Glossary
Racing & betting terms, defined
The vocabulary of racing and betting, in plain English. Bookmark it.
- SP (Starting Price)
- The official price of a horse at the moment the race starts, set from the on-course market. Used to settle bets and as the benchmark for closing-line value.
- BSP (Betfair Starting Price)
- The starting price calculated by the Betfair exchange by matching all SP bets at the off. Often differs from the official SP and is widely used to measure value.
- CLV (Closing Line Value)
- Whether the price you took beat the market's final (closing) price. The single hardest metric to fake and the best long-run signal that a process has an edge.
- Each-way
- Two bets in one: half your stake on the win, half on the horse finishing in the places. Place terms (number of places and the odds fraction) depend on the race.
- Place terms
- How an each-way place pays — a fraction of the odds (e.g. 1/4 or 1/5) across a set number of places that scales with field size and race type.
- NAP
- A tipster's single most confident selection of the day. 'Next Best' is the second tier; Algohorse also publishes Speculative selections.
- Favourite–longshot bias
- The persistent market pattern where longshots are overbet and favourites underbet relative to their true chances.
- Steamer
- A horse whose price is shortening (being backed in) in the market before the off — the opposite of a drifter.
- Drifter
- A horse whose price is lengthening (drifting out) in the market before the off, suggesting weakening support.
- Handicap
- A race in which horses carry different weights based on official ratings, intended to give every runner a theoretically equal chance.
- Going
- The condition of the racing surface, from firm to heavy, which materially affects which horses are suited.
- Draw bias
- An advantage or disadvantage conferred by a horse's stall position, depending on course, distance and conditions.
- ROI (Return on Investment)
- Profit or loss as a percentage of total stakes. Reported by Algohorse with confidence intervals once the sample is meaningful.
- Edge
- A genuine, repeatable advantage over the market — evidenced over time by positive closing-line value, not by a run of winners.
- Calibration
- Whether stated probabilities match reality: if a model says 70%, do those selections place 70% of the time? Good calibration means honest confidence.