If you bet on the exchange, Rule 4 does not apply to you. That sentence alone separates the exchange world from the bookmaker world, and it is the starting point for understanding how Betfair handles non runners. Instead of the fixed deduction scale that traditional bookmakers use — a table of pence-per-pound tied to the withdrawn horse’s starting price — Betfair applies a reduction factor calculated from the horse’s actual traded odds on the exchange.
The reduction factor is a more precise instrument. It reflects what the market genuinely believed about the horse’s chances, not a standardised grid that approximates them. But precision does not mean simplicity. The RF calculation has its own quirks, thresholds, and edge cases that catch exchange bettors off guard — especially when multiple horses are withdrawn from the same race or when a non runner is declared late in the market cycle.
This is a deep dive into how the reduction factor works on the exchange, where it diverges from Rule 4, and what it means for your net payout when a horse is pulled out.
How the Reduction Factor Is Calculated — Traded Odds to Percentage
The reduction factor starts with a single question: what was the withdrawn horse’s implied probability based on its traded price? On the exchange, every horse has a price determined by the market — the aggregate of what backers and layers are willing to accept. That price translates directly into an implied probability.
A horse trading at 2.0 (evens) has an implied probability of 50%. A horse at 5.0 (4/1) implies 20%. A horse at 11.0 (10/1) implies roughly 9.1%. The reduction factor is derived from this figure. Betfair takes the withdrawn horse’s weighted average traded price — not just the last matched price, but a volume-weighted average across all matched bets — and converts it into a percentage. That percentage, after some adjustment, becomes the RF applied to all winning bets in the race.
The adjustment accounts for overround. Because the sum of implied probabilities in a horse racing market typically exceeds 100% (the exchange overround, though much smaller than bookmaker margins), Betfair normalises the figure. The exact formula is proprietary, but the principle is transparent: the RF represents the withdrawn horse’s share of the market, and it is subtracted proportionally from every winning payout.
In practical terms, if a horse is withdrawn with an RF of 30%, a winning bet that would have paid £100 in profit now pays £70. The calculation is applied to the profit portion of the bet — your stake is returned in full, and the RF reduces only the winnings. That distinction from bookmaker deductions becomes important when comparing the two systems directly.
The RF is published by Betfair immediately after the non runner is confirmed. It appears in the market alongside the NR label, and it is fixed at the point of withdrawal. If the remaining horses’ prices shift after the NR — as they always do — the RF does not change. It is locked to the moment the horse was removed.
The 2.5% Threshold — When Betfair Ignores the Non Runner Entirely
Not every non runner triggers a reduction factor. Under Betfair’s exchange rules, if the calculated RF for a withdrawn horse is less than 2.5%, the reduction is not applied. The horse is treated as immaterial to the market — its removal does not meaningfully change the probability distribution among the remaining runners.
In practice, this threshold kicks in for extreme outsiders. A horse trading at 40.0 (39/1) or longer has an implied probability under 2.5%, and its withdrawal will not reduce your payout at all. This is a significant difference from Rule 4, where even a 14/1 shot triggers a 5p deduction in the bookmaker world. On the exchange, that same horse — if it was trading at 15.0 or higher — might clear the 2.5% threshold and leave your winnings untouched.
The threshold exists for a practical reason: applying tiny reductions to every non runner in a large field would create noise without meaningful impact. In a 20-runner handicap at a Saturday meeting, one or two outsiders might be withdrawn. Without the 2.5% floor, Betfair would need to process micro-adjustments across potentially thousands of matched bets — a computational cost that adds no real fairness to the settlement.
For the exchange bettor, the threshold is a quiet advantage. In races with many runners, the NR of a long-priced horse simply does not affect you. On the bookmaker side, the same withdrawal would still trigger a Rule 4 deduction, however small. It is one of several structural reasons why exchange settlement tends to leave more money in the bettor’s pocket when non runners are involved.
Reduction Factor vs Rule 4 — A Direct Comparison
Rule 4 and the reduction factor solve the same problem — compensating for a shortened field — but they use fundamentally different methods. The differences matter more than most bettors realise.
Rule 4 is a fixed scale. It was codified under Tattersalls Rule 4(c) and assigns deductions based on the withdrawn horse’s starting price at the time of removal. The scale runs from a maximum of 90p in the pound for a horse at odds of 1/9 or shorter, down to 5p in the pound for horses between 10/1 and 14/1. At odds longer than 14/1, no deduction is applied. The scale is blunt by design — it groups horses into broad price brackets and applies a uniform deduction regardless of what the market actually thought about that specific runner.
