Set Betting Strategy: Beyond Match Winner
Exact set-score forecasting must stop when a player clinches the match. A match-win probability is not a per-set probability. The calculations below use explicit hypothetical per-set inputs and enumerate only valid match paths.
Best-of-three: enumerate the terminal paths
Assume independent sets with a constant probability p = 0.70 that A wins each set, and no retirement or other interruption. A 2–0 requires WW, with probability p² = 49%. A 2–1 has exactly two valid paths: WLW and LWW. Each has probability p²(1−p) = 14.7%, so together they give 2p²(1−p) = 29.4%. WWL is invalid: the match already ended after WW.
| Final score | Valid paths (A’s perspective) | Probability |
|---|---|---|
| A 2–0 | WW | 49.0% |
| A 2–1 | WLW, LWW | 29.4% |
| B 2–0 | LL | 9.0% |
| B 2–1 | LWL, WLL | 12.6% |
The four probabilities sum to 100%. A’s match-win probability is 49.0% + 29.4% = 78.4%, different from the 70% per-set assumption. There is no undefined correction factor.
Best-of-five and changing set probabilities
Under the same constant independent-set model, A 3–k has probability C(2+k,k) × p³ × (1−p)^k for k = 0, 1, or 2. The final set must be a win for A, and the earlier sets contain exactly two A wins and k losses. At p = 0.70, the A probabilities are 34.300%, 30.870%, and 18.522%; B’s are 2.700%, 5.670%, and 7.938%, totaling 100% across both players.
If set probabilities change with score, fatigue, or serving order, multiply the relevant conditional probabilities along each valid path and sum the paths. Do not apply an invented momentum correction or an unsourced fifth-set win rate. Retirements and match tie-break formats require additional rules and states.
Connect the modeled score to a price
As a separate hypothetical, suppose p = 0.72 per set and A 2–0 is offered at +120. Under independence its probability is 0.72² = 51.84%; break-even is 100/220 = 45.4545%; net EV is 0.5184 × 2.20 − 1 = +14.048%. That expectation is conditional on the model, not proof of a forecasting edge.
Inspect how plausible changes in p, dependence, format, or retirement settlement affect the result. Complexity and a lower number of competing quotes do not establish mispricing. Calculate margins from actual complete-market prices instead of asserting a universal set-market range.
OwnTheLines does not expose an exact-set-score placement workflow. Use the example as an external modeling exercise; do not confuse a match-winner selection with a set-score contract.
Continue reading: Surface Tension · Building Your Own Forecasting Model.
Frequently Asked Questions
What is A’s 2–1 probability at an assumed 70% per set?
With constant independent sets, the valid paths are WLW and LWW. Each is 0.70 × 0.30 × 0.70 = 14.7%; their sum is 29.4%. WWL is invalid because A has already clinched.
Can I use a 70% match-win estimate as the per-set input?
No. They describe different events. In this example a 70% per-set probability produces a 78.4% match-win probability.
Are complex set markets inherently mispriced?
No. Evaluate the actual prices, format, settlement rules, and forecast uncertainty. Complexity alone is not evidence of value.