• The issue has been fixed and hopefully you can post screen shots again
  • Hi Guest Just in case you were not aware I wanted to highlight that you can now get a free 7 day trial of Horseracebase here.
    We have a lot of members who are existing users of Horseracebase so help is always available if needed, as well as dedicated section of the fourm here.
    Best Wishes
    AR

Horse Racing Data, Strategy, and AI Harnessing

Here is what greed looks like.

Of course, I didn't change the trading parameters, and the trades were running without a stop loss — and of course, one such losing trade occurred. Instead of a €99.95 profit, I ended up with a €0.06 loss.

Lesson learned.

I have been working on different prompts, adding those two new data sources.

Tomorrow I will test the LLM prompt and give it free rein to write whatever it thinks is the better approach — it suggested a trailing stop loss.

1791316342607.png
 
Today was the first day of running my strategy, the code for which was generated by the prompt I had been testing for a couple of days.

Even though in the prompt I asked, "The data is the deliverable — use all of it," the generated code told a different story; but what is more important is that the strategy generates profit.

1791736626521.png
 
Kotari ~ 14.50 | Loser ~ 0.86
Gortmore Lady ~ 10.00 | Loser ~ 0.71
Dawnmore ~ 7.00 | Loser ~ 2.20
Gili Island ~ 36.00 | Loser ~ 20.00
Our Sara ~ 18.00 | Loser ~ 16.13
Persian Land ~ 30.00 | Loser ~ 15.38
Drop Dead Gorgeous ~ 14.00 | Loser ~ 0.93
Final Entry ~ 10.00 | Loser ~ 16.28
Sudden Flight ~ 23.00 | Loser ~ 0.43
Snapretend ~ 46.00 | Winner ~ 15.00
Royal Acclamation ~ 1.79 | Winner ~ 4.07
Abbotsford ~ 19.50 | Winner ~ 18.21
Gower Prince ~ 4.70 | Winner ~ 4.11

My trading profit: 130.59
When it bets, it makes a profit: 5799.00 | 4/14
 
It seems my trading strategy is picking winners. What would you do, switch to betting or keep trading?

Any new strategy I try, I run as a trading strategy, since psychologically I don't hate to lose.


Trade or Bet Profit:

Kamille Au Kanada ~ 5.00 | Loser ~ 16.28
Kotari ~ 14.50 | Loser ~ 0.86
Gortmore Lady ~ 10.00 | Loser ~ 0.71
Dawnmore ~ 7.00 | Loser ~ 2.20
Gili Island ~ 36.00 | Loser ~ 20.00
Our Sara ~ 18.00 | Loser ~ 16.13
Persian Land ~ 30.00 | Loser ~ 15.38
Drop Dead Gorgeous ~ 14.00 | Loser ~ 0.93
Final Entry ~ 10.00 | Loser ~ 16.28
Sudden Flight ~ 23.00 | Loser ~ 0.43
Snapretend ~ 46.00 | Winner ~ 15.00
Royal Acclamation ~ 1.79 | Winner ~ 4.07
Abbotsford ~ 19.50 | Winner ~ 18.21
Gower Prince ~ 4.70 | Winner ~ 4.11
Karrygrant ~ 4.80 | Loser ~ 17.07
Motawaared ~ 2.66 | Loser ~ 9.68
Olympiacos ~ 15.00 | Loser ~ 15.38
Contrary To Law ~ 6.40 | Winner ~ 2.31
Moonlightinthepark ~ 16.00 | Loser ~ 0.96
Status Symbol ~ 9.20 | Loser ~ 1.17
Goldrush Kid ~ 5.50 | Loser ~ 1.30
Astracornus ~ 5.80 | Loser ~ 2.94
Much ~ 6.20 | Loser ~ 2.93
Zoustar Dreams ~ 4.80 | Loser ~ 3.91
Siam Park ~ 6.00 | Loser ~ 2.71
Keep Me Stable ~ 13.50 | Loser ~ 1.01
Lauderdale King ~ 4.70 | Loser ~ 4.12
Muskoka ~ 11.00 | Loser ~ 17.02
Alberich ~ 4.00 | Loser ~ 4.41
Alpine Sierra ~ 5.60 | Loser ~ 3.20
Goomah ~ 9.80 | Loser ~ 1.61
Inver Bay ~ 9.40 | Winner ~ 1.80
Siouxperb ~ 4.30 | Winner ~ 4.65
Juan Les Pins ~ 12.50 | Winner ~ 1.20
Archer Royal ~ 12.50 | Loser ~ 1.20

