Reading a Greyhound Racecard: Traps, Grades, Going and Form

A greyhound racecard packs a lot of information into one line per dog. If you've never opened one before, the codes and abbreviations look impenetrable. This guide walks through every column on a typical UK or Irish racecard, explains what each piece means, and points out which fields actually move the model.

The race header

Above the runners you'll see:

• Track + race number — e.g. "Romford R7" identifies the meeting and the slot within the card.
• Off time — the scheduled start, in local time.
• Distance — in metres. Sprint (<400m), standard (400–500m), middle (500–600m), stayer (600m+).
• Grade — the class of the field. A1 is highest at most UK tracks; D is lower-class; OR is an open race; HP is a handicap.
• Going — fast / standard / slow, telling you how the surface is running.

These four define the event — every other column describes a runner within it. See the grade glossary entry and going entry for the practical implications.

The runner row — trap and dog

The leftmost columns identify the dog:

• Trap number (1–6 in standard UK racing). Trap 1 is on the inside rail, trap 6 on the outside. Colours: red, blue, white, black, orange, black-and-white. See trap for trap-bias details.
• Dog name in all caps. Punctuation and apostrophes are stripped (the dog "O'Hara's Hero" appears as "OHARAS HERO" on most printed cards).
• Trainer — the licensed handler.

A trap colour clash isn't accidental — it's how trackside crowds tell runners apart at speed.

Form figures — the most-read column

The next-to-trap column is usually the last 5 finishing positions, most recent on the right:

32145

reads as: 3rd, 2nd, 1st, 4th, 5th — with the 5 being the most recent run. Special characters:

• 0 — finished 7th or worse (or unplaced).
• - — line break / season change.
• R — retired part way (rarely).
• N — non-runner from a recent meeting.

A pattern like 11132 says the dog has won 3 of its last 5 and just dropped form a little. 45634 is the opposite — bouncing around the back of the field.

What the model does with this: the average finish position over 3 and 5 runs, the win rate over 5, and the recency-weighted form are all features in the LightGBM predictor. Bias toward recent runs is automatic — older runs weigh less.

Distance, weight and SP-related columns

Different cards lay these out differently, but typically you'll see:

• Weight — the dog's race-day kilogram weight, taken at the kennel hours before the race. Sudden weight swings (e.g. >1kg below recent average) can be a sharp tell.
• Sectional time — split time to a fixed mark (usually the first bend). Smaller number = faster.
• Best time — the dog's personal best at this distance, perhaps with a track abbreviation.
• Forecast price — what Timeform (or the racing media) expects the SP to be. See forecast-price.

Sectional times are particularly valuable: they isolate early pace from late running. A dog with a strong sectional but bad finishing positions is a known frontrunner — predictable in races where it can lead, vulnerable in races where it can't.

Form remarks and trip notes

Below or beside the form figures, many cards print a short verbal note for each recent run: "Crowded 1st, ran on", "Led throughout", "Slow away, weaker". These are human read-outs of trip trouble — small but useful context.

TrapStats doesn't currently parse these notes (we have a remarks-NLP module flagged for future work), but a sharp punter will read them carefully — a dog with consistent "crowded" notes may be a luckless improver underneath the form figures.

Putting it together

When you read a UK racecard you're scanning, in order:

1. Top line — is this a race I want to bet at all? (Grade, distance, going.)
2. Form figures — who has been winning recently?
3. Trap — do the form-pickers fit the bias of this track?
4. Forecast price — who is the market siding with?

The model does this in milliseconds, but doing it by eye is the only way to develop a feel for the cards. Once you've read 200 of them, you'll see the patterns immediately.

Where TrapStats fits

On every track hub page (e.g. /en/tracks/romford) we render today's racecard with the AI win probability and EV alongside the official columns. The "Why this pick" panel on the /denis page breaks down the decision: which form features pushed each pick up, which gates it passed, what the segment-prior base rate would predict, and where the bet lands in the EV-vs-breakeven plot.

Use the racecard to develop your intuition; use the model as a second opinion. They're stronger together.