Improvements to bot?

[chrissyb1977]

Hi,

I am currently using bots to try and take a few ticks profit whenever a horse hits odds of 3.6 in running. It places a back bet straightaway and also a lay bet 4 ticks below. I have put conditions in so that the same horse can't be backed/layed more than once and also historical odds conditions so that I'm not including horses where odds absolutely steam in as this can cause jumps with the lay. I also have it running for only the first 60-75% of a race so the same thing doesn't happen near the end.

I'm still getting a few backs placed that don't have a matching lay (I know this will happen from time to time) and I'm adjusting things as the days go by and I check the logs. I seem to be getting too many for my liking though.

Is there anything else that anyone can suggest? Is my fill or kill time ok? Whatabout the actual odds and the conditions?

Any suggestions at all would be MASSIVELY appreciated. I have attached two of my bots. They are all the same apart from different distances/hurdle races are active for different lengths of time during the race.

[jimibt]

it may be an idea to also add a Historic Relative Odds condition to ensure that the spread between the back and lay price is tight. this will help mitigate placing trades when the spread is wide open. something along the lines below should help:

Capture.PNG

[chrissyb1977]

That's a good idea... As even if the odds are trading below the conditions I've set it there could be a large gap between back and lay.

Thanks for your help mate, much appreciated!

[jimibt]

That's a good idea... As even if the odds are trading below the conditions I've set it there could be a large gap between back and lay.

Thanks for your help mate, much appreciated!

of course, my post was a bit incorrect in the implementation detail, the final section should of course be minus 2 ticks (and not 2 odds as i have in the image doh!! ) - less haste, more speed as my granny used to say...

[chrissyb1977]

I did think that maybe 2 odds was a bit too big but I'm still a novice at this

So will this avoid bets being placed basically near the end of the race when the odds are all over the place?

[jimibt]

I did think that maybe 2 odds was a bit too big but I'm still a novice at this

So will this avoid bets being placed basically near the end of the race when the odds are all over the place?

yes and no. it would certainly prevent trades being placed when the odds were swinging around but of course, they could momentarily settle for enough time for a trade to be placed at the wrong moment. to avoid that you'd need to put in place a set of signals, per race type that triggered to let you know that the given race was now at the 90% mark (for example) and was therefore out of bounds for placing trades etc. i did something on this a while back and have added it to your original rule to give you an idea of what i'm on about...

each distance flag is set to trigger, per race type, at a known moment in time (in this case at approx 50% and again on the 85-90% mark). i've basically added a few new conditions to your rule to say that the back bet rule can only fire when the race is between 50% and 90% complete. you can of course remove the 50% condition, i just added it to show the flexibility you could have.

this allows your single back bet rule to remain simple and rely on the distance flags to drive its activity.

reviewing the uploaded file will clarify the objectives. anyway, just food for thought...

[chrissyb1977]

That's brilliant, thanks for your help mate. Very much appreciated.

[jimibt]

That's brilliant, thanks for your help mate. Very much appreciated.

hope you find it a useful building block for this and future rules. i've found that being able to limit the viable activity band for placing trades has removed a lot of *noise* from my rules and thus allows me to focus in on the real points of interest within the race.

[chrissyb1977]

Yeah, it's been SO much more consistent today. In fact, until the last race just gone I had 16/16 greens and all expected levels. The last race has just wiped it all out but I'm actually much happier with how it's gone.

Now to work on increasing the number of races it bets on (only 16/34 races so far today) and also increasing the number of times it bets per race.

[jimibt]

Yeah, it's been SO much more consistent today. In fact, until the last race just gone I had 16/16 greens and all expected levels. The last race has just wiped it all out but I'm actually much happier with how it's gone.

Now to work on increasing the number of races it bets on (only 16/34 races so far today) and also increasing the number of times it bets per race.

chrissy -glad to hear that. small steps i always feel. sort the infrastructure, bullet proof the logic, then implement the rule to run against only the races that you know have a high strike rate and the rest will follow.

