Thought i would write a post on how i go about making my automation rules for the UK horse racing and show how you may be able to maximize your own rules and be able to run them on a wider range of races.
For the purposes of this post i am going to use the well known strategy of Laying The Field (LTF) - if you don't know what this is search for it on BA TV or there are tons of vid's on you tube or you can google it.
Most people will of at some point created this rule with instructions to place a lay bet on all runners just after the off, - for this example where going to say that's a £10 stake @ 1.75 (£7.50 liability)
Applying any single rule to all races will not return a profit long term, so you need to select the races that best suit the rule and hopefully the strike rate will be such that it returns a profit but how can we limit the losses and maximize this further?
As it stands first you are ONLY running the rule on a few carefully selected races and ONLY one of two outcomes can happen - only one horse trades below 1.75 and there for it produces a loss of £7.50 or two or more horses trade below 1.75 and a profit is returned of £2.50, £12.50 or in rare cases £22.50.
By adding a few simple parts to this we can increase the amount of races we can apply it to and also increase the chances of a profit at the end of the race.
Lets now say we include a condition that if 1 horse trades at 2.0 and the price of the net runner is over 10.0 (so looks unlikely that a few will now trade down as low as 1.75) the rule now cancels all these lays @ 1.75 (this alone may increase your overall strike rate & profit) but instead of just leaving it at that what we can do is have the rule do something different due to the new make up of the race so now it places a back bet with a total stake of £7.50 on all the runners priced between 14.0 and 25.0 believing that its unlikely that all will continue to drift and that one or more of them will come back in a few dozen ticks with a condition also added and set to close and green the trade if a total profit reaches £2.50 or more.
Now there are 3 possible outcomes - the same first two as above still apply only one horse trades below 1.75 as it does so one or more of the other runners are trading less than 5 but never get as low as 1.75 so we still take the loss, second option also the same as above a few runners are in contention so all trading around 3s and 1 at a time drop in price with at least two getting matched so we get a profit.
Our 3rd option now is that only one horse looks like trading that low and going on to win so our rule cancels the lays and instead backs the runners between 14.0-25.0 worst case they all continue to drift and £7.50 is lost - this is exactly what we would of lost if the lays where left alone and only the one traded that low but we now have a 2nd bite of the cherry and need just one or more of the runners now backed to come in a hand full of ticks by having the fav bounce back out just for a few seconds which will be enough for the total profit to equal more than £2.50 and close our trade this way.
What all this means is that with a bit of thought and few extra parts applied to your existing rules if it looks likely its going to trigger and end in a losing trade the rule can take advantage of the current circumstances and do something different - in addition the rule can be applied to a wider range of races because the race may not unfold one way but may another giving you chance to still make something so at least 90% of races can now be covered with the above rule.
Many of you probably even have rules from a while ago that just did nt make a profit or close to break even so with a few added bits here and there to cover changing eventualities in-running you could get a completely different set of results.
I de just like to just say the above is just an example please dont go and do it and expect a guaranteed profit - my point is make your rules flexible and able to changes to the condition of the race as it unfolds this will lower your chances of a loss and increase your strike rate and ultimately your profits.
How To Make a Profitable Automation Rule in Guardian
-
convoysur-2
- Posts: 1110
- Joined: Thu Jan 12, 2012 10:00 am
Thanks for posting that dallas its very informative,
marc
marc
I have to oppose this idea.
Simple mathematics say : Any combination of non profitable strategies is non profitable. The outcome can be obtained as
Result_Of_Strategy1 + Result_Of_Strategy2
where
Strategy1 is Lay the field and Cancel if conditions above are satisfied
Strategy2 is Back runners with odds 5-9 if conditions above are satisfied
If one of these two strategies is not profitable itselves, there is no point in creating combination with this strategy, it will be just worse then profitable strategy alone.
So if you have non profitable strategy, keep it forgotten...
Simple mathematics say : Any combination of non profitable strategies is non profitable. The outcome can be obtained as
Result_Of_Strategy1 + Result_Of_Strategy2
where
Strategy1 is Lay the field and Cancel if conditions above are satisfied
Strategy2 is Back runners with odds 5-9 if conditions above are satisfied
If one of these two strategies is not profitable itselves, there is no point in creating combination with this strategy, it will be just worse then profitable strategy alone.
