Two winners. Same card. Both identified before racing began. March 24th at Southwell was exactly the kind of day that shows what a data-driven approach to horse racing tips can deliver — not luck, not a hunch, but a systematic process that flagged both horses before the market had fully caught up.
Here is how Horse Racing Oracle AI found them, and why the selections made sense before a single race was run.
The System Behind the Selections
Horse Racing Oracle AI analyses over 200 variables for every race on the card — going conditions, course and distance form, trainer and jockey strike rates, official ratings versus Racing Post Ratings, class movements, recent trajectory, and more. The output is not a list of every horse worth considering. It is a single NAP of the day: the one selection where the data alignment is strongest and the case is clearest.
On March 24th, that process pointed to Southwell — and it pointed there twice.
Shadowfax Of Rohan — NAP of the Day, 14:30 Southwell, Won at 1.83
The data case for Shadowfax Of Rohan was built around three signals that the system weighted heavily. First, the horse was returning off the same official rating of 86 that had seen him win his chase debut at Catterick — the handicapper had not raised him, meaning he was getting a second chance off a mark already proven to be winnable. Second, trainer Patrick Neville was operating at a 30% strike rate from recent runners, one of the highest in the country at that point in the season — a signal the system treats as meaningful evidence of deliberate, confident placement. Third, the field lacked depth at Class 5 level, meaning the rating advantage was not being offset by a stronger-than-usual opposition.
The result confirmed what the analysis suggested. Shadowfax Of Rohan won, delivering returns of £50 from a £25 stake at the boosted evens price and £23.72 profit from a £30 exchange single at 1.79.
Hillberry Hill — Same Card Bonus, 16:00 Southwell, Won at 1.72
Hillberry Hill was not the NAP — but the system's wider card analysis flagged it as a horse in an unusually weak race for its class. Having won a maiden hurdle in Ireland, Hillberry Hill was carrying a penalty into a Southwell novices' hurdle that the Racing Post's own racecard described as a very weak race to do so in. When the opposition is assessed as significantly below the quality of the selection, that is a signal worth acting on. It won comfortably.
Those who combined both selections into a double saw a £20 stake return £63.33 at combined odds of 3.17.
Why Data Beats Gut Feel
Both selections were identified through process, not instinct. The yard strike rate, the handicap mark, the class of the opposition, the going conditions — none of these are things a casual punter checks systematically for every race every day. That is exactly what Horse Racing Oracle AI does, automatically, before publishing a single tip.
The system does not win every day. Horse racing never allows that. But by tracking the variables that matter most and identifying the races where the evidence aligns most clearly, it consistently surfaces selections that the market undervalues — and yesterday was a clean example of that process working exactly as it should.
Get Tomorrow's NAP Before the Market Moves
The Horse Racing Oracle AI NAP is published every morning before racing begins. By the time the market shortens on the selections the system has already identified, it is too late to get the best price. Sign up free and get the tip in your inbox before the bookmakers catch up.
Want free AI-powered tips every morning? Sign up free at horseracingoracleai.com →
Gambling involves risk. Only bet what you can afford to lose.