CSF Dividend Calculation: From SP to Settled Exotic Slip
Available in US
The Starting Price Algorithm Behind Forecast Settlement
Ask ten regular UK punters how the CSF dividend gets calculated and you’ll get ten answers – most of them wrong. «It’s like the Tote.» «It’s just SP multiplied by SP.» «The Press Association just makes it up.» None of these is accurate, and the confusion costs people money because they back the wrong forecast product without realising the algorithm is doing something specific that neither the Tote nor any rule-of-thumb captures.
The Computer Straight Forecast – CSF – is the bookmakers’ fixed-odds answer to the Tote Exacta. It’s a deterministic formula run by the Sporting Life and Press Association on every UK race, producing a single dividend per pound staked based on the starting prices of the first two home and the field structure. There’s no pool. No takeout in the parimutuel sense. The dividend the formula returns is what every operator pays out. Understanding how the algorithm builds that number from SPs is the difference between a forecast punter who knows their product and a forecast punter who’s gambling on a black box.
Where the CSF starts – the starting price forming process
The CSF takes Industry Starting Prices as its inputs. Those SPs are themselves the product of a market – a real one, formed at the racecourse and on the digital exchanges that feed back into the on-course market. The Racing Post and the major price reporting services aggregate the prices offered by on-course bookmakers in the ring at the moment the starting stalls open (or the tape goes up for jumps), apply the long-standing SP rules, and publish a single SP for each horse.
The integrity of that SP-forming process matters because the CSF is only as reliable as the SPs it ingests. On a quiet Wednesday at a small course, with limited on-course bookmaker money in the ring, the SPs can drift in unrepresentative ways. As one Dragonbet bookmaker put it after the 2026 Cheltenham Festival, «It’s funny how the betting ring ebbs and flows. For a long time all the sharp money used to be in the betting ring. It then disappeared but, with all the regulation online, the sharp money has really come back to the course and the bookies are forced to have an opinion.» That return of sharp money to the on-course market has tightened SP formation again – and through it, the CSF dividends that flow from those SPs.
BHA punctuality data backs the same picture. The proportion of races starting within two minutes of the scheduled off-time climbed to 87.6% in Q1 2025 from 72.7% in 2023. That tighter on-time discipline means SPs are formed with less late drift and more reliable convergence, which feeds through to CSF dividends that match the apparent market more closely than they did even five years ago.
The SP-derivation formula – what the algorithm actually does
The CSF formula in its standard published form is not a secret. It builds a forecast dividend from the SPs of the winner and second, weighted for whether the favourite is involved, the strength of the second favourite, and the number of runners in the field. The published version produces a dividend per unit staked that ranges from low single digits on short-priced forecasts to thousands of pounds on longshot pairings.
At a structural level, the algorithm rewards rarity. A 10/1 horse winning ahead of a 16/1 horse produces a dividend much larger than 10×16 – because the formula weights the unlikeliness of the combination, not just the multiplied prices. Conversely, two short-priced horses finishing in their forecast order produce a dividend that’s often slightly lower than naive multiplication would suggest, because the algorithm captures how often that result occurs and prices it accordingly.
The deduction for two-horse forecasts where both selections are favoured – what bookmakers call the «dual forecast deduction» – applies in the calculation when both top-two finishers are themselves co-favourites. The effect is a modest reduction in the published CSF, reflecting the algorithm’s view that the result was the most likely outcome.
A worked example with two short prices and one 12/1
Take a 10-runner Saturday handicap on the Flat. The result: a 7/2 favourite, a 12/1 outsider, and the third home a 6/1 horse who isn’t part of the forecast calculation since the CSF only covers the first two. The forecast pair is 7/2 and 12/1.
Naive multiplication would give 3.5 × 12 = 42, so a £1 forecast at SP would seem to pay around £42. The CSF on this kind of pairing typically returns a dividend in the £55 to £70 range on a £1 stake, depending on the field size and the strength of the second favourite. The algorithm’s uplift over naive multiplication reflects the rarity of a 12/1 horse finishing second after the favourite has won – a result that’s less common than the multiplied price implies.
The same result through a Tote Exacta would depend on the pool. On a Saturday handicap pool of £85,000 with takeout of around 27% and twelve winning units, the Tote Exacta would pay £85,000 × 0.73 ÷ 12 = roughly £5,170 per unit. But across a wider sample, the Tote+ Exacta beats the CSF in 77% of cases with an average +21% value per bet, so most of the time the pool is the structurally correct product. The CSF wins when the pool is thin or the result is «expected enough» that the pool dilutes its dividend below the algorithm’s number.
Where the formula breaks down – shock SPs and thin markets
The CSF assumes the SPs it ingests are accurate reflections of the market. When that assumption fails, the dividend can look strange. The classic case is a horse whose SP doesn’t match the market shape – a horse heavily backed online but not in the ring, finishing in the forecast, with the on-course SP catching up to the real market only after the off.
The second failure mode is thin markets. On a 7-runner conditions race at a small evening Flat fixture, with minimal on-course bookmaker presence, the SPs are formed from a sparse market. The CSF still runs the formula but the inputs are noisier than usual, and the dividend can be 20 to 40% higher or lower than the equivalent Tote Exacta on the same pair.
The third edge case is a withdrawn horse handled by Rule 4. When a horse withdraws between off-time and the race actually running, Rule 4 deductions cascade through the SPs of remaining horses. The CSF picks up the post-deduction SPs and produces a dividend based on those, which can look surprisingly low to a punter who’d been mentally pricing the bet at the pre-Rule-4 prices in the morning paper.
For the broader product comparison this article fits into, the head-to-head guide on CSF and Tote Exacta dividends covers when each product systematically outperforms in light of the algorithm’s structural quirks.
Does the CSF use morning prices or SP for settlement?
Industry Starting Prices, not morning prices. The CSF formula ingests the SPs as published after the race goes off and runs the algorithm on those. Morning prices and earlier on-course prices have no role in the CSF settlement – the dividend depends entirely on the final SPs returned for the first two horses home.
Can two horses at the same SP produce different CSF dividends in different races?
Yes. The formula weights the field size, the strength of the wider market, and the relationship between the two prices in the context of the whole race. A 7/2 and 12/1 forecast on a 10-runner handicap produces a different dividend from the same prices in a 6-runner conditions race because the algorithm’s field-size factor is different.
Escrito por los editores de «box bet in Horse Racing».
