The Computer Straight Forecast Formula: How the CSF Sets Its Dividend

html

Where the CSF Number Actually Comes From

The first time I tried to explain a Computer Straight Forecast dividend to a friend, I drew the formula on the back of a Doncaster racecard and watched his face go slowly blank. He had assumed the number on the screen was decided by a bookmaker in a back room. It is not. The CSF is a settlement product calculated by an algorithm that nobody at the bookmaker actually controls on the day of the race.

That single misconception sits behind half the complaints I see about “stingy” CSF payouts on social media. Punters compare two races where the first two home looked similar, see wildly different dividends, and assume someone trimmed the price. In reality the entire dividend is mechanical – every shilling of it driven by the Starting Prices and a formula that has been public since the 1970s. By the end of this piece you will know exactly what feeds that number, why two similar-looking finishes can pay £200 apart, and where a CSF actually has the legs to outrun a Trifecta on the same race.

The CSF Formula in Plain English

Strip away the maths textbook tone and the CSF is doing one thing – turning the Starting Prices of the first two finishers into a number that estimates how unlikely that pairing was. The less likely the pairing, the bigger the payout, scaled to a £1 unit stake.

The core mechanic runs in three layers. First, the Starting Prices of the two horses are multiplied together with a small adjustment that accounts for the chance the second horse would have beaten the first. Second, the formula applies a “uniformity factor” that tightens the payout when the field is dominated by very short prices and loosens it when the prices are spread thinly. Third, a takeout-equivalent deduction shapes the final dividend so the implied margin sits roughly in line with the bookmaker industry’s expected return on forecast pools.

The reason this matters is that the CSF is never quoted before the race. The dividend lands the moment the SP is declared and the formula runs. That gives the CSF a property fixed-odds bets cannot have – the price reflects how the live market actually closed, not how it looked an hour earlier. A horse that drifted from 5/1 to 9/1 in the last five minutes is rewarded in the CSF in a way the morning-priced Tricast simply cannot match.

I treat the formula as a black box that rewards two things – long combined prices and unusual price configurations. If the second favourite beats the favourite in a competitive handicap, the CSF will usually feel fair. If a 33/1 shot beats a 50/1 shot in a small field, the CSF will pay a number that looks almost theatrical, and that is the algorithm doing its job, not a generous gesture.

How Starting Prices Feed the Calculation

Forget the morning shows and the early prices. The CSF has one input that matters – the Starting Price returned for each runner, as agreed by the SP supervisors on the live market that closes at the off. Every other quote a punter sees during the day is meaningless to the settlement.

The Starting Price is not pulled from a single screen. It is sampled from a panel of on-course and digital sources, then audited by an independent SP team that finalises the figure for settlement. That means a horse traded heavily online but quietly on-course can land at a different SP from the one a punter saw on a single bookmaker app. The CSF then uses that final, settled SP – not the price the punter was watching.

What this does to the dividend is best shown by example. Imagine two handicaps with near-identical morning markets. In race one, the favourite holds firm at 3/1 and the second favourite at 9/2. In race two, the favourite drifts to 5/1 and the second favourite tightens to 4/1 – same approximate combined book, very different price shape. A CSF on the 3/1 and 9/2 horses pays roughly what you would expect from a moderate combined chance. A CSF on the 5/1 and 4/1 pair in race two pays a noticeably different number, because the formula treats a flat top of the market differently from a steep one. The same finish, the same combined implied probability – the CSF rewards them unevenly because the market shape itself is data.

This is why two CSFs on superficially similar prices can pay dividends a clear distance apart. The Starting Price configuration of the entire market – not just the two finishers – informs the uniformity adjustment.

A Worked CSF Example from a UK Handicap

Picture a 12-runner Class 4 handicap at Newbury. The first two home, declared at SP, are a horse that started at 8/1 and a horse that started at 14/1. The market behind them was broken – a 5/2 favourite, a 100/30 second, then a long tail of doubles and trebles. A £1 CSF on the 8/1 to 14/1 finish lands at roughly £104 on a market shape like this. Multiply through to the £61 region if the favourite had figured in second instead of being unplaced.

Now hold that next to a real CSF that captured the imagination of UK forecast players – the 2019 Cheltenham handicap that produced first three home at 50/1, 66/1 and 40/1. The Computer Tricast on that race paid £73,711.25 for a £1 stake. The Trifecta on the same finish paid £23,748.90. The CSF was settled on just the first two of those longshots, and the dividend reflected what the algorithm regarded as an almost impossible combined price configuration in a competitive championship handicap.

What both examples share is the way the CSF rewards unusual market shape. The Newbury fiction pays because the favourite was short and the placed horses long. The Cheltenham reality pays because the top of the market collapsed and the formula treated the survival of three outsiders as a freak outcome – which, on the prices, it was.

How the CSF Differs from the Computer Tricast

The CSF and the Computer Tricast are siblings, not twins. Both are settlement-only products. Both feed off Starting Prices. Both are calculated by an algorithm rather than priced live. But the Tricast extends the bet to a third finisher and applies its own uniformity logic across a three-horse permutation, which makes the dividend grow non-linearly with the prices.

That extension is exactly why the Cheltenham 2019 number stretched to £73,711 on the Tricast but only £23,748 on the same race’s Trifecta. The Tricast formula effectively compounded the improbability of all three placed runners, while the Tote Trifecta pool – driven by what punters actually staked, not by a price formula – settled to a lower number because enough Tote players had at least one of those three on their slip.

For the punter, the practical line between the two products comes down to one question – do you want a settlement product that pays you whatever the algorithm decides, or do you want a pool product whose dividend depends on how many other people backed your combination? On a chaotic-priced race with a quiet Tote pool, the Tricast wins. On a competitive race with deep Tote liquidity, the Trifecta tends to. I keep an eye on both, and on the days the CSF is in play I will often take a small CSF as a hedge against a two-horse line I already have inside a wider Trifecta box. If you want to dig deeper into the Tote-side mechanics that sit on the other side of this trade-off, the breakdown of the pool maths is in our UK Trifecta payout calculator guide.

The single biggest myth I want to retire is the assumption that the CSF and Trifecta are interchangeable. They are not the same bet. They reward different inputs, settle through different machinery, and on a long enough run they will not converge. Knowing which one to back on which race is where the edge sits – and that decision starts with reading the Starting Price configuration, not the morning show.