So if you have non profitable strategy, keep it forgotten...
DavDvo you have mis-read/understood the post, i did not say that you can just combine losing stratergies and expect them to now make a profit when combined into one rule, my point was when making a rule dont do it with just one purpose in mind and limit your options instead allow it to be flexable and change to different conditions to maximize the strike rate and overall profitablity.
In my example using LTF was for demontration purposes - forget about it being called LTF and instead call it "All Rounder" Yes its first purpose it to try and lay the field but if the movement of the odds suggest it looks certain its now only going get one horse layed and therefore be a loser this race instead of just allowing this to happen and take the loss have the rule cancel this part and if possible have it take advantage of the new situation and perform a different task based on how the race looks to be shaping up from the make up of the odds - this may still lose but you wont be any worse off than leaving it as it was but will give you extra chances of success so decrease the number of loses you have and increase the winning strike rate.
It does not have to be complex to cover multiple eventuallties, the rule I use the most has just 3 lines (rule types) and only uses 4 different conditions and can apply it to 90% of races each day and think i have had 4-6 losing days the whole of this year across both Flat and NH on everything from Grade 1 to nurseries.
What i did say at the bottom of my post was that some people MAY already have ran a rule that was just short of producing a profit or breaking even and if so it MAY be possible to make it more flexable and that COULD lead to different results - Nothing at all about bundling just any bunch of losing stratergies together and expecting a profit.
In my example using LTF was for demontration purposes - forget about it being called LTF and instead call it "All Rounder" Yes its first purpose it to try and lay the field but if the movement of the odds suggest it looks certain its now only going get one horse layed and therefore be a loser this race instead of just allowing this to happen and take the loss have the rule cancel this part and if possible have it take advantage of the new situation and perform a different task based on how the race looks to be shaping up from the make up of the odds - this may still lose but you wont be any worse off than leaving it as it was but will give you extra chances of success so decrease the number of loses you have and increase the winning strike rate.
It does not have to be complex to cover multiple eventuallties, the rule I use the most has just 3 lines (rule types) and only uses 4 different conditions and can apply it to 90% of races each day and think i have had 4-6 losing days the whole of this year across both Flat and NH on everything from Grade 1 to nurseries.
What i did say at the bottom of my post was that some people MAY already have ran a rule that was just short of producing a profit or breaking even and if so it MAY be possible to make it more flexable and that COULD lead to different results - Nothing at all about bundling just any bunch of losing stratergies together and expecting a profit.
The only thing I would add to the first post would be to try it in Practice mode too. (or keep minimal funds in your account).
I think the other key point (Which Dallas has mentioned) is to keep it as simple as possible. Its underated imo
Ive built systems that had so many conditions and dependencies, they were bullet proof...the only problem was they virtually only triggered once every blue moon!
If you have a basis concept or idea, try it, and then hone and refine it.
Over the years, Ive always had BA running alongside my main system I use for work and I must have seen tens of thousands of in running markets. Quite often when it running, you can see something that isnt quite right (or could be just that little bit better and can capitalise on it) and you can virtually change it on the fly.
Above all; have some fun with it!
Regards
Peter
I think the other key point (Which Dallas has mentioned) is to keep it as simple as possible. Its underated imo
Ive built systems that had so many conditions and dependencies, they were bullet proof...the only problem was they virtually only triggered once every blue moon!
If you have a basis concept or idea, try it, and then hone and refine it.
Over the years, Ive always had BA running alongside my main system I use for work and I must have seen tens of thousands of in running markets. Quite often when it running, you can see something that isnt quite right (or could be just that little bit better and can capitalise on it) and you can virtually change it on the fly.
Above all; have some fun with it!