The reduction factor, by contrast, is continuous. It is calculated to multiple decimal places from the horse’s exact traded price on the exchange. There are no brackets, no rounding to the nearest 5p. A horse with a 12.3% implied probability generates a 12.3% RF — not a deduction chosen from a menu.
The second difference is what gets reduced. Rule 4 deductions apply to total returns — your stake plus winnings. The reduction factor applies to profit only. This distinction means that for identical races and identical non runners, the exchange bettor typically keeps more money than the bookmaker bettor. The gap is most visible when the withdrawn horse was a mid-priced contender: large enough to trigger a meaningful deduction, but not so dominant that the two methods converge.
A third difference is the treatment of outsiders. Rule 4 applies no deduction beyond 14/1. The exchange applies no reduction below 2.5% implied probability (roughly 40/1). In the range between 14/1 and 40/1, Rule 4 is silent while the exchange may still apply a small RF. In practice, this narrow band rarely matters, but it is a technical distinction worth knowing if you operate on both platforms.
Worked Example — Your Payout Before and After Reduction Factor
Numbers make this concrete. Suppose you back Horse A at 6.0 (5/1) on the exchange for a £20 stake. If Horse A wins without any non runners, your profit is £100 (5 × £20), plus your £20 stake back, for a total return of £120.
Now Horse B, which was trading at 4.0 (3/1), is declared a non runner. Its implied probability is 25%, and Betfair applies a reduction factor of 25.0% to the race. Your winning bet on Horse A is settled as follows: your £20 stake returns in full. Your £100 profit is reduced by 25%, leaving you with £75 in profit. Total return: £95.
Compare that to the bookmaker treatment. At Rule 4 deductions, a horse at around 3/1 triggers a 25p-in-the-pound deduction. But Rule 4 applies to total returns, not just profit. Your bookmaker return of £120 is reduced by 25p per pound: £120 × 0.75 = £90. You lose £30 with the bookmaker versus £25 on the exchange. The exchange left you £5 better off on a £20 bet — a difference that scales with stake size.
The gap narrows when the withdrawn horse is a strong favourite. If a horse at 1.5 (1/2) is declared NR, the RF will be in the region of 65-70%, and the Rule 4 deduction will be 75p or higher. At these extremes, both systems take a large chunk of the payout, and the practical difference shrinks. But in the most common scenario — a non runner at single-figure prices — the exchange consistently returns more to the winning bettor.
One caveat: exchange bettors also pay commission on net winnings, typically 2% to 5% depending on the user’s discount rate. After the RF is applied and commission is deducted, the net advantage over Rule 4 is smaller — but in most scenarios it still exists.
Settlement Quirks — Multiple NR, Late Withdrawal and Void Markets
Single non runners are the clean case. Multiple non runners in the same race add complexity. When two or more horses are withdrawn, Betfair applies reduction factors sequentially — each in the order the horse was removed. The first RF is calculated on the original market. The second RF is calculated on the adjusted market after the first non runner has been removed. This sequential approach means the total reduction is not simply the sum of the individual RFs; it is compounded.
Late withdrawals add another wrinkle. If a horse is removed close to the off, the traded price may reflect earlier market conditions rather than the final state of play. Betfair uses the volume-weighted average, which smooths out last-second spikes, but it can still produce an RF that feels disconnected from the horse’s true chances at the time of withdrawal. Exchange regulars accept this as a feature of a dynamic market, but it can surprise casual users who expected the RF to match the last visible price.
In extreme cases — typically when multiple favourites are withdrawn from the same race — Betfair reserves the right to void the entire market. This is rare and usually involves extraordinary circumstances: a mass withdrawal due to unsafe ground, a false start that leads to multiple at-start NRs, or an administrative error in the declarations process. When a market is voided, all matched bets are cancelled and stakes are returned to both backers and layers. No one wins, no one loses, and the race effectively never happened for betting purposes.
For the exchange bettor navigating non runners, the core takeaway is that the reduction factor is a better-calibrated tool than Rule 4, but it is not immune to edge cases. Understanding the sequential application, the 2.5% threshold, and the possibility of market voiding gives you a clearer picture of what your payout will actually be when the racecard starts thinning out.