just two people,. Person A can only win what person B loses. If I win £10 then you can't lose 5 otherwise where did my extra fiver come from? And you can't lose 15 otherwise where did the extra 5 go?

If we look at any single market in isolation, I 100% agree with you – if the back side of the market wins (or loses) £5,000 then the lay side loses (or wins) £5,000 in return. That’s a given. My observation was simply that when we add all the profits and losses together, the back side lost a lot of money, over £325 million, which was matched exactly by lay profits.

In one sense we could say the market was perfectly balanced, back losses matching lay profits, but in another it could be argued it was unbalanced in that neither side of the market broke even, or close to even, at least in the context of how tight the margins are on BF. In the large sample, the back side lost 2.75% of turnover and in the smaller, more recent sample the loss was just over 4%.

The driver behind the above figures was the average weighted averages prices being overround - just over 104% for the large sample and over 105% for the small. As an aside, I’ve checked my football data and the average WAP is 100%, near as damn it.

None of which should be taken as advice to the OP to flip from backing unprofitable favs to laying them as we all know the problem is a lot more complex than that.

[burntheory]

You won't be the first to think that and definately not the last.

Usually when faced with these "surely" questions I just remember that millions of people have spent 1000s of years studying gambling, from fools to Nobel Prize winners. If making money was as simple as doing the opposite as when losing money...believe me you'd know about it. I'm not saying anyone shouldn't try to have original thoughts, but it's going to take more than simply betting on red instead of betting on black.

Always worth remembering too that a market needs an equal amount won and lost, it's a competition. So it's not that there isn't a magic way to win, it's that there can't be. That doesn't need any maths, it's just logic.

It doesn't follow, though, that both sides of the book - back and lay - are going to perform equally in the long run. I've just run an analysis of 32,000 races from a few years ago to compare P/L figures for the back side of the book with those of the lay. Taking into account BF's practice of doubling the back amount on each runner, the total amount wagered per race was compared with the total amount returned in 'winnings', based in the weighted average price of the winner/s, producing an average loss per race of £8,900 for backers, i.e. the average amount wagered per race was £321.800 while the average amount returned in winnings was £312.900. The lay bets averaged £8,900 profit, obviously. The total loss to backers over this period was over £283 million.

About a year ago I collected another 6 months of data, taking it up to May this year. The sample size was only 3,800 races. This time the average loss per race for the backing book was a fraction over £11,000. The total backing loss was over £42 million. All of the figures quoted above exclude commission from the calculations.

Not sure if this has any relevance to the OP, but I thought it was kind of interesting.

I'm not sure this has any relevance at all, I mean for this information to be of use wouldn't people need to predict the future? Maybe I'm missing the point but surely you would need to know the outcome beforehand, that goes for the original comment too, how can you do the opposite of an outcome that hasn't yet happend.

Just to add, I'm not trying to be a twat with the above comment, I'm just confused by it.

If the observation had any relevance, and I’m not claiming it does, it would be that over the 4-year period analysed, lay money ‘invested’ in the market yielded a positive return, excluding commission, etc whilst back money yielded a significant loss. Does that mean anybody choosing to lay in UK races over that period, rather than back, would have had fewer hurdles to overcome to reach profitability, then my answer is probably yes, but that’s about as much as can be said. I’m a non-trader, so accept my perspective probably doesn’t have much relevance to that discipline.

In terms of predicting the future, isn’t a betting exchange just a bunch of people buying and selling the relative chances of possible outcomes to a future event, whether that be the event itself, taking place at a pre-determined date and time, or market movements in the lead up to that event? Aren’t these people backing themselves to out-perform the market’s assessment of those possible outcomes? By their very nature, betting exchanges are about trying to imagine possible future outcomes and compare them to prevailing market assessments in search of that elusive edge. Understanding some of the underlying dynamics of various markets may or may not help in that quest.

[jamesg46]

I think that is open to interpretation, i'm sure there are people who have systems built on predicting outcomes, and I mean the actual result of the event, on the flip side, my assumption of someone who trades order flow may have an idea of a likely outcome but is ultimately trading what they actually see, this isn't imo "predicting" it's actually happening.

I think I understand a little better now your original comment & i'm sorry if my reply seemed a little "shitty" it wasn't intended that way, I was genuinely trying to understand what you had said... maybe lack of sleep or IQ was my downfall.

[Euler]

The maximum amount that can be won or lost in a market is more or less the spread between the back side of the market and the lay side. Individually and on a strategy basis you are going to have wide variations and periods when the market swings into favour on one side or the other. But in the long term the market is generally efficient.