Calculating Correlated Selection Margins in Bet Builders
A friend texted me a screenshot last August during the Vitality Blast. Six legs on a single Lancashire match, total odds 38/1, stake £5, potential return £195. He wanted to know if I liked it. I did not. Every leg looked defensible in isolation. Stack them together, and the implied probability of the whole thing collectively was around 2.5 percent, with most of the apparent value coming from the book applying its margin six times over rather than from any compounding edge in his reading of the game.
That’s the bet builder problem in three sentences. The UI is friendly, the legs feel intuitive, the final price looks attractive, and the structural maths is brutal. With the UK online betting sector pulling in £2.6 billion in gross yield in FY 2024/25 — a rise of 10.9 percent in a single year — the book has every incentive to keep this product front and centre. It’s high-margin, it’s sticky, and it converts casual punters far better than a single straight bet ever would.
How a Cricket Bet Builder Is Priced
The book starts with the underlying selection prices, then applies a correlation adjustment, then layers on its margin per leg. Take three selections from a T20 match: side A to win at 1.80, top batsman from side A at 5.50, and total runs over 320 at 1.90. If you multiplied those decimals naïvely you’d get a combined price of about 18.81, which would imply a probability of around 5.3 percent.
That’s not the price you’ll see in the slip. The book applies what amounts to a discount based on positive correlation — if side A wins, top batsman from side A and high total are both more likely, so the legs aren’t independent. The slip price might come back at 14.50, implying around 6.9 percent. The difference between the naïve multiply and the actual price is the correlation adjustment, and books don’t share their formula.
The margin is the more important part. Each leg in a bet builder typically carries a per-leg overround of 4 to 6 percent for sports like cricket. Stack six legs and you’re paying the book a structural toll well into the high teens on the final price, even before correlation adjustments take a further bite. That’s the hidden cost: the more legs you add, the further the “true” price drifts from the slip price, and almost all the drift is in the book’s favour.
Correlated Selections: When the Book Lets You, When It Does Not
The book’s correlation rules are the most interesting thing in a bet builder, because they tell you exactly where the operator believes its margin is most exposed. Strong positive correlations are usually blocked outright. You typically cannot combine “team A wins” with “team A first-innings total over 180” on the same slip, because the second event being true makes the first event drastically more probable. The book would be paying out compound probability without compound margin to defend.
Weaker correlations get through. “Top batsman from team A” and “team A wins” both correlate positively, but the relationship is loose enough that books allow the combination and just apply a discount. “Top bowler from team B” and “team A wins” are negatively correlated — wickets for B make A’s victory harder — and the slip price reflects that with what looks like a modest premium versus the naïve multiply.
Where I see most punters lose value is on combinations the book lets through with no obvious flag. Pairing “over 320 total runs” with “top batsman scores 50+” in a T20 is a textbook correlated stack — both legs benefit from the same underlying high-scoring conditions. The book will let you have it because the correlation isn’t strong enough to be a structural risk, but the slip price has been quietly dampened to account for the connection. The punter sees a single-bet combined price and assumes independence. The book knows better.
Practical Bet Builder Recipes for T20, The Hundred, ODI
I’ll be direct about what I’ll actually back as a multi-leg cricket bet. The structure that survives the margin maths is two to three legs, lightly correlated, with each leg defensible on its own merits.
A T20 build that has earned its place on my slip: side batting first to score over 170, plus their top batsman to be one of two specific openers. Two legs, mild positive correlation, around 4/1 combined on a typical match. The reasoning has to come from the conditions — flat surface, in-form openers, opposing seam attack ill-suited to early death-over death — not from price gravity.
For The Hundred, a clean two-leg recipe is “set 1 to 5 total runs under 35” plus “match favourite to win”. You’re betting that the pacy bowling unit will get through the powerplay with damage, and the underlying favourite goes on to convert. The legs are independent enough to dodge serious correlation discounting and the conditions read is shared.
In ODIs the bet builder market is thinner because totals settle more linearly across 50 overs and the variance per leg is lower. The recipe that still pays is “match winner” plus “top wicket-taker from the winning side”. A single bowler bias on a known surface can be worth pairing with the match result. Three legs maximum in ODIs — anything more and the per-leg margin compounds past the point where any pre-match read can rescue the slip.
