Festival weeks in horse racing are where reputations are made and lost. The Dante Festival at York draws some of the finest Flat horses in training, fields tighten up, and the margin for error shrinks. It is precisely the kind of week that separates a disciplined, data-driven approach from gut-feel guessing. The week of May 12–16 was a clean demonstration of what Horse Racing Oracle AI is built to do.
Three Winners Across Five Days
The week opened at York on May 13 with Legacy Link in the 15:05. This was not a straightforward win selection — the play was structured as a back win and each-way combination, which is exactly the kind of nuanced staking the system recommends when the form points to a horse likely to place at minimum but with genuine each-way upside. The return came in at +£68.55. A strong start to a big week.
May 14 brought See The Fire to the 14:30 at York — Dante Festival day itself, with the stands packed and the market efficient. Odds of 2.34 were available, and on a £50 stake the return landed at +£67.21. This is where the system earns its credibility. Identifying a winner at 2.34 on one of the highest-profile racing days of the Flat calendar is not lucky. It is the product of analysing going conditions, trainer trajectory, RPR versus market price, and recent form trajectory in combination — not in isolation.
Road to Wembley completed the hat-trick on May 15 at Aintree, returning +£48.56 at odds of 1.97 on a £50 stake. Three consecutive days. Three winning selections. Different tracks, different race types, same systematic process behind every one of them.
What the System Is Actually Doing
The mistake most punters make during big festival weeks is overcomplicating their approach. They take on races they do not need to be in, chase enhanced odds on horses whose form does not fully stack up, and spread stakes too thinly across too many selections. The result is noise — a string of small losers punctuated by occasional winners that do not cover the losses.
Horse Racing Oracle AI narrows the focus. More than 200 variables are processed for every race — official ratings versus Racing Post Ratings, course and distance win percentages, trainer strike rates over the preceding 14 days, jockey booking patterns, going suitability, class movement, and recent finishing trajectory. The output is not a long list of possibles. It is a small number of high-confidence selections with a clear staking rationale attached.
During Dante Festival week, that process pointed clearly at York and Aintree. The selections came with a reason for each one, and each reason held up when the race was run.
The NAP of NAPs — Our Guide at Stratford
The week closed on May 16 with Our Guide at Stratford. Priced at 1/3, this was always going to be a modest return on a £10 stake — £13.60 — but that is not the point of a selection like this. Our Guide was the NAP of NAPs for the week: the single highest-confidence selection across all five days, the kind of horse you include in an accumulator to anchor the odds, or back as a standalone statement of certainty. He won. The week ended four from four in terms of horses the system had identified as genuine NAP-level selections.
What This Week Proves
One good week does not define a tipping system. What matters is consistency over time, across different tracks, going conditions, race types, and price ranges. But a week like May 12–16 is a useful illustration of what the system looks like when it is working — because it shows that big-festival racing, which intimidates many punters into inaction or recklessness, is exactly where a data-driven approach finds its edge. The market gets noisy during Dante week. The form book does not.
If you want to receive tomorrow's NAP before the market prices it in, the sign-up is free. The selections are published every morning before racing begins.
Want free AI-powered tips every morning? Sign up free at horseracingoracleai.com →
Betting involves risk. Please gamble responsibly. Visit BeGambleAware.org.