Regards
Peter
Amazing effort, Dallas
You'll still get forum users like Haichless criticise you though, as in his eyes, experienced traders should be going to people's houses and actually installing it on their computers for them
If there's one person who makes money through your generosity then that should be commended
You'll still get forum users like Haichless criticise you though, as in his eyes, experienced traders should be going to people's houses and actually installing it on their computers for them
If there's one person who makes money through your generosity then that should be commended
Dallas I can understand well what you mean, but have a different look on this.Dallas wrote: In my example using LTF was for demontration purposes - forget about it being called LTF and instead call it "All Rounder" Yes its first purpose it to try and lay the field but if the movement of the odds suggest it looks certain its now only going get one horse layed and therefore be a loser this race instead of just allowing this to happen and take the loss have the rule cancel this part and if possible have it take advantage of the new situation and perform a different task based on how the race looks to be shaping up from the make up of the odds - this may still lose but you wont be any worse off than leaving it as it was but will give you extra chances of success so decrease the number of loses you have and increase the winning strike rate.
Your example from first post can be devided into two separate tasks:
LTF + cancel (S1)
Back 5-9 (S2)
After one month of doing this take all bets placed by S1 and evaluate the profit and do the same for S2 as well. If you see that one of this two strategies has negative result, just forget it and continue just with positive one.
In this example lets assume, that combination above is profitable: S1+S2 is generating profit but if you evaluate separately all bets placed by S1 the result is negative and all bets placed by S2 are in total profitable. So you can easily continue running S2 separately (without placing Lay bets and cancelling them) under the same set of conditions and your result will be better.
I hope you agree wit this.
Of course if you find by separate evaluation, that S1 is profitable and S2 as well, it is very good idea to combine S1 and S2 as you save a lot of commission.
I am not trying to criticise successful trader and I don't expect anyone sharing golden eggs with me. I believe I am not a rookie, my average is about 1.2k/week this year using automation only. I am just sharing my experience as well as Dallas does.
I am not trying to criticise successful trader and I don't expect anyone sharing golden eggs with me. I believe I am not a rookie, my average is about 1.2k/week this year using automation only. I am just sharing my experience as well as Dallas does.
Dallas said in his first post that it wasn't a guaranteed way of getting a profit, it's put forward as an idea that people can play about with.
I'v not done any inplay trading for about a year but i never thought about the idea of if one stratagy isn't working to change to another, i think it's very genorous of Dallas to put this forward, in the two years i'v been on Betangel i don't recall anyone else doing this?
I'v not done any inplay trading for about a year but i never thought about the idea of if one stratagy isn't working to change to another, i think it's very genorous of Dallas to put this forward, in the two years i'v been on Betangel i don't recall anyone else doing this?
DavDvo I welcome and respect posts and views from anyone but still think you have miss understood the overall point, or did until you posted your 3rd comment yesterday as you kept referring to "combining losing strategies together", the only part of my original post that this could even remotely have been mistaken for was clarified for you in my last message to you, I respect your right to oppose mine or any other forum post but opposing it on the grounds of something not actually said "Any combination of non profitable strategies is non profitable" if im to explain it further for you can you at least quote the exact part of my post where i actually say "you can combine two losing strategies into one rule and turn a profit"?
I have to say I am skeptical.
I can see what Davdvo is saying, unless both strands of the strategy are viable then the one will drag down the profits on the other. What you seem to be saying is that you have come up with not one but two successful systems.
I have been trying (on and off) for over 10 years to work out one system that produces even a small profit let alone two systems!
I am convinced that there must be a system that where X and Y happens then Z will happen enough times to make a small profit in the long run. Despite my conviction I am still searching for X and Y.
I have tried something similar to your lay the field example but the reality is that you end up filtering out as many winners as you do losers. You would think that if no other horse is trading below 5 then the chances are that the favourite will go onto win but in fact a fair few races from that scenario will produce 2 or more horses below 1.75 and in those races where 2 horses trade below 5 many will end up with only one below 1.75. So it's as long as is it short.
Can you honestly say that this strategy has produced long term profits and if so what is your motivation for handing out free money when in my 10 years of research I have never come across someone willing to give a way their golden goose.
I note that you are very active on the forum and are always willing to help, so please do not take this as an attack on your integrity.
I can see what Davdvo is saying, unless both strands of the strategy are viable then the one will drag down the profits on the other. What you seem to be saying is that you have come up with not one but two successful systems.
