BACK/LAY by liability, in play BOT.!shock!

Share your automation file with other users.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

decomez6 wrote:
Sat Jul 18, 2020 6:21 am
And the only way to convert a simple strategy into a winning one is by timing and conditioning,

The ‘ELEPHANT in the ROOM is , how can a collective imaginative capacity of a forum like this cannot Convert a simple strategy like the one proposed here into a profitable one?

This video from Peter is a brilliant way of converting a simple strategy into a winning one.
That first paragraph simply isn't correct. And for someone with so little experience you should be careful of making assertions like that because you'll close your eyes to all the other considerations.

As for the 2nd paragraph then your strategy appears to be based on hope rather than anything fundamental. You can't just pick a random idea and coax it into life. Your idea that you'll make it work by collective imagination is falling into that trap.

I haven't watched that video but I assume you're referring to devising a strategy and then addressing the losers, leaving you with just the winners. Again that's naive and over simplistic.

The one question you haven't addressed is why this strategy should work. Your enthusiasm is great but you're missing so much, even when it come to results several days or even several weeks of wins means nothing without you being able to measure if that gain is coming from an edge or natural varience. You've just picked a random idea and are making random changes, that's not how it's done.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

Have you read the threads about strategy development and perhaps those on modeling too? Plus all the threads about the fundamentals of developing automation, incl sample sizes..Seems like you've just watched a couple of videos and that's it.
User avatar
decomez6
Posts: 651
Joined: Mon Oct 07, 2019 5:26 pm

ShaunWhite wrote:
Sat Jul 18, 2020 1:58 pm
decomez6 wrote:
Sat Jul 18, 2020 6:21 am
And the only way to convert a simple strategy into a winning one is by timing and conditioning,
That first paragraph simply isn't correct. And for someone with so little experience you should be careful of making assertions like that because you'll close your eyes to all the other considerations.
i stay corrected , ofcourse there are countless ways to flip a coin with equal chances of loosing and winning minus the Noise ( standard deviation). i am an enthusiast, abressive teachers can only make me sharp. i mean this in a good way.

i have thrown a glance at modellinng and backtesting , amazed at the level of sophistication.this forum is rich with senior experienced members and the discussions are defensive with a suggestive approach. MARTINGALE is a crime! :)
User avatar
decomez6
Posts: 651
Joined: Mon Oct 07, 2019 5:26 pm

ShaunWhite wrote:
Sat Jul 18, 2020 1:58 pm
decomez6 wrote:
Sat Jul 18, 2020 6:21 am

This video from Peter is a brilliant way of converting a simple strategy into a winning one.
I haven't watched that video but I assume you're referring to devising a strategy and then addressing the losers, leaving you with just the winners. Again that's naive and over simplistic.

The one question you haven't addressed is why this strategy should work. , even when it come to results several days or even several weeks of wins means nothing without you being able to measure if that gain is coming from an edge or natural varience. You've just picked a random idea and are making random changes, that's not how it's done.
that video brought me to bet angel , was my introduction to sports trading. its pure strategy ,i actually made the following servant to try and capture a random WIN and LOOSE streaks , which will correspond to black and red colours on the cards respectively.
the servant SCRATCHES and SWITCHES direction, to try and catch a trend. its very basic .
IN theory all what one needs is place any bet and OFFSET outside the current asking price with a 200% netstake .
You do not have the required permissions to view the files attached to this post.
betx
Posts: 15
Joined: Thu Jul 02, 2020 11:59 am

I did decide to use real stakes because I saw it was making money on practice but I should never have left it alone.

So Anyway...

You seem to have some good ideas.

Why don't you explain how to actually use the bot :lol:

1. Is it just on the FAV
2. Can you use it on horses, dogs, football?
3. What is it actually doing

Just be nice to know then we can maybe understand how to tweak it.

Thank you.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

betx wrote:
Mon Jul 20, 2020 12:11 pm
I did decide to use real stakes because I saw it was making money on practice
How many races did you let it run for? Your sample size needs to be proportional to the expected strike rate. A strategy that has a 50% strike rate will need the fewest races to test it, and as that moves towards 1% or 99% that sample size needs to increase.

