Close Trade If In Loss Over 10 Secs Guardian Automation Bot

Post Reply
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

Radio Rod wrote:
Sun Sep 04, 2022 9:47 pm
Interesting insight on the flip Shaun. I find that at the business end of a race when the losing horses are starting to fade away the lay odds rise steadily but the back odds can vary hugely from a couple of ticks away from the lay odds to suddenly 10-20 ticks away from the lay and then immediately back again. This can be enough to trigger my stop prematurely, where it looks as if it shouldn’t have triggered at all.

It’s making my head hurt figuring ways around this so maybe I need a new approach.
There's no easy way round market efficiency. If its 2.0 it's 50% going to win/lose. If its 4.0 then it's 25/75.

It's not much of an insight really, you only make money betting (trading) by getting value bets. The situation you describe eg 10 or 20 tick moves in very short timescales probably doesn't represent the real world change in likelihood, so the appropriate approach is to maintain offers say 15 ticks either side of the current price..... and a lesser offset earlier in the race when there's less volatility.

Don't get hung up on greening and stops, if you're not doing it at a fair price then it's going to cost you money. Also, are you taking prices or offering? If you're taking then you're filling the pockets of the people making smart offers. If you're offering then that's less likely. You mention a new approach, the first thing would be to realise the market doesn't care about your position therefore its not something you should use as a decision to bet.

Imo losing new traders aren't doing much more than having multiple bets, people who do OK have understood their part in the ecosystem. It should be trading 101 but rarely if ever gets mentioned here.
Radio Rod
Posts: 15
Joined: Sun Aug 15, 2021 10:16 pm

"The situation you describe eg 10 or 20 tick moves in very short timescales probably doesn't represent the real world change in likelihood, so the appropriate approach is to maintain offers say 15 ticks either side of the current price..... and a lesser offset earlier in the race when there's less volatility."

I'll keep an eye on this one Shaun, thank you.
icey1962
Posts: 21
Joined: Wed Jul 20, 2022 2:14 pm

Hi I'm wondering if someone can help me? I have tried automation but I seem to just have more bail outs of small amounts that amount to a large amount. It either bails out too soon in a race or if I extend the time I have a bigger loss. I'm just wondering if any one has it working well for them. I just trade the UK races. Is it better to trade the shorter odds horses? I usually L2B. Any guidance would be appreciated TIA
icey1962
Posts: 21
Joined: Wed Jul 20, 2022 2:14 pm

Hi Can anyone help me understand why the automation fired so late that I ended up losing £137 in one race, I have the Automation file set to fire if £-49.99 and limit to £-50 for 2 second rather than the default 10sec?
Is it a fault with my set up or the way the market went or duration left in race? I'm not sure what the log is telling me

14/12/2022 12:51:40: Market has been suspended
14/12/2022 12:51:40: Market has been unsuspended
14/12/2022 12:51:40: Market is in-play
14/12/2022 12:52:37: [G_Auto 1] : Automation Signal for market: loss = 1
14/12/2022 12:52:38: [G_Auto 1] : Automation Signal for market: loss = 2
14/12/2022 12:52:39: [G_Auto 1] : Automation Signal for market: loss = 3
14/12/2022 12:52:39: [G_Auto 1] : Automation Signal for market: stop = 10
14/12/2022 12:52:39: [G_Auto 1] : Cancelling bets 290233539800 290233543089
14/12/2022 12:52:40: [G_Auto 1] : Greened up Swiss Pride by Backing 146.42 at 1.13
14/12/2022 12:52:42: [G_Auto 1] : Automation Signal for market: stop = 10
14/12/2022 12:52:43: [G_Auto 1] : Greened up Swiss Pride by Laying 10.21 at 1.68
14/12/2022 12:52:45: [G_Auto 1] : Automation Signal for market: stop = 10
14/12/2022 12:52:46: [G_Auto 1] : Cancelling bets 290233700216
14/12/2022 12:52:46: [G_Auto 1] : Greened up Swiss Pride by Laying 5.62 at 3.05
14/12/2022 12:52:51: Market has been suspended
User avatar
Mastermind
Posts: 103
Joined: Mon Jul 27, 2020 8:39 pm