[starjock]

Hi Jim, Thanks for the file , Its not placing the bet , and I get this log after the race
9/03/2024 11:58:16: Guardian has detected that the market is suspended
9/03/2024 11:58:17: Guardian has detected that the market is in-play
9/03/2024 11:58:17: Guardian has detected that the market is now unsuspended
9/03/2024 11:58:18: [G_Auto 1] : Automation Signal for market: percentcomplete = 0
9/03/2024 11:58:19: [G_Auto 1] : Automation Signal for 3. My Ruby: wasintop2 = 1
9/03/2024 11:58:46: [G_Auto 1] : Automation Signal for market: percentcomplete = 50
9/03/2024 11:58:48: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:48: [G_Auto 1] : Automation Signal for market: percentcomplete = 50
9/03/2024 11:58:50: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:52: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:54: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:56: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:59:02: [G_Auto 1] : Automation Signal for 8. Little Bit Chai: wasintop2 = 1
9/03/2024 11:59:05: [G_Auto 1] : Automation Signal for 1. Lunar Hero: wasintop2 = 1
9/03/2024 11:59:11: Guardian has detected that the market is suspended

Can you help me with this please
Kim

[jimibt]

Hi Jim, Thanks for the file , Its not placing the bet , and I get this log after the race
9/03/2024 11:58:16: Guardian has detected that the market is suspended
9/03/2024 11:58:17: Guardian has detected that the market is in-play
9/03/2024 11:58:17: Guardian has detected that the market is now unsuspended
9/03/2024 11:58:18: [G_Auto 1] : Automation Signal for market: percentcomplete = 0
9/03/2024 11:58:19: [G_Auto 1] : Automation Signal for 3. My Ruby: wasintop2 = 1
9/03/2024 11:58:46: [G_Auto 1] : Automation Signal for market: percentcomplete = 50
9/03/2024 11:58:48: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:48: [G_Auto 1] : Automation Signal for market: percentcomplete = 50
9/03/2024 11:58:50: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:52: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:54: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:58:56: [G_Auto 1] : Automation Signal for market: percentcomplete = 90
9/03/2024 11:59:02: [G_Auto 1] : Automation Signal for 8. Little Bit Chai: wasintop2 = 1
9/03/2024 11:59:05: [G_Auto 1] : Automation Signal for 1. Lunar Hero: wasintop2 = 1
9/03/2024 11:59:11: Guardian has detected that the market is suspended

Can you help me with this please
Kim

apologies, i no longer have a BA subscription, otherwise would dive in and help. maybe one of the BA active members could jump in and have a look. I do recall this being a pretty useful ruleset and strategy.

[Fugazi]

Sorry not.fully.read the thread.

My question / criticism would be why specifically those odds? Is there some.analysos behind it? Does it work all markets ?

Otherwise youre just trading randomly and will be eaten by comission and losing out a lot of profit locking in a trade for a short profit, but big losses when you dont match

Also, interesting download to replies ratio. Shame a lot of lurkers arent more active on the forums (appreciate some just wont wanna clog the thread with thank yous).

[jamesedwards]

apologies, i no longer have a BA subscription, otherwise would dive in and help. maybe one of the BA active members could jump in and have a look. I do recall this being a pretty useful ruleset and strategy.

I think perhaps the issue is the back bet trigger requires the signal "percentcomplete" to be >50 and <90, but as far as I can tell the rules file only ever sets this signal to 0, 50, or 90. Unless I'm missing something, it will therefore never trigger.

Capture 27.PNG

Presumably the condition should be "Signal percentcomplete for the market = 50", and the <90 condition should be removed ???

[wearthefoxhat]

Sorry not.fully.read the thread.

My question / criticism would be why specifically those odds? Is there some.analysos behind it? Does it work all markets ?

Otherwise youre just trading randomly and will be eaten by comission and losing out a lot of profit locking in a trade for a short profit, but big losses when you dont match

Also, interesting download to replies ratio. Shame a lot of lurkers arent more active on the forums (appreciate some just wont wanna clog the thread with thank yous).

It was posted in the spirit of helping the OP to get a bot working and be able to tweak and improve upon. Much like the many Dallas has done over the last few years.

FWIW it's worth, In-play is awkward to do with a bot that opens and closes a trade. You're fighting the in-play delay and fast moving odds along with some that have faster pictures or use TPD.

IMO, the Bet Angel in-play trader is best to use, especially when L2B and used for the manual entry letting the software close out the trade. The key, as with any strategy, is to find the right type of runner(s) and the right odds range.