The BTC Horse Racing Thread
-
Nothing matched yesterday. A bit of slow horses action today
Also the Zero Paced lays are doing ok live trading as well. I'll start posting those after another couple of weeks monitoring.
-
No match yesterday. Just the one again today. Currently inside the 3.5-9 odds range. Going to have another bash at the short inrunning today. Cinvinced I can get those going
-
Not a lot going on today. Just the one potential Trob. 3.5-9 odds range only.
-
@kevin-pepper said in The BTC Horse Racing Thread:
@john-folan said in The BTC Horse Racing Thread:
Quite a good day yesterday. No dobs today. The shortinrunning has a massive list of potentials which can only be handled if you have a bot so I won't post.
Just the one Trob. 3.5-9 odds range
Hi John
Are the shortinrunning selections from your own filters or is there a set of BTC software filters? How are they doing?
Thanks
KevinThey are from the btc filters. I’ve got an issue with results not matching but Adams sorting that. I just load them in Bfbotmanager everyday and let it handle placing them if they match the odds filters. They are in 3 points in profit after a slightly iffy start(4points downturn). Just letting them run along on a 1% compounding of stakes.
Same with the short odd lays I mentioned a while back. They’ve gone from a 10 point initial drop to 17 points in profit. They are only able to be done on a bot basis as well.
Slow horses and zero paced dobs are poodling along at break even. Only expect 8-10 points a month from slow horses and 15-20 from the zero paced dobs. These don’t need to be bot based but it makes it easier on races with a few selections.
-
@john-folan said in The BTC Horse Racing Thread:
Quite a good day yesterday. No dobs today. The shortinrunning has a massive list of potentials which can only be handled if you have a bot so I won't post.
Just the one Trob. 3.5-9 odds range
Hi John
Are the shortinrunning selections from your own filters or is there a set of BTC software filters? How are they doing?
Thanks
Kevin -
Quite a good day yesterday. No dobs today. The shortinrunning has a massive list of potentials which can only be handled if you have a bot so I won't post.
Just the one Trob. 3.5-9 odds range
-
Trobs: Trob any between 3.5-9
Short Inrunning. Place a lay bet at 3.1 for any horse between 3.1 to 100. Bet persistence to keep in play
-
Jusy Short Inrunning and Slow Horses today
Slow Horses. Dob any over 5.0 as close to the off as possible.
Short Inrunning. Place a lay bet at 3.1 for any horse between 3.1 to 100. Bet persistence to keep in play
-
Much more manageable.
Trobs: Trob any between 3.5-9
Slow Horses. Dob any over 5.0 as close to the off as possible.
ShortInrunning
-
@martin said in The BTC Horse Racing Thread:
@john-folan said in The BTC Horse Racing Thread:
A general discussion.
Course Filters? What’s everyone’s view? I just stuck them on the shortinrunning strategy and my god the difference. Obviously to do with the style and shape of the course.
I’ve never been a fan before but thought I’d look seeing as everyone else is into them. I would like a strategy to be profitable without them first so there is less chance of a backfit.
Thoughts?
I like them but it's easy enough to run two versions one with course filtering and one without and see what performs best moving forward. I love the version part of the strategies you can split test multiple things extremely easily all within the same strategy!
Funny, I’ve literally started to split my strategies into different versions. Great isn’t it? Nice and easy to keep track of as well.
-
@john-folan said in The BTC Horse Racing Thread:
A general discussion.
Course Filters? What’s everyone’s view? I just stuck them on the shortinrunning strategy and my god the difference. Obviously to do with the style and shape of the course.
I’ve never been a fan before but thought I’d look seeing as everyone else is into them. I would like a strategy to be profitable without them first so there is less chance of a backfit.
Thoughts?
I like them but it's easy enough to run two versions one with course filtering and one without and see what performs best moving forward. I love the version part of the strategies you can split test multiple things extremely easily all within the same strategy!
-
@keith said in The BTC Horse Racing Thread:
@john-folan said in The BTC Horse Racing Thread:
@keith-driscoll said in The BTC Horse Racing Thread:
@john-folan If you understand 100% why something works at a specific track, then you can use a strategy with limited data that backs it up. For other strategies, using filter for courses can be hit-and-miss in some cases. Which is why it is worth doing a bit more research.
I would generally make a list of tracks that show a loss, and then look at them one-by-one. Usually, I would use 5 years of data. If they show 4 or 5 out of 5 years of loss, then I'd exclude it. Otherwise, I would keep it in the strategy. It does depend on how many races it is looking at though, 2-3 years data can be enough for some.
The software has enough data for AW tracks. Nearly enough for NH. It needs this season flat season to make course turf form meaningful. That all means you can use the course data for AW tracks, and extrapolate for NH. Ignore it for the turf till October.
Thanks that makes excellent sense. Kind of where I was wandering to with it. Adam does have about another 7-8 years of data he will be adding in. I’m going to do as you suggest though. I like a bit of research.
Can't wait to get this extra data!!
Very excited for the extra 8 years of data!!
-
@john-folan said in The BTC Horse Racing Thread:
@keith-driscoll said in The BTC Horse Racing Thread:
@john-folan If you understand 100% why something works at a specific track, then you can use a strategy with limited data that backs it up. For other strategies, using filter for courses can be hit-and-miss in some cases. Which is why it is worth doing a bit more research.
I would generally make a list of tracks that show a loss, and then look at them one-by-one. Usually, I would use 5 years of data. If they show 4 or 5 out of 5 years of loss, then I'd exclude it. Otherwise, I would keep it in the strategy. It does depend on how many races it is looking at though, 2-3 years data can be enough for some.