Hi everyone, hello Dallas,

I have merged my greyhound bot I am working on with this "Close Trade if In Loss Over 10 Sec" Bot... then a problem occured ... when the now in my bot implemented stoploss after these 10 seconds in loss is triggered (Live mode, 2 Pound Stakes, Stoploss 10 Cents) and it´s greened up, then the bot starts again to fire a new back bet on the runner into the market. How can I fix this? I want the stoploss after its triggered to stop this trade, not to green up and start to back the selected runner again in the same race.

Some help would be nice, I appreciate it like always.

Best regards.
User avatar
Dallas
Posts: 22674
Joined: Sun Aug 09, 2015 10:57 pm
Location: Working From Home

Mastermind wrote:
Sat Sep 23, 2023 11:01 am
Hi everyone, hello Dallas,

I have merged my greyhound bot I am working on with this "Close Trade if In Loss Over 10 Sec" Bot... then a problem occured ... when the now in my bot implemented stoploss after these 10 seconds in loss is triggered (Live mode, 2 Pound Stakes, Stoploss 10 Cents) and it´s greened up, then the bot starts again to fire a new back bet on the runner into the market. How can I fix this? I want the stoploss after its triggered to stop this trade, not to green up and start to back the selected runner again in the same race.

Some help would be nice, I appreciate it like always.

Best regards.
You'll need to add another 'Set/Modify Signal' rule
and on the signal tab set a signal (use any name you want ie, stop) on the market

and on that rule add a Signal Value condition
to test the signal named loss
for the market
value is equal to 10


Then on the rules you want to stop triggering once this one has add a 'Signal Set condition'
to test the signal named stop (or whatever you named it)
for the market
is set with a value
User avatar
Mastermind
Posts: 103
Joined: Mon Jul 27, 2020 8:39 pm

Thanks .. tested on six US-Horse Races .. six times stopped in loss ...I have to tweak around now with it ... maybe because of the spread the stop has to be a liitle bit farer away. But yes, the merger itself of the two bots works technically very well. thanks for your help, Dallas.
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

Mastermind wrote:
Sun Sep 24, 2023 1:08 am
Thanks .. tested on six US-Horse Races .. six times stopped in loss ...I have to tweak around now with it ... maybe because of the spread the stop has to be a liitle bit farer away.
Are you offering a price or taking?
User avatar
Mastermind
Posts: 103
Joined: Mon Jul 27, 2020 8:39 pm

