Bot Optimisation - The Basics

We've gone to the dogs.
Post Reply
Nacho
Posts: 5
Joined: Wed May 06, 2020 6:10 pm
Location: Scotland

Hi everyone, 

Cool to see frequent posts in the Greyhound section. I'm at the stage of starting to trial some systems on modest stakes and I wonder if someone would be kind enough to answer a few basic questions that I have. In my case I am looking at UK only and laying the more liquid, shorter-priced runners.

Will "take SP" leak value compared to matching pre-off or is it a legitimate approach?   

To avoid leaking value -what is the optimal time to lay? 10 seconds, 30 seconds, 60 seconds?

To avoid leaking value  - ought I to take available prices or offer price on opposite side of the book?

Is dripping in stake recommended over a certain size? 


Basically, I'd be grateful for any "best practice" tips to keep the efficiency as high as possible. 

Thanks in advance to anyone taking the time to answer. :)

Cheers,

Nacho
User avatar
gazuty
Posts: 2547
Joined: Sun Jun 26, 2011 11:03 am
Location: Green land :)

Nacho wrote:
Sat Jan 09, 2021 4:12 am
Will "take SP" leak value compared to matching pre-off or is it a legitimate approach?   


Take SP is a legitimate approach in my view. Or you could choose not to green out with SP and let things ride. Do not go in play.

Nacho wrote:
Sat Jan 09, 2021 4:12 am
To avoid leaking value -what is the optimal time to lay? 10 seconds, 30 seconds, 60 seconds?


You will need to experiment to determine optimal time, might change by event and be dependent on velocity of money in the event.

Nacho wrote:
Sat Jan 09, 2021 4:12 am
To avoid leaking value  - ought I to take available prices or offer price on opposite side of the book?


Offer on the opposite side of the book (or take the very last of the available price before it moves to the other side of the book). It depends how much money is left and what is happening.

Nacho wrote:
Sat Jan 09, 2021 4:12 am
Is dripping in stake recommended over a certain size? 
Yes because every action you take impacts the market. Dump in a huge order the market will react. Dump in a small order the market will still react but probably not by as much and perhaps in a much more subtle way.
User avatar
The Silk Run
Posts: 902
Joined: Mon May 14, 2018 12:53 am
Location: United Kingdom

Hi Hi Nacho

"In my case I am looking at UK only and laying the more liquid, shorter-priced runners"

Have you considered Australian greyhound markets. Opposing favourites 1.20 - 1.80 with filtering is a good START. I used to use soft recovery, but level staking is fine ...
spreadbetting
Posts: 3140
Joined: Sun Jan 31, 2010 8:06 pm

Make sure you check any tweaks before letting them run unattended too.

Did a few tweaks last night to one of mine let it loose on the aus stuff and went to bed :x

Untitled.png
Back to the drawing board for that one.
Nacho
Posts: 5
Joined: Wed May 06, 2020 6:10 pm
Location: Scotland

Thanks very much indeed to you guys for replying.

@ spreadbetting - s&%t, that is a brutal hit. Respect for posting the screenshot. You seem like an experienced trader but that's a bitter pill to swallow for anyone. I'll definitely heed your warning.

@gazuty - cheers for the detailed answers. Everything you say makes sense. I should have clarified that I am looking at straight lays, albeit filtered ones, so nothing too complicated. I think I will run several bots on identical strategies but different entry criteria (one of them taking SP) so as to be able to compare. I am wary of a lack of liquidity in the Greyhound markets and you are right to remind me of the basic point that liquidity will vary between events. I guess I wondered - for people with proven profitable lay systems - what size of straight lay is the maximum before problems start to arise. I won't be betting with anything more than testing stakes to start with regardless. I think I saw Peter mention once that he found there was a kind of staking tipping point.