Trade Profit: 231.18
Bet Profit: 6959.00 | 8/35

Yes, a couple of days of trading is nothing from a statistical point of view; it may just be luck.

The betting profit is mostly influenced by the horse: Snapretend ~ 46.00
 
What i am trying to do today is anyrca ei have picked one or two horses out if i dont like the fav dutch them looking for odds of no less that 4/1 on any one horse.
 
StefanBelo said:
It seems my trading strategy is picking winners. What would you do, switch to betting or keep trading?

Any new strategy I try, I run as a trading strategy, since psychologically I don't hate to lose.


Trade or Bet Profit:

Kamille Au Kanada ~ 5.00 | Loser ~ 16.28
Kotari ~ 14.50 | Loser ~ 0.86
Gortmore Lady ~ 10.00 | Loser ~ 0.71
Dawnmore ~ 7.00 | Loser ~ 2.20
Gili Island ~ 36.00 | Loser ~ 20.00
Our Sara ~ 18.00 | Loser ~ 16.13
Persian Land ~ 30.00 | Loser ~ 15.38
Drop Dead Gorgeous ~ 14.00 | Loser ~ 0.93
Final Entry ~ 10.00 | Loser ~ 16.28
Sudden Flight ~ 23.00 | Loser ~ 0.43
Snapretend ~ 46.00 | Winner ~ 15.00
Royal Acclamation ~ 1.79 | Winner ~ 4.07
Abbotsford ~ 19.50 | Winner ~ 18.21
Gower Prince ~ 4.70 | Winner ~ 4.11
Karrygrant ~ 4.80 | Loser ~ 17.07
Motawaared ~ 2.66 | Loser ~ 9.68
Olympiacos ~ 15.00 | Loser ~ 15.38
Contrary To Law ~ 6.40 | Winner ~ 2.31
Moonlightinthepark ~ 16.00 | Loser ~ 0.96
Status Symbol ~ 9.20 | Loser ~ 1.17
Goldrush Kid ~ 5.50 | Loser ~ 1.30
Astracornus ~ 5.80 | Loser ~ 2.94
Much ~ 6.20 | Loser ~ 2.93
Zoustar Dreams ~ 4.80 | Loser ~ 3.91
Siam Park ~ 6.00 | Loser ~ 2.71
Keep Me Stable ~ 13.50 | Loser ~ 1.01
Lauderdale King ~ 4.70 | Loser ~ 4.12
Muskoka ~ 11.00 | Loser ~ 17.02
Alberich ~ 4.00 | Loser ~ 4.41
Alpine Sierra ~ 5.60 | Loser ~ 3.20
Goomah ~ 9.80 | Loser ~ 1.61
Inver Bay ~ 9.40 | Winner ~ 1.80
Siouxperb ~ 4.30 | Winner ~ 4.65
Juan Les Pins ~ 12.50 | Winner ~ 1.20
Archer Royal ~ 12.50 | Loser ~ 1.20

Trade Profit: 231.18
Bet Profit: 6959.00 | 8/35

Yes, a couple of days of trading is nothing from a statistical point of view; it may just be luck.

The betting profit is mostly influenced by the horse: Snapretend ~ 46.00
Click to expand...
Don't just watch it, bet small stakes and monitor.
Well done by the way.
 
gerry said:
What i am trying to do today is anyrca ei have picked one or two horses out if i dont like the fav dutch them looking for odds of no less that 4/1 on any one horse.
Click to expand...
In these races i looked at 12 races today where i thought could be tough for fav and out of those 12 races the fav only won one of those races which made it possible to do well but surprising 4 of the races a big price horse won it out with acceptable but the other 7 races with my dutching i backed 6 winners i did in theory hope to get 7 winners dutching 3 per race to show good profit, but did show profit just the same and would i have shown profit betting normal i honestly think no chance today.
 