ShaunWhite wrote:
Sun Sep 24, 2023 12:44 pm
Mastermind wrote:
Sun Sep 24, 2023 1:08 am
Thanks .. tested on six US-Horse Races .. six times stopped in loss ...I have to tweak around now with it ... maybe because of the spread the stop has to be a liitle bit farer away.
Are you offering a price or taking?
Thats one of the next points to tweak ... tbh I have so many zones under construction in the Bet Angel Universe .. it´s hard to find the right focus in case of what to do in which order ... even if you are a crazy workaholic like me. Last night I was focussing on how it technically works and it works from the technical point of view very well. And to duplicate it to a second bot to lay first. For now .. no misbehaviour .. but yeah .. in relatively illiquid markets. I used - because of the night time I had time to tweak around - I used the greyhound bot in illiquid US-horse racing markets. Technically it looks like it worked well, but because of the spread the stoploss had to be farer away then normally and there is also the question if I in general should use a stop loss. Because .. the stop loss safes you in theory from losing MORE money (people say throwing good money behind the one which is already gone) ... and in how many cases does the drifting really goes further after after are stopped out - and not coming back a little bit. In one trade I had a stop at 20 Cent loss (5 Pound Stake, Practice Mode) and the drift was so f**king fast that the time stop 10 seconds leaded to green out at 48 cents (instead of triggerded at 21 and then lets say filled at 25. I saw a lot in pre-off markets .. but that stop-loss triggered at 21 and filled at 48 is a nice spike. It was directly at the start of the time period the bot was planned to work, so it has really 120 seconds or more to come back a liitle bit .. what it in fact did .. around 40% of the ticks I lost with the stop would have been still on the virtual (practice mode) account if I just would have been stubborn in the position in loss and red out over time short to the off with the orders the bot always have to do anyway . So .. there has to be really reliable data if I should use a stop or not. No one has that data. Normally I should do an excel with all races I let it run in practice mode .. and not only make notice profit and loss .. in each case the stop is triggered, I should make notice at which point its triggered, where the order is filled and much more important .. what happened in XXX seconds rest time to the off ... I mean .. does it come back after stoploss or not and if yes or no .. how much. Thats all work for a excel professor like Peter (and I mean this as a compliment, I love his deep thoughts into the excel data) ... but we speak here about hundreds of races which will be needed to find these data in the way I mentioned. If I want to have a look how drifters come back after stoploss, I need a reliable amount of races stopped out .. but how many percentage are stopped out? So to find 100 which are stopped out I need 200 races .. or more or less? and they must be comparable .. not some in the US .. some in AUS and some in UK .. I have to take the factor illiquidity out in the first place ... so I can only collect this kind of data in the UK and Irish markets. Haha .. so yes, you all know this .. some one can be very, very busy in this excel things ... hahaha .. but yeah, its fun and if someone has a little autistic touch like me .. so why not .. haha. So back to your question Shaun, it takes at the final green up instead of there possibly also offering for some seconds first. If I want to offer .. I need more time .. lets say .. not taking 5 seconds before the off .. instead .. offering 15 seconds before the off as reverse price, let it stand there 8 seconds and if its not matched ... take it then at best price. But here the same ... I offer and when its taken, then I have one tick profit more. Thats right. BUT ... in busy markets .. this 1 tick costs me 10 seconds around before the off .. in the craziest market time ... in how many cases would I make another matched bet in the same time period to take the spread of two ticks my bot is working with? Hahaha .. so yes .. you see I am laughing .. so the first priority was now to get it technically merged and of course come deeper in the understanding of how the signaling and everything of the automation works. I think the rest of the tweaking ideas are coming watching it working and of course one of the next priorities is then getting the work of the bot(s) recorded to study them, especially in the final phase of closing the bets, greening out before the off and so on. Also interesting would be the question with this small stakes .. if I should possibly change some things a little bit and not take the best price to close the trade but to tweak a little bit around with taking instead the best price or offering the reverse price simply taking the BSP to close. These ideas are coming during watching the bot working I think. Excuse some bad english .. I have to tweak this as well, haha. But noting my thoughts on this down is helpful I think, at least for me a little bit .. haha. :D
User avatar
Mastermind
Posts: 103
Joined: Mon Jul 27, 2020 8:39 pm

Thats exact what I meaned ... Towcester 16.08 h (Practice Mode, 5 Pound Stake backing the Fav .. stopploss at 0,40) the SL-counter started three (!) times standing at 0,42 some seconds, but it never get the 10 seconds finished .<(counted till 1,6,4) .. then the dog was backed again .. it ended then after heavy laying in the last 20 seconds with a loss of 0,29 after red up, BUT .. around 20 sec before off it was back around zero, 2 cents or so in profit for 2 or 3 seconds ... so it could also make sense to merge a third bot with it. With the rule ... IF the counter of the stop loss has started to count twice or more .. and the trade is still alive ... (then this trade is clearly hanging like a boxer in the ropes, does not work like expected) .. then grab a green if its back in profit, say thank you and full stop in this market no matter what the clock says. But you all know the possibilities with Bet Angel Professional are as gigantic as the universe, in fact endless. :lol:
User avatar
Mastermind
Posts: 103
Joined: Mon Jul 27, 2020 8:39 pm