@Silk Run - I have basic bots on greyhounds for a couple of months including Australia. However, like many people, I recently stitched together the 2020 and 2019 BSP data and have spent more time looking at the UK data. It seems to me that laying favourites offers a better starting point in the UK than down under. However, I will take your advise and look into it. :)

As for recovery staking, I have tried this too but I want to concentrate on developing edges at level stakes. Recovery systems might work well with a solid edge but it would seem to me that heightened commission would be a big downside.

Thanks again for the responses.
rik
Posts: 1583
Joined: Sat Jan 25, 2014 5:16 am
Location: London

spreadbetting wrote:
Sat Jan 09, 2021 1:01 pm
Make sure you check any tweaks before letting them run unattended too.

Did a few tweaks last night to one of mine let it loose on the aus stuff and went to bed :x


Untitled.png

Back to the drawing board for that one.
sorry, i have lower exposure limit on the accounts i run automation and dont let the balance run up higher than the bots needs or is likely to lose over the day/week that should stop it misfiring badly
Anbell
Posts: 2004
Joined: Fri Apr 05, 2019 2:31 am

spreadbetting wrote:
Sat Jan 09, 2021 1:01 pm
Make sure you check any tweaks before letting them run unattended too.

Did a few tweaks last night to one of mine let it loose on the aus stuff and went to bed :x


Untitled.png

Back to the drawing board for that one.
That's most unkind.
spreadbetting
Posts: 3140
Joined: Sun Jan 31, 2010 8:06 pm

Woke up to another -£450 this morning :(

Thought I'd fixed it yesterday but thankfully at least the logging I'd added identified where it was messing up. Fingers crossed things will now be back to normal service. It'll teach me not to tweak code without fully testing or thinking it thru.
User avatar
gazuty
Posts: 2547
Joined: Sun Jun 26, 2011 11:03 am
Location: Green land :)

spreadbetting wrote:
Sun Jan 10, 2021 1:11 pm
Woke up to another -£450 this morning :(

Thought I'd fixed it yesterday but thankfully at least the logging I'd added identified where it was messing up. Fingers crossed things will now be back to normal service. It'll teach me not to tweak code without fully testing or thinking it thru.
Yikes. Glad you’ve identified the problem.
spreadbetting
Posts: 3140
Joined: Sun Jan 31, 2010 8:06 pm

Back to winning ways this morning so looking good, will start upping the stakes gradually over the week as Monday is never that big a day. I know where I'd cocked up the code so should be fine.
puntinrolls
Posts: 70
Joined: Mon Aug 31, 2020 3:01 pm

gazuty wrote:
Sat Jan 09, 2021 10:36 am
Do not go in play.
I was asking similar questions as OP in an other thread and this has helped me also. Thanks.

I'm confused, what do you mean about not going in play, I thought you can't go in play with the dogs
User avatar
gazuty
Posts: 2547
Joined: Sun Jun 26, 2011 11:03 am
Location: Green land :)

puntinrolls wrote:
Fri Jan 15, 2021 9:30 pm
I'm confused, what do you mean about not going in play, I thought you can't go in play with the dogs
You can't, but then I am not to know if OP strategy is restricted to dogs or also to horses.
puntinrolls
Posts: 70
Joined: Mon Aug 31, 2020 3:01 pm

gazuty wrote:
Sat Jan 16, 2021 1:41 am
puntinrolls wrote:
Fri Jan 15, 2021 9:30 pm
I'm confused, what do you mean about not going in play, I thought you can't go in play with the dogs
You can't, but then I am not to know if OP strategy is restricted to dogs or also to horses.
Thanks, just wasn't sure I was missing something
Nacho
Posts: 5
Joined: Wed May 06, 2020 6:10 pm
Location: Scotland

This is perhaps not the correct place to post but as it carries on from my efforts to design a greyhound bot I will go ahead.

I successfully stitched together all the Betfair CSV results for 2019, and 2020 (using the downloader tool that a forum user had kindly provided). Both are looking pretty good and I have found some angles which look promising.

However, the 2018 files are jumbled after I merge them (mixing across columns). I wondered if anyone had found a solution for this?

Any suggestions gratefully received. :)
Post Reply

Return to “Trading Greyhound racing”