The software has enough data for AW tracks. Nearly enough for NH. It needs this season flat season to make course turf form meaningful. That all means you can use the course data for AW tracks, and extrapolate for NH. Ignore it for the turf till October.
Thanks that makes excellent sense. Kind of where I was wandering to with it. Adam does have about another 7-8 years of data he will be adding in. I’m going to do as you suggest though. I like a bit of research.
Can't wait to get this extra data!!
-
@keith-driscoll said in The BTC Horse Racing Thread:
@svajunas-bernotas-0 said in The BTC Horse Racing Thread:
just an idea, not to backfit is to create a strategy lets say using 2021 data and see how it performed in 2022
You can also do alternating months. You should never do all 20&21. Always use random years or months, and then see how the system works on all of them. Usually, if it fails that stage, I scrap it, then move on to research my next idea.
I like that. Nice idea.
I’ve never used them before. In two of mine there are obvious crap ones. The good thing about one of them is it is the polar opposite of Keith’s back the favourites as in laying the short odds ones and the courses where his are good, mine are shite.
As I said I would only use them to make a strategy that was profitable better and not try and polish a turd as it were.
If something isn’t profitable to the base idea I leave it
-
@svajunas-bernotas-0 said in The BTC Horse Racing Thread:
just an idea, not to backfit is to create a strategy lets say using 2021 data and see how it performed in 2022
You can also do alternating months. You should never do all 20&21. Always use random years or months, and then see how the system works on all of them. Usually, if it fails that stage, I scrap it, then move on to research my next idea.
-
@keith said in The BTC Horse Racing Thread:
@john-folan said in The BTC Horse Racing Thread:
A general discussion.
Course Filters? What’s everyone’s view? I just stuck them on the shortinrunning strategy and my god the difference. Obviously to do with the style and shape of the course.
I’ve never been a fan before but thought I’d look seeing as everyone else is into them. I would like a strategy to be profitable without them first so there is less chance of a backfit.
Thoughts?
It's a fine line between back fitting and not mate but after hearing @Keith-Driscoll speak about certain strategies for certain courses I'm wondering if there are systems that only work at specific courses due to the track itself?
I've used them in my systems I've posted on here bit I will be forward testing for a bit before going liveTies in nicely with what Keith said. I’m going to do some research. I think horses for courses is actually a thing tbh. It’s the same as the greyhounds.
-
@keith-driscoll said in The BTC Horse Racing Thread:
@john-folan If you understand 100% why something works at a specific track, then you can use a strategy with limited data that backs it up. For other strategies, using filter for courses can be hit-and-miss in some cases. Which is why it is worth doing a bit more research.
I would generally make a list of tracks that show a loss, and then look at them one-by-one. Usually, I would use 5 years of data. If they show 4 or 5 out of 5 years of loss, then I'd exclude it. Otherwise, I would keep it in the strategy. It does depend on how many races it is looking at though, 2-3 years data can be enough for some.
The software has enough data for AW tracks. Nearly enough for NH. It needs this season flat season to make course turf form meaningful. That all means you can use the course data for AW tracks, and extrapolate for NH. Ignore it for the turf till October.
Thanks that makes excellent sense. Kind of where I was wandering to with it. Adam does have about another 7-8 years of data he will be adding in. I’m going to do as you suggest though. I like a bit of research.
-
@keith said in The BTC Horse Racing Thread:
@john-folan said in The BTC Horse Racing Thread:
A general discussion.
Course Filters? What’s everyone’s view? I just stuck them on the shortinrunning strategy and my god the difference. Obviously to do with the style and shape of the course.
I’ve never been a fan before but thought I’d look seeing as everyone else is into them. I would like a strategy to be profitable without them first so there is less chance of a backfit.
Thoughts?
It's a fine line between back fitting and not mate but after hearing @Keith-Driscoll speak about certain strategies for certain courses I'm wondering if there are systems that only work at specific courses due to the track itself?
I've used them in my systems I've posted on here bit I will be forward testing for a bit before going livejust an idea, not to backfit is to create a strategy lets say using 2021 data and see how it performed in 2022
-
@john-folan If you understand 100% why something works at a specific track, then you can use a strategy with limited data that backs it up. For other strategies, using filter for courses can be hit-and-miss in some cases. Which is why it is worth doing a bit more research.
I would generally make a list of tracks that show a loss, and then look at them one-by-one. Usually, I would use 5 years of data. If they show 4 or 5 out of 5 years of loss, then I'd exclude it. Otherwise, I would keep it in the strategy. It does depend on how many races it is looking at though, 2-3 years data can be enough for some.
The software has enough data for AW tracks. Nearly enough for NH. It needs this season flat season to make course turf form meaningful. That all means you can use the course data for AW tracks, and extrapolate for NH. Ignore it for the turf till October.
-
@john-folan said in The BTC Horse Racing Thread:
A general discussion.
Course Filters? What’s everyone’s view? I just stuck them on the shortinrunning strategy and my god the difference. Obviously to do with the style and shape of the course.
I’ve never been a fan before but thought I’d look seeing as everyone else is into them. I would like a strategy to be profitable without them first so there is less chance of a backfit.
Thoughts?
It's a fine line between back fitting and not mate but after hearing @Keith-Driscoll speak about certain strategies for certain courses I'm wondering if there are systems that only work at specific courses due to the track itself?
I've used them in my systems I've posted on here bit I will be forward testing for a bit before going live