StefanBelo said:
It seems my trading strategy is picking winners. What would you do, switch to betting or keep trading?

Any new strategy I try, I run as a trading strategy, since psychologically I don't hate to lose.


Trade or Bet Profit:

Kamille Au Kanada ~ 5.00 | Loser ~ 16.28
Kotari ~ 14.50 | Loser ~ 0.86
Gortmore Lady ~ 10.00 | Loser ~ 0.71
Dawnmore ~ 7.00 | Loser ~ 2.20
Gili Island ~ 36.00 | Loser ~ 20.00
Our Sara ~ 18.00 | Loser ~ 16.13
Persian Land ~ 30.00 | Loser ~ 15.38
Drop Dead Gorgeous ~ 14.00 | Loser ~ 0.93
Final Entry ~ 10.00 | Loser ~ 16.28
Sudden Flight ~ 23.00 | Loser ~ 0.43
Snapretend ~ 46.00 | Winner ~ 15.00
Royal Acclamation ~ 1.79 | Winner ~ 4.07
Abbotsford ~ 19.50 | Winner ~ 18.21
Gower Prince ~ 4.70 | Winner ~ 4.11
Karrygrant ~ 4.80 | Loser ~ 17.07
Motawaared ~ 2.66 | Loser ~ 9.68
Olympiacos ~ 15.00 | Loser ~ 15.38
Contrary To Law ~ 6.40 | Winner ~ 2.31
Moonlightinthepark ~ 16.00 | Loser ~ 0.96
Status Symbol ~ 9.20 | Loser ~ 1.17
Goldrush Kid ~ 5.50 | Loser ~ 1.30
Astracornus ~ 5.80 | Loser ~ 2.94
Much ~ 6.20 | Loser ~ 2.93
Zoustar Dreams ~ 4.80 | Loser ~ 3.91
Siam Park ~ 6.00 | Loser ~ 2.71
Keep Me Stable ~ 13.50 | Loser ~ 1.01
Lauderdale King ~ 4.70 | Loser ~ 4.12
Muskoka ~ 11.00 | Loser ~ 17.02
Alberich ~ 4.00 | Loser ~ 4.41
Alpine Sierra ~ 5.60 | Loser ~ 3.20
Goomah ~ 9.80 | Loser ~ 1.61
Inver Bay ~ 9.40 | Winner ~ 1.80
Siouxperb ~ 4.30 | Winner ~ 4.65
Juan Les Pins ~ 12.50 | Winner ~ 1.20
Archer Royal ~ 12.50 | Loser ~ 1.20

Trade Profit: 231.18
Bet Profit: 6959.00 | 8/35

Yes, a couple of days of trading is nothing from a statistical point of view; it may just be luck.

The betting profit is mostly influenced by the horse: Snapretend ~ 46.00
Click to expand...
Keep trading, the sample is way too small and just to add to this I have not found a model that predicts IR price droppers that wins to BFSP
 
My system works autonomously, so any betting or trading is executed without my attention.

When developing a new strategy, I'm used to checking online betting/trading: while sitting at my computer and working on code, I see how the Betfair market reacts to my bets and, of course, watch the price movement on the ladders.

Here the main approach is trading, and new ideas come up around trade execution, how to set profit/loss targets, or how to dynamically change the target as the race-off approaches, since on many trades the profit target was set too high, so such execution could still close a profitable position without the risk of going in-play.
 
Today I have added agentic support to my "Horse Racing Browser" tool, so when adding tipsters' selections I can query the data and analyze it, adding a new data source to the base data the tool is using.

This allows me to find what patterns tipsters might use to select a horse for betting, as this approach is very important in my trading strategy, where selections are filtered by simple rules so that when the rules are met my trading strategy can be executed — on every race, which is not desired.