Say you're laying 100/1 shots, 99 or even 199 races won't tell you much. Same with backing 1.01 shots, you'll have a few hundered good results in a row and the loser or 6 straight losers might not happen for 500 races. But if you expect a strike rate of say 55% then you should be able to start to get an impression of how it's doing in as few as 2 or 300 races.

Set up a little spreadsheet with 100 or even 10,000 lines that are randomly 1 or 0 and plot the running total. Then press f9 a few times and see how varied the results can be even with a coin toss. You'll see just how big a sample need to be to get a fair impression. This example returns a 1pt win or a 1pt loss, it has a strike rate of 55%... it's a winning strategy. It wasn't too hard to find a run of 50 that were losing and similarly if it had a strike rate of 40% it wouldn't have been hard to find a run of 50 that made a profit.


Edit, :) this example was wrong, the chart is showing a 45% win rate, can't really be bothered to change it ...but you get the point.
Screenshot_6.png
It's easy to be fooled by randomness. It might be worth you reading a book by Nassim Taleb on that very subject.
You do not have the required permissions to view the files attached to this post.
betx
Posts: 15
Joined: Thu Jul 02, 2020 11:59 am

Maybe we shouldn't trade :)

Honestly though I have been in this game along time...

I believe only a small percentage make a living from it.

Peter and a handful more.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

betx wrote:
Mon Jul 20, 2020 5:10 pm
Maybe we shouldn't trade :)

Honestly though I have been in this game along time...

I believe only a small percentage make a living from it.

Peter and a handful more.
It's a small percentage, and less than 4-5% of all accounts even show a profit, but that's still in the thousands making a reasonable living. Not everyone uses BA and the forum and if they do not everyone feels the need to talk about it. You tend to just hear from people struggling who need help. It's all fine though, if it was easy there wouldn't be the same rewards for cracking it. It's only the funneling of money from the 95% to the 5% that makes it worth doing. If it was 50/50 there wouldn't be enough money from the 50% losing to feed the other 50%.
User avatar
decomez6
Posts: 651
Joined: Mon Oct 07, 2019 5:26 pm

betx wrote:
Mon Jul 20, 2020 12:11 pm
You seem to have some good ideas.
Why don't you explain how to actually use the bot :lol:
Just be nice to know then we can maybe understand how to tweak it.
Thank you.
:) nice to see you`r back in the ring ,rolling with the punches . You can never loose ,can only learn.
i dont know about good ideas ,i seem to have a new one everytime i watch a video or read a post like the one by Shaun.

as for the bot :LAY by liabilty back by NETSTAKE , it simply lays all selections and offset them by X ticks 20,30,40,etc.
Best case scenario.:
all your horses will close on a profit. this will only happen if your winning horse is greened up before the race ends.and this where you need to do a lot of testing to see which kind of races and race tracks this is happening more often than not.
in all other cases : all your horses will green up apart from the winner. this throws in a spectrum of possibilities, . In handicap races , the competition is high and you will have horses layed at lower odds in close ranges, and a possibility of the winner dobbing and getting matched up. in other races you have one clear favourite and your p/L is only going to show green just before the end of a race as the prices are settling down and your layed horses are summing up againsst the winners total liability.from what i have observed( i can be wrong)ts a very small window and can be very costly if not taken. thats why race selection is important.

the other BOT: LAY and BACK by LIABILITY ; is the crazy one i have no clue what s going on there. the swings are too crazy. so its only for experiment, but it does exactly as the one above but offsets using liability stakes instead of net stakes.
i actually put them out there for a collective experiment, hopefully we can all bring our minds to one place and develop one profitable bot. Not to mean there are non in this forum , but an additional one wont harm.

shaun has given a great break down on randomness and its elusive nature.i believe randomness is just another way to define infinite continuity.WATCH the video above , it will give you a clear guidance on how to deal with occuring patterns , this will give you an upperhand on using frequency to distort continuity.the card counting game is a great strategy on its own merit.CHECK IT OUT.

strike rate is directly proportaional to implied probablity .the price of a selection is a clear indication of what you will get in return . anything under 2.0 is always a good value to lay. so if you are only dealing with only one selection you can only loose what you stake, but anything above 2.0 can only gain you more than your initial bet if you back it and check the strike rate of this selection, if its more than 50% and you can afford to pay the betfair commission with a surplus in your back pocket, thats a wiining strategy.