ShaunWhite wrote:
Sun Sep 24, 2023 12:44 pm
Mastermind wrote:
Sun Sep 24, 2023 1:08 am
Thanks .. tested on six US-Horse Races .. six times stopped in loss ...I have to tweak around now with it ... maybe because of the spread the stop has to be a liitle bit farer away.
Are you offering a price or taking?
Short version .. the results of my bot are a horrorshow ..and I think with timestop more worse then without. But this needs to be tested with two devices parallel next time. First I have to turn it around to lay first .. and Shaun, of course to offer instead of taking .. and the timerange I let him work is for greyhound to long. But I wrote in Excel now a nice spreadsheet to notice all results to compare and tweak around. So the sunday is not really wasted, but the results with the two merged bots is in the current adjustment a complete mess. But now I know this excact .. will change some things .. and then the next tests will follow. Haha .. :lol:
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

Mastermind wrote:
Sun Sep 24, 2023 7:32 pm
ShaunWhite wrote:
Sun Sep 24, 2023 12:44 pm
Mastermind wrote:
Sun Sep 24, 2023 1:08 am
Thanks .. tested on six US-Horse Races .. six times stopped in loss ...I have to tweak around now with it ... maybe because of the spread the stop has to be a liitle bit farer away.
Are you offering a price or taking?
Short version .. the results of my bot are a horrorshow ..and I think with timestop more worse then without. But this needs to be tested with two devices parallel next time. First I have to turn it around to lay first .. and Shaun, of course to offer instead of taking .. and the timerange I let him work is for greyhound to long. But I wrote in Excel now a nice spreadsheet to notice all results to compare and tweak around. So the sunday is not really wasted, but the results with the two merged bots is in the current adjustment a complete mess. But now I know this excact .. will change some things .. and then the next tests will follow. Haha .. :lol:
Analysing your bets is key. Just looking at market totals doesn't look deep enough into where your strengths/weaknesses are.
User avatar
Mastermind
Posts: 103
Joined: Mon Jul 27, 2020 8:39 pm

I mirrored it and use now the lay bet first, it was running on 14 races and was this evening slightly profitabel from scratch in the first version in practice mode. But .. you see there a nice winning streak of 7 in a row .. so this 14 races is no reliable data. Too much randomness possible. Looks nice, but smelss random. Lets see what happens over a whole UK-Session and maybe more days. BUT .. the interesting .. this was only switching from back to lay. I made now after UK markets are closed some other slight tweaks including offering green up 8 seconds to the reverse price before if unmatched taking best. So this should make some more ticks over the long run towards profitability on top of todays more or less random positive result. Excited to see what happens over a whole UK-Session. Also interesting ... B2L 17 out of 40 races are closed by the time-stopploss (10sec) ...horrible amount ... but L2B only 1 out of 14. Thats very interesting and this result looks much more then random. Lets see what happens next.
You do not have the required permissions to view the files attached to this post.
User avatar
Kai
Posts: 6092
Joined: Tue Jan 20, 2015 12:21 pm

