Performance To Date
(since 1st August 2026)
day
84
7
losers
13
winners
557
£2,327
£327
+16.4%
bank
|
97.7%
selection accuracy
77
winning streak
(days)
winning streak
(days)
Free Tips
Subscription Model
Don't try to beat the bookmakers. Copy them!
I wanted to make money from gambling (it's tax-free in the UK), but I'm not an expert at horse racing, nor do I want to spend hours every day analysing form and statistics to make selections. I wanted a system to gamble without emotion.
No 'gut feelings', no wild picks, no betting on horses with a funny name. I call this 'emotional betting' and the big winners are the bookmakers, not us. I wanted a computer to do the picking for me. Unemptionally.
Consistently picking winners to make a profit is difficult.
Bookmakers make millions every year.
Picking losers is much easier.
I want to copy the bookmakers.
With AI's help, I created a mathematical model that does just that.
I fed over 150,000 results from two years of UK and Irish horse races into ChatGPT and asked it to give me criteria that I could use to select horses with a high probability of losing. ChatGPT analysed the data and produced a set of criteria in less than 60 seconds! I feed the day's races into a spreadsheet and using the ChatGPT criteria, the model makes the selections that I place lay bets on. This is the information that I give to my subscribers.
Performance Summary
So, how is the model doing?
September 2026 was good...
-
Starting bank: £2,000
-
Ending bank: £2,235
-
Profit: £235 (11.8%)
-
No winners: 204
-
No. losers: 6
-
Selection accuracy: Good at 97.1
The right chart above shows the monthly profit return.
This shows you the ROI% each month. Consider than banks will pay you 4-5% per year... but banks are 100% safe, we are gambling and can lose everything.
Earlier History...
I've been running the model since 1st January 2026 and it has performed well each month, but I took profits at the end of July 2026, so reset everything to start again on 1st August with £2,000.
-
January 2026. An 18% profit, making £247.
-
February 2026. Terrible month, with just 2% profit, so I made alterations to the model.
-
March 2026. Better with a 14% profit, making £124.
-
April 2026. Best return with a 30% profit, making £303.
-
May 2026. The model made a 14% profit. I deposited £5,000 into the bank, so risking and making larger amounts, which equated to £812.
-
June 2026. Good returns continued with a 16% profit, making £824.
-
July 2026. More good returns with a 20% profit. First ever 100% selection accuracy with 149 losers and 0 winners. We made £1,152. I had lost my job in April, so needed to withdraw some money to pay the bills. The bank dropped from £7,000 to £2,000 at month end.
-
August 2026. 6% profit making just £66, with 188 losers and 4 winners (97.9% accuracy)
-
September 2026. 12% profit making £235, with 204 losers and 6 winners (97.1% accuracy)
My personal goal is to generate an average of £300/day or £9,000/mo, with no tax due (in the UK). That pays all my monthly bills, the mortgage, money for food, petrol and football, and the never-ending bills of having an 18-year-old daughter who likes changing her hair every 4 days and wearing the same clothes outside twice is a sin!
Don't forget...
This is gambling. You can certainly lose, as the figures above show, I have losing days.
Just because the model has made a profit every month this year, past performance is no guarantee of future profit.
Expect to lose money some days.
Please gamble responsibly, only risking money you can afford to lose.
This model is definitely a 'Get Rich Slow' scheme. You need to be patient.
Think of it like a medium-term investment.
If the stockmarket returns 10-15% a year, that's a really good year.
Banks will offer you 4-5% a year on savings at the moment.
Since 01JAN 2026, our AVERAGE RETURN PER MONTH has been 12%. That's per month, not per year.
But it's still no guarantee of future performance.
The left chart above shows our daily profit total.
This shows you how much money we have made since 01AUG 2026 using the same tips we send to subscribers.