there are many ways to achieve a strike rate of 50% or more. normally selecting more than one horses , especially the top two will easily get you there.
how i would personally do it. is . lay three horses at lowest odds anything under 4.0.. if iam using £10 fixed stakes my liability is always around £30 and i stand to win or loose £ 30 , regardless of the outcome.
so my risk to reward ratio is 1:1. 50/50.... i can then use the card counting trick on the video above to catch a pattern.
............
MORE BOTS ON THE WAY , ..........
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

decomez6 wrote:
Mon Jul 20, 2020 5:44 pm
.... anything under 2.0 is always a good value to lay. .
I'm sorry but this is utter nonsense. Value occurs at any price.

And strike rate has ZERO to do with winning, avg win amount > avg lose amount is what makes money.

But if you want to continue with your lay everything with an offset bot, then you'll probably need some sort of race jepardy index whereby the going, jumps/flat, distance, horse experience and course terrain might all cause the eventual winner to falter at some point. But I don't even see why you need an offset, you're just cutting your lay gains short and letting your loser go on to win.
User avatar
decomez6
Posts: 651
Joined: Mon Oct 07, 2019 5:26 pm

ShaunWhite wrote:
Mon Jul 20, 2020 7:48 pm
decomez6 wrote:
Mon Jul 20, 2020 5:44 pm
.... anything under 2.0 is always a good value to lay. .
I'm sorry but this is utter nonsense. Value occurs at any price.
at any price including ,2.0 , yes.
laying 3 horses under 2.0 will bag a guranteed profit, -Any 4 horses under 3.0, -Any 5 under 4.0...you get the drift.
all those position represent the same RETURN on INVESTMENT but a varied RISK to REWARD. if any of the bets goes against me , its only at 2.0 and below where i loose less than my initial stake, hence less risk.
AT 1.01 yous are only risking a 1p to apound, and this is why its profitable to keep your money laying there instead of keeping it a bank account earning silly savings that are taxable and have very little compounding effect. any one with cash to spare can feel the real magic of compouding at 1.01...
there are many ways to mitigate any risk ,, some may have complicated formulaes that calculate the exact amount needed to break even at any given point, or others may wait for the right moments while using fixed stakes.

strike rate has ZERO to do with winnin ... s money.

strike rate is tied to the risk and reward ratio, and only then does it determine the outcomes and how much you stand to loose or win.
another way to look at it, every strike rate has is corresponding staking plan , while every bank size has its corresponding level of confidence. i sit at the riverside doing my fishing using a fishing rod while the big Sharks got deep pockets to apply the staking plans at thier will and disposal. and BOY they are plenty!
You do not have the required permissions to view the files attached to this post.
User avatar
decomez6
Posts: 651
Joined: Mon Oct 07, 2019 5:26 pm

ShaunWhite wrote:
Mon Jul 20, 2020 7:48 pm
decomez6 wrote:
Mon Jul 20, 2020 5:44 pm
.... anything under 2.0 is always a good value to lay. .
But if you want to continue with your lay everything with an offset bot, then you'll probably need some sort of race jepardy index whereby the going, jumps/flat, distance, horse experience and course terrain might all cause the eventual winner to falter at some point. But I don't even see why you need an offset, you're just cutting your lay gains short and letting your loser go on to win.
if i cut my greens iam also cutting my reds that would potentially be working against me if iam not greened up , value is found at all prices,odds and positions ,remember. ;) in addition, i will be presesented with a potential of greening up a winner before the race even begins and i end up with a free bet plus a whole field layed and greened . this in return gives me ample time to close all trades and green all at any time of my liking.
that bot is water under the bridge , i have another better version. or so i think.

THE JOB AT HAND
you are well versed in programming and software development, correct me if am wrong. iam about to give you my understanding of what programming is all about, then explain why is relevant to take a programming approach to developing a profitable BOT.
HARDWARE
electrical flip swittches on a motherboard of any computer consist of Central Processing Units , Random Accesorry Memories etc.
at thier base level, they all work the same way. they are just electrical swithes that go ON/OFF.
SOFTWARE
its the ability to controll this millions of electrical flip swithces found in the computer ,so that they are projected on a monitor/Screen in a way that our brains can put the puzzle back together and make sense of what is being shown.
PROGRAMMER
you will be like an instructor of an ORCHESTRA, you tell the the the flip switches when and how to ON/OFF.

