I Love Horse Racing – So I Taught Claude to Build Better Tickets. Here’s What Actually Happened.

⚠️ Before anything else — a serious word. Horse race betting can be
addictive and you can lose real money fast. What follows is a research and
analysis tool, not a money machine, and no prediction system can guarantee a profit.
Only stake what you can comfortably afford to lose, never chase losses, and treat any winnings as a
bonus, not an income. If betting stops being fun, stop — and if it starts to feel out of control,
reach out to a service such as GamCare or BeGambleAware. 18+ only. Bet responsibly.

I like horse racing. The form, the going, the silks, the maths of a good each-way ticket — I enjoy
all of it. And a few months ago I had the thought a lot of us have had: could I use AI to build
better tickets? Not to “beat the bookies forever” — I'm not naïve — but to stop betting on gut
feeling and start betting with a method.

So I tried. And my first attempts were honestly mediocre. I'd paste a racecard into a chatbot, get
back a confident-sounding paragraph, and realise it was just rephrasing the favourites. No structure,
no memory, no sense of value. It told me what might win; it never told me whether the price
was worth taking.

So I kept going — and slowly turned those experiments into a proper system: a structured knowledge
base the AI reads before it scores a single horse. It didn't make me rich. What it
did do was make me disciplined — the same factors, every race, no tilt, no favourite
bias. That system became the Professional Horse Race Betting AI Skill, and this is
how it works.

👉 Get the skill here:
https://ko-fi.com/s/d3452efe52

What actually makes this skill special

This isn't a vague “AI tipster.” It's a complete prediction framework. Here's the part that
should make any serious bettor lean in:

• A 10-factor scoring engine (F1–F10). Every runner is graded on form &
  recency, distance fit, going/surface, class & weight, raw time/speed, jockey or driver quality,
  draw & race shape, experience, market confirmation, and — for the big staying chases — landmark
  race experience. The weights change by discipline: speed dominates in trotting, stamina
  and going dominate in a 6,000m steeplechase.
• Clear S / A / B / C tiers and a full rank order — not one “banker,” but a
  defensible top-5 you can actually build tickets around.
• A Value Gap calculation. The model compares its own rank to the market rank, so
  you instantly see which favourites are false positives and which outsiders the crowd has
  underpriced. That's the difference between backing winners and backing value.
• It learns. After each result, a post-race routine maps what the model got right
  and wrong back to the exact factor, then writes a new rule. Real lessons are already baked in — like
  hard-capping a horse drawn in the widest stall of a 16+ runner handicap, or the “elite-driver
  rescue” pattern where a top driver lifts a poorly-formed trotter straight into the placings.
• Deep, reusable horse and rider profiles. The French jump and trotting circuits
  reuse the same pool of horses, jockeys, drivers and trainers season after season. The skill carries
  persistent profiles — a finite pool of knowledge that compounds into a genuine edge.

It covers what serious bettors actually play: UK, French and international racing,
across flat, jumps (hurdles and steeplechase) and harness/trotting — all in plain
English.

Why an AI out-thinks a human bettor (on ideas, not luck)

AI doesn't have a crystal ball, and neither does anyone else. But on the quality and consistency
of ideas, a well-instructed AI has advantages I simply don't have at 9pm after a long day:

1. Zero emotion, zero tilt. It never chases a loss, never doubles up out of
   frustration, and never backs a horse because the name is cool or it “feels due.” It scores the 18th
   race of the day as carefully as the 1st.
2. Total recall. It remembers every horse, jockey and trainer profile across seasons —
   which trainer's “target race” quotes are reliable, which horse falls when the ground turns sticky,
   which driver rescues badly-drawn trotters. Humans forget; the knowledge base doesn't.
3. Speed and cross-referencing. In seconds it can build a mini-ranking from every
   shared reference race in the field, adjust for weight and going, and compare beaten lengths — the
   grind most punters skip.
4. Discipline about value. It rigidly separates “will it run well?” from
   “is it worth the odds?” via the Value Gap, instead of blindly backing the favourite.
5. It thinks in tickets, not single picks. The skill reasons about coverage risk —
   which horse could wreck your Quinté/Kincsem+ ticket even if it won't win — and builds spread lines
   accordingly.
6. It improves on purpose. Human “experience” is fuzzy gut feeling. The skill turns
   each result into a written, repeatable rule. The system you use next month is measurably sharper
   than the one you use today.

What you need to run it

The skill is a set of plain-text knowledge files, so it works with any AI that can read documents.
For the best experience, use a tool that can load a whole folder at once:

• Claude (Desktop app or Claude Code) — ideal. Point it at the skill folder and it
  reads the entire knowledge base before answering.
• OpenAI Codex / ChatGPT — works well; attach the files or paste the master
  framework into a project.
• Google Antigravity or any other agentic coding assistant that can read a project
  directory.
• Any file-aware chat AI — if it lets you upload or paste several documents, you're
  good to go.

Rule of thumb: the more of the knowledge base the AI holds in context at once, the sharper the
prediction. Agentic tools that read the full folder give the strongest results.

First steps: load the skill into context

1. Download the skill from the Ko-fi
   link and unzip it. You'll get the framework plus the discipline-specific knowledge files.
2. Keep all the files together in one folder. They're designed to work as a set — the
   master framework references the galopp, trotting and horses-and-riders files.
3. Bring them into your AI's context. In an agentic tool, open the folder as your
   working directory. In a chat app, upload or paste the files — at minimum the master framework, plus
   the file matching the race type you're predicting.
4. Give it the standing instruction: “Read the knowledge base before scoring.
   Score every runner with the 10 factors and give me tiers, a rank order, Value Gaps and a
   ticket.”

How to make your first prediction

You don't need to be technical. There are two easy ways to feed the AI a race:

• Option A — give it a link. Paste the URL of the racecard (for example a PMU, Geny
  or official racecourse page) and say: “Predict this race using the skill.” The AI pulls the
  runners, odds and conditions and goes to work.
• Option B — copy-paste the basics. If web access isn't available, just paste the
  race title and date (e.g. “Grand Handicap de Lamorlaye, Chantilly, 14 June 2026”)
  and, ideally, the runner list with draws, jockeys and odds. The model scores from what you give it.

Then ask for the full verdict. In moments you'll get back:

1. A tier list (S / A / B / C) for the field.
2. A ranked top-5 projection with the reasoning behind each placing.
3. Value Gaps showing where the model disagrees with the market.
4. A Kincsem+ / Quinté ticket plan — a core line and a spoiler/coverage line — plus
   the horses most likely to upset your ticket.

The honest bottom line

I'll say it plainly, because I wish more people did: this skill won't make betting risk-free, and it
hasn't made me rich. What it changed is how I bet — consistent, transparent, value-led, and a
little sharper after every result instead of a little more emotional. If you're the kind of bettor who
wants to back value with a clear head rather than chase favourites on a feeling, this is the analyst I
built for exactly that.

🏇 Get the Professional Horse Race Betting AI Skill:
https://ko-fi.com/s/d3452efe52

And one last time: bet for entertainment, set a budget before you start, and walk away when it
stops being fun. The real edge is discipline — on and off the track.