I have been trying (on and off) for over 10 years to work out one system that produces even a small profit let alone two systems!
I am convinced that there must be a system that where X and Y happens then Z will happen enough times to make a small profit in the long run. Despite my conviction I am still searching for X and Y.
I have tried something similar to your lay the field example but the reality is that you end up filtering out as many winners as you do losers. You would think that if no other horse is trading below 5 then the chances are that the favourite will go onto win but in fact a fair few races from that scenario will produce 2 or more horses below 1.75 and in those races where 2 horses trade below 5 many will end up with only one below 1.75. So it's as long as is it short.
Can you honestly say that this strategy has produced long term profits and if so what is your motivation for handing out free money when in my 10 years of research I have never come across someone willing to give a way their golden goose.
I note that you are very active on the forum and are always willing to help, so please do not take this as an attack on your integrity.
For the purposes of explaintion in this post I used the Lay the Field stratergy as its the most well known and one that almost everyone can relate to, the other part i used about backing was a on the fly scenario which i made up as i was writing the post and i did say not to expect a profit from doing this and it was solely for example purposes.
Davdvo did PM me a while back just after his posts and the confusion was cleared up.
I would never give away any profitable automation stratergy to people i dont know but i will offer general advice on how i go about something and highlight things people might want to take into account which they may not of previously considered.
This post was to get people to be more open minded and not to just blindly create a rule with one function in mind and be limited to how many races it can be run on and worse still allowing the rule to still trigger even if the way things are shaping in the race its highly unlikely to produce a profit.
What i will say is i have three profitable automations one of which i can use on most races and all three have multiple parts to them, one does inclued a profitable LTF part but for obvious reasons is confugured completly different to what i used in the explanation.
Too simplfy things down if we forget about the additional parts/strands if someone was to run a basic LTF rule on every race and i did the same but added just one condition i could guarantee the following two things
The rule they run would produce a long term loss
Mine would produce a much smaller loss
Also as I posted earlier I always respect the posts or views of everyone and welcome any comments or questions
Davdvo did PM me a while back just after his posts and the confusion was cleared up.
I would never give away any profitable automation stratergy to people i dont know but i will offer general advice on how i go about something and highlight things people might want to take into account which they may not of previously considered.
This post was to get people to be more open minded and not to just blindly create a rule with one function in mind and be limited to how many races it can be run on and worse still allowing the rule to still trigger even if the way things are shaping in the race its highly unlikely to produce a profit.
What i will say is i have three profitable automations one of which i can use on most races and all three have multiple parts to them, one does inclued a profitable LTF part but for obvious reasons is confugured completly different to what i used in the explanation.
Too simplfy things down if we forget about the additional parts/strands if someone was to run a basic LTF rule on every race and i did the same but added just one condition i could guarantee the following two things
The rule they run would produce a long term loss
Mine would produce a much smaller loss
Also as I posted earlier I always respect the posts or views of everyone and welcome any comments or questions
I created a little test today using a basic strategy inspired by the original post by Dallas, whilst I was out and about, its had a fair day. (that last race at Wolv 16:45 had 4 under 2 for instance).
Worth having a crack, even if you just use tiny stakes, you wont lose much (Test in practice etc first)
Good fun!
Regards
Peter
(PS One thing I find useful is if you have two or more strategies running and they all fire, it can sometimes be difficult to unravel it at the end of the day to see what worked and waht didnt. One tip: You can vary your stakes, so say you are running two lay stratgeies for instance, system 1 may use £5 stakes whereas systen 2 uses £5.10 stakes...when you review your bets at the end of the day, you can clearly see which strategy worked by the stake value)
Worth having a crack, even if you just use tiny stakes, you wont lose much (Test in practice etc first)
Good fun!
Regards
Peter
(PS One thing I find useful is if you have two or more strategies running and they all fire, it can sometimes be difficult to unravel it at the end of the day to see what worked and waht didnt. One tip: You can vary your stakes, so say you are running two lay stratgeies for instance, system 1 may use £5 stakes whereas systen 2 uses £5.10 stakes...when you review your bets at the end of the day, you can clearly see which strategy worked by the stake value)