MY TAKe
the fundamentals of any market is that the prices will either GO up or Down ,,from a nuetral starting point. just like the BINARY functions of ON/OFF of the computer. thats why i believe there is a simple way of catching any trend in a trading platform.

SOLUTION
make a BOT, that places a BET, with a FILL or KILL,.
it will then OFFSET with a reverse bet placed 1 ticks outside the BEST reverse price.
(in ecxel this is a-1 reverse offset bet)
stakes of 200% of the netstake.
this will guarantee to SWITCH and SCRACTCH.
The bot is up there, please have a look and tell me what i could modify and make it catch a trend as soon as it switches UP /DOWN.

WE ARE ALL PART OF A PUZZLE and thats what iam bringing to the table.
HAVE a look at this video< its only 3 mins long, hope it speaks to you a languange ,that inspires comradeship.
my apologies for this loooong posts, i shall keep it short and to the point next time.

https://youtu.be/O5DACrSq5oU
elofan0
Posts: 316
Joined: Fri Jan 13, 2017 4:44 pm

The proof has to be when the bot is created that is consistent .......... i will keep watcing this post ... its interesting if nothing else for now ................... im sure something positive may eventually come from this ... if not you are a great salesman ...... :)
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

you are well versed in programming and software development,

:) I've dabbled. Compared to some here I'm 2nd division but I did a couple of decades in fund management IT last century so I get the general idea.

I'm fine with the idea of comradeship, solidary, being helpful etc and although my observations are a bit direct I wouldn't be making them without that in mind. But...... If you're looking for any sort of highly detailed collaboration I don't think you'll find it here.

The thing is, sports markets are tiny and the total available profit is even smaller. And it's not a battle with the market per se, rather a battle between each of us to extract some of that profit. Lots of us here have been working full-time+ for years to develop relaibale sustainable edges.... years, and risking that edge being diluted simply isn't a sensible choice.

If I were to tell you exactly how to exploit your idea and edit your bot for you, there's going to be occasions where your order might be ahead of mine, and as helpful as I am, I'm not inclined to be sharing my wages with you on a regular basis. This whole concept of trading being a competitive sport in its own right is why people spend huge amounts of time and money trying to be the fastest to market, writing their own software (including lightweight operating systems) and researching the ideal VPS service to shave off a few milliseconds here and there. This obviously isn't such a concern for manual trading because there are other far more important considerations and speed isn't a substitute for skill. But when you're talking about bots and algorithmic trading there comes a point when you're on your own. It's why you find so much more advice about trading manually than algorithmically.

I wish you luck and you're obviously keen but my suggestion would be for you to slowdown just a touch and really try to fully understand risk/reward, money management, statistics around sample sizes, the development life cycle, ways to distinguish between parameters that are irrelevant or relevant, the real world things that happen in racing etc etc. Building a profitable bot isn't something you're likely to do by the end of the year, so use this time to learn all you'll need to get to the point where that can happen, rather than bashing out random bots and making random changes hoping they'll work.

It would also be massively beneficial if you spent a few months manual trading first to improve your understanding and to generate ideas. And COLLECT data! So much more of your strategy development can be done via analysis that way, and it's much cheaper and faster. Don't worry about it not being hyper accurate because that's not relevant.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

elofan0 wrote:
Tue Jul 21, 2020 3:04 pm
The proof has to be when the bot is created that is consistent .....
... And you put your feet up in the garden feeling self satisfied. But 2 weeks later the results take a suicidal nosedive because someone somewhere is now doing something that affects you, and you're back to having zero income and a piece of paper with some scribbled ideas. Cue more months of graft, false dawns and disappoints.... Until the day you find the next thing :D. Just being able to fire up the PC to manually trade whatever there is using your skill and judgement starts to look very appealing.
Post Reply

Return to “Bet Angel Automation - Shared files”