Thanks for the reply Peter. I have identified a type of race that does not work with the strategy that i deploy so i knocked that on the head and now i am producing small but consistent profits at £12 stakes. I could scale it up by x 10 and would still get matched easily. I only need to win 40% of races which is basically what is happening so i have decided to take on board your comment of eliminating losses before i ramp up the stakes. i'm going to try and get my strike rate up ( ie eliminate more losses ) before i ramp up the stakes.
Just one last question - why have you stopped your strategy if its a winner ? What made you think it would not carry on indefinetly ?
Building A Strategy
-
- Posts: 8
- Joined: Fri May 10, 2013 3:25 pm
after only having the software a short while now I would have to say that I think the automation function is a fantastic tool which can lead to all sorts of possibilities,it was suggested to me when looking for some help setting up a rule that I should take time to gain knowledge of the software and the automation and to access the practice mode for result finding that would justify my progress,"thanks again rodger" but in reading this reply it struck me that my first mistake was not learning enough about betfair ,so this was my next port of call I have simply examined as much resource,feedback,instruction,and information regarding trading within the set limitations of betfair as I could and I have to say that it simply opened the door to how my approach to automation would start and progress ,some of the theseis papers written on betfair are remarkable and very insightful this has helped me make decisions so much quicker and given me independance of thought when writing down on a sheet of paper exactly what i want to create ,test,and hopefully make work for me yes i am a newbie and i know my opionion will not mean alot but reading the forum posts will relate a feeling of mixed message from newbies,experianced and people who have given up and people who have been successful all i wanted to say to anyone starting new is learn the size of the pitch and its boundarys before you get the ball because it has helped me ,yes ,helped me dont look for answers look for help ones given the other is earned
Your 100% right there. I worked this out sometime ago and when i did it became the key for me. No-one that has worked hard on a strategy is going to tell you what it is and why should they ?guidedog777 wrote: dont look for answers look for help ones given the other is earned
Some of the pro's on here are very generous with their posts - they actually do provide the answers but you have listen and then work out what they are for yourself. Mugsgame , Jolly Green , Peter Le and particularily Euler all offer incredible insight into what it takes to be a proffessional - all you have to do is listen and apply a tremendous amount of hard work - you get nothing for nothing.
Seek and ye shall find.................
It actually becomes quite enjoyable when you start building something from the ground up. Its like nurturing a plant from seed!
Yesterday I was running a "Lay to back" system that I had been tweaking on and off for ages..
It waits for an awful lot of conditions to be met before it It places a lay and then offsets looking for large odds. The first race I tried the new version on was at Windsor, it laid a horse called "Cafetiere" for a combined total of 9.90 at the bottom of its range and then managed to get £4 matched in the offset. It doesn't use greening
I didn't catch it right at the top, but got £4 matched. According to Timeform it traded at 200 at the top (although looking at the video i recorded I thought it peeked around 450?
I can now use that profit to develop the system risk free
You only need a couple of these a month
Yesterday I was running a "Lay to back" system that I had been tweaking on and off for ages..
It waits for an awful lot of conditions to be met before it It places a lay and then offsets looking for large odds. The first race I tried the new version on was at Windsor, it laid a horse called "Cafetiere" for a combined total of 9.90 at the bottom of its range and then managed to get £4 matched in the offset. It doesn't use greening
I didn't catch it right at the top, but got £4 matched. According to Timeform it traded at 200 at the top (although looking at the video i recorded I thought it peeked around 450?
I can now use that profit to develop the system risk free
You only need a couple of these a month
I'm pretty sure some people dont trade tracks like Wolverhampton due to the extreme volatility. Runners can be both big steamers and big drifters all in the space of a few minutes which ruins the approach of " trend " traders. After a while i suppose you work out which tracks suit your style and which dont and it makes sense to not trade the ones that dont fit.
Ive pretty much mastered BA bot creation, aside from using spreadsheets yet anyway. I have made bots before for others things such as virtual dogs and the dating nice so I had that as an advantage in learning Betangel.
Can be a bit frustrating sometimes. I'm not doing it to win money however. I just like making bots and I think thats the best attitude to have. Then if I do win money down the road that is just a bonus.
However, I do see big potential to make a lot of money still.
I bet on a dog, I think it was a dog, called Harambe. 2nd fav @ 14/1. I just knew it would win. SImply because it that price was out of place looking. I meant to back for 70 pence but I entered £70 by mistake. I back out back down to the original amount I wanted but of course it came in. LOL
So, I can so the potential but I am not expecting to make any money per se. Just takes a lot of testing, patience and a bit of luck. Some people will get their lucky break some wont.
I find in general that the more complex I make the bots the worse they do rather than the opposite. Since they start to rely more on whether they will match or not, ETC. As I said on another post, I think it's better to have an idea first then make the bot for it but I guess some people have done it the other way around with success and it is fun making them anyway.
Can be a bit frustrating sometimes. I'm not doing it to win money however. I just like making bots and I think thats the best attitude to have. Then if I do win money down the road that is just a bonus.
However, I do see big potential to make a lot of money still.
I bet on a dog, I think it was a dog, called Harambe. 2nd fav @ 14/1. I just knew it would win. SImply because it that price was out of place looking. I meant to back for 70 pence but I entered £70 by mistake. I back out back down to the original amount I wanted but of course it came in. LOL
So, I can so the potential but I am not expecting to make any money per se. Just takes a lot of testing, patience and a bit of luck. Some people will get their lucky break some wont.
I find in general that the more complex I make the bots the worse they do rather than the opposite. Since they start to rely more on whether they will match or not, ETC. As I said on another post, I think it's better to have an idea first then make the bot for it but I guess some people have done it the other way around with success and it is fun making them anyway.
- ruthlessimon
- Posts: 2094
- Joined: Wed Mar 23, 2016 3:54 pm
What about if the coding was predictable?
A 2.0 fav will win, lose, win, lose, win, lose etc
There was a study done, where professional statisticians had to pick random data from non-random - & they were consistently biased by "non-streaky data" (i.e. they consistently assumed data with streaks was non-random). Perhaps the coders on the virtual races are too, never know
It’s all audited:ruthlessimon wrote: ↑Mon Jul 23, 2018 5:45 pmWhat about if the coding was predictable?
A 2.0 fav will win, lose, win, lose, win, lose etc
There was a study done, where professional statisticians had to pick random data from non-random - & they were consistently biased by "non-streaky data" (i.e. they consistently assumed data with streaks was non-random). Perhaps the coders on the virtual races are too, never know
http://www.gamblingcommission.gov.uk/fo ... oring.aspx
You always have to throw a spanner in the works!ruthlessimon wrote: ↑Mon Jul 23, 2018 5:45 pmWhat about if the coding was predictable?
A 2.0 fav will win, lose, win, lose, win, lose etc
There was a study done, where professional statisticians had to pick random data from non-random - & they were consistently biased by "non-streaky data" (i.e. they consistently assumed data with streaks was non-random). Perhaps the coders on the virtual races are too, never know
Virtual racing is bound to be randomised with the odds against you and of all the available sports where it's possible to win to choose from, why pick virtual racing?
- ruthlessimon
- Posts: 2094
- Joined: Wed Mar 23, 2016 3:54 pm
I wouldn't do it because, BF would quickly plug any hole in a dodgy algorithm. & the effort it'd take to understand how the algo's worked would be a complete nightmare.
But I am fascinated by virtual markets - literally, thousands can be made in excel, which can then be compared to the real data - & where they differ, is arguably, very interesting "seed" info