In the screenshot we can see how a simple prompt is processed by the LLM: "Get data context AllRacesHorsesData, and in the Tipsters field count which tipster has how many selections."

AllRacesHorsesData.png
 
No more racing for today. From the 14 selections I took from 2–3 threads on the forum, Chesham Chesham and Cashelman Cashelman found the winner Trustintimes, which was also my model's best selection at 16.82%.

Congratulations to Tufnel Tufnel, whose selection George Wickham was only 5.35% by my model. This selection was not such a clear one, as we all may use similar data; for now I do not know why it was selected by Tufnel, and my AI is just trying to analyze all the raw data to see whether we missed something in NRSStrategyEngine.
 
StefanBelo said:
No more racing for today. From the 14 selections I took from 2–3 threads on the forum, Chesham Chesham and Cashelman Cashelman found the winner Trustintimes, which was also my model's best selection at 16.82%.

Congratulations to Tufnel Tufnel, whose selection George Wickham was only 5.35% by my model. This selection was not such a clear one, as we all may use similar data; for now I do not know why it was selected by Tufnel, and my AI is just trying to analyze all the raw data to see whether we missed something in NRSStrategyEngine.
Click to expand...
Thanks. I also put Trustintimes up.

George Wickham had the best form LTO at the distance and in higher class. Was only carrying 1 lb more today on the same course. Also figured in the consistency and Erin probables.
 
App looks excellent but I'm afraid I'm a Linux guy. Very much interested in any collaboration of ideas.
I have researched pretty much everything I can think of to research across many sports and I always end up at the same point - the market still has this "unknown" that you can't model for consistently across multiple years.
Even if you run to de-vigged markets, given enough time the market will still win.

That's the game though right?
 
Interesting what you just said A Ajit the market still has this unknown as that's what exactly it does have the known and that along with the hidden value is exactly what i have been running the last week i started it on last saturday what a dangerous day to start something but i thought well its usually the toughest day for me why not and i am happy to say that just passed my 8 days so far with yet a other saturday which i felt would tell me a lot 8 days with two saturdays in there and happily past my first hundred bets with it and nice profit.
So i would say if the next hundred do the same stats i will be delighted and your right the market is solid crack that might crack racing .
 
Thanks, Tufnel Tufnel, for the additional information. I run 4 strategies on different data, and actually what you describe is available in my data; unfortunately, yesterday I tested a different strategy engine.

On the other hand, for now I trade the selections my model suggests, and in this race no trade was executed.
 
A Ajit, the Betfair market reflects different opinions and different knowledge in the odds/prices that we may not know about, and actually yes — in George Wickham's case, the Betfair market accumulated this information. So, as for the price, my model evaluated all my data, and the market said -EV, so the market was already favouring this horse against my model.

George%20Wickham_Winner.png
 
StefanBelo said:
Today I have added agentic support to my "Horse Racing Browser" tool, so when adding tipsters' selections I can query the data and analyze it, adding a new data source to the base data the tool is using.

This allows me to find what patterns tipsters might use to select a horse for betting, as this approach is very important in my trading strategy, where selections are filtered by simple rules so that when the rules are met my trading strategy can be executed — on every race, which is not desired.

In the screenshot we can see how a simple prompt is processed by the LLM: "Get data context AllRacesHorsesData, and in the Tipsters field count which tipster has how many selections."

AllRacesHorsesData.png
Click to expand...
I simply don't understand the logic here StefanBelo StefanBelo , you're describing members as tipsters and choosing to call selections for the race of the day as tips worth recording for some reason or other, even accounting for those that name 3 horses for the same race, there's little or no value in any of it.
I've already made the point elsewhere that the race of the day shouldn't be classed as a real tip unless somebody specifically says so, reason being obvious in that people are simply encouraged to take part by throwing in a selection for a horse or horses that they have no intention of backing themselves so why would you want to use them for any true intent on behalf of the partisipants ?
 
You must log in or register to reply here.
Back
Top