Mastermind wrote:
Sun Sep 24, 2023 3:14 pm
Thats one of the next points to tweak ... tbh I have so many zones under construction in the Bet Angel Universe .. it´s hard to find the right focus in case of what to do in which order ... even if you are a crazy workaholic like me. Last night I was focussing on how it technically works and it works from the technical point of view very well. And to duplicate it to a second bot to lay first. For now .. no misbehaviour .. but yeah .. in relatively illiquid markets. I used - because of the night time I had time to tweak around - I used the greyhound bot in illiquid US-horse racing markets. Technically it looks like it worked well, but because of the spread the stoploss had to be farer away then normally and there is also the question if I in general should use a stop loss. Because .. the stop loss safes you in theory from losing MORE money (people say throwing good money behind the one which is already gone) ... and in how many cases does the drifting really goes further after after are stopped out - and not coming back a little bit. In one trade I had a stop at 20 Cent loss (5 Pound Stake, Practice Mode) and the drift was so f**king fast that the time stop 10 seconds leaded to green out at 48 cents (instead of triggerded at 21 and then lets say filled at 25. I saw a lot in pre-off markets .. but that stop-loss triggered at 21 and filled at 48 is a nice spike. It was directly at the start of the time period the bot was planned to work, so it has really 120 seconds or more to come back a liitle bit .. what it in fact did .. around 40% of the ticks I lost with the stop would have been still on the virtual (practice mode) account if I just would have been stubborn in the position in loss and red out over time short to the off with the orders the bot always have to do anyway . So .. there has to be really reliable data if I should use a stop or not. No one has that data. Normally I should do an excel with all races I let it run in practice mode .. and not only make notice profit and loss .. in each case the stop is triggered, I should make notice at which point its triggered, where the order is filled and much more important .. what happened in XXX seconds rest time to the off ... I mean .. does it come back after stoploss or not and if yes or no .. how much. Thats all work for a excel professor like Peter (and I mean this as a compliment, I love his deep thoughts into the excel data) ... but we speak here about hundreds of races which will be needed to find these data in the way I mentioned. If I want to have a look how drifters come back after stoploss, I need a reliable amount of races stopped out .. but how many percentage are stopped out? So to find 100 which are stopped out I need 200 races .. or more or less? and they must be comparable .. not some in the US .. some in AUS and some in UK .. I have to take the factor illiquidity out in the first place ... so I can only collect this kind of data in the UK and Irish markets. Haha .. so yes, you all know this .. some one can be very, very busy in this excel things ... hahaha .. but yeah, its fun and if someone has a little autistic touch like me .. so why not .. haha. So back to your question Shaun, it takes at the final green up instead of there possibly also offering for some seconds first. If I want to offer .. I need more time .. lets say .. not taking 5 seconds before the off .. instead .. offering 15 seconds before the off as reverse price, let it stand there 8 seconds and if its not matched ... take it then at best price. But here the same ... I offer and when its taken, then I have one tick profit more. Thats right. BUT ... in busy markets .. this 1 tick costs me 10 seconds around before the off .. in the craziest market time ... in how many cases would I make another matched bet in the same time period to take the spread of two ticks my bot is working with? Hahaha .. so yes .. you see I am laughing .. so the first priority was now to get it technically merged and of course come deeper in the understanding of how the signaling and everything of the automation works. I think the rest of the tweaking ideas are coming watching it working and of course one of the next priorities is then getting the work of the bot(s) recorded to study them, especially in the final phase of closing the bets, greening out before the off and so on. Also interesting would be the question with this small stakes .. if I should possibly change some things a little bit and not take the best price to close the trade but to tweak a little bit around with taking instead the best price or offering the reverse price simply taking the BSP to close. These ideas are coming during watching the bot working I think. Excuse some bad english .. I have to tweak this as well, haha. But noting my thoughts on this down is helpful I think, at least for me a little bit .. haha. :D
Holy wall of text Batman :shock:

Paragraphs are your friends :!:
User avatar
ShaunWhite
Posts: 9731
Joined: Sat Sep 03, 2016 3:42 am

Kai wrote:
Mon Sep 25, 2023 9:23 am
Mastermind wrote:
Sun Sep 24, 2023 3:14 pm
Thats one of the next points to tweak ...
Holy wall of text Batman :shock:

Paragraphs are your friends :!:
I asked chatGPT to summarise it.... ;)
"The speaker is deeply engaged in optimizing and tweaking multiple zones in the Bet Angel Universe and is specifically working on technical refinements and strategic adjustments for bot performance in various racing markets. He/she contemplates the efficacy of using stop-loss, considers the variability of market liquidity, discusses strategy adjustments in busy markets, and indicates a need for extensive data analysis to make informed decisions while expressing enjoyment in the detailed and technical nature of the work."
Mastermind wrote:
Sun Sep 24, 2023 3:14 pm
I offer and when its taken, then I have one tick profit more. Thats right. BUT ... in busy markets .. this 1 tick costs me 10 seconds around before the off
Yes it all depends on the situation. If you want an easy ½ a tick then close at BSP which will be in the spread, and it's a known 0ev bet so harmless.
Post Reply

Return to “Bet Angel Automation - General”