This article explains what a three-way or 1X2 market is and why clear settlement conventions matter for reliable evaluations. It walks through converting odds to implied probability, removing bookmaker margin, calculating expected value, and applying realistic staking rules like fractional Kelly. The guide is aimed at sports fans and analytics-minded participants who want disciplined, repeatable procedures for checking edges and sizing stakes while respecting model calibration and settlement rules.

Three-way markets list home win, draw, and away win and usually settle on regulation time only.

Convert odds to implied probabilities, remove overround, then compute EV with calibrated forecasts.

Use fractional Kelly or capped stakes to limit drawdown while exploiting positive EVs.

What a three-way market is and why it matters for betting strategies

Definition and quick example

A three-way market, often called 1X2, lists three mutually exclusive outcomes for a match: home win, draw, or away win. These outcomes are typically priced separately, so the market gives a distinct price for each result rather than a binary win-or-lose line; for many readers this structure is the fundamental difference that shapes how you estimate probabilities and spot value. (see Sports Illustrated's explanation of three-way odds).

Because three-way prices normally settle on regulation time only, modelers must be explicit about the settlement horizon used in their forecasts. For soccer this usually means 90 minutes plus any stoppage time, not extra time or penalty shootouts, which affects how you map model outputs to market prices IFAB Law 7: The Duration of the Match.

Explore challenge-based testing and disciplined forecasting

Keep reading to see step-by-step procedures for converting odds, removing margin, and sizing stakes responsibly without assuming short-term certainty.

View FundedPlays Challenges

When three-way differs from two-way lines

Three-way markets require a different modelling approach compared with two-way or totals markets because they explicitly allocate probability mass to draws as a third outcome. That third outcome can materially shift implied probabilities and the perceived value of home or away selections, so you cannot directly treat a 1X2 price as if it were a binary moneyline without appropriate conversion and normalization.

For practical work this means your forecast must return three probabilities that sum to one under the settlement rules you intend to use. If your model instead outputs a single margin or an implied distribution for full-time including extra time, you must adjust it before comparing to regulation-only 1X2 prices.

How settlement rules change model inputs for betting strategies

Soccer: 90-minute law and its implications

Soccer markets labeled as 90 minutes or without specification are widely settled on regulation time, which the Laws of the Game define as 90 minutes plus any stoppage time; extra time and penalties are separate events and are excluded unless the market explicitly includes them IFAB Law 7: The Duration of the Match.

In practice that means when you build a predictive model for a 1X2 soccer market you must ensure the model's horizon and event definitions match the market settlement, or you will misestimate edge when converting probabilities to expected value.

Hockey: regulation, overtime, and separate market types

NHL rules distinguish regulation results from outcomes that include overtime and shootouts, and many operators offer both regulation-only and full-game (OT-inclusive) market types; selecting the wrong settlement convention when comparing a model to a market can distort implied probabilities and the EV calculation Official Rules of the National Hockey League 2024-25 (see bet365 hockey rules).

When working with hockey, either produce a regulation-only probability distribution from your full-game model or restrict your comparisons to markets that share your model's horizon. Documenting this alignment is a simple but essential control step before you compute edges.

Converting odds to implied probability and normalizing for fair comparison

Decimal, fractional, and American odds conversions

Converting market quotes to implied probability is the first numerical step: for decimal odds the formula is straightforward, implied probability = 1 / decimal odds. That conversion gives the market's break-even probability for each outcome before any adjustment for margin Implied Probability: Definition, Formula, and How to Calculate. Also see an additional guide on implied probability from OddsIndex.

Other formats map to decimal equivalents before conversion; fractional and American odds are common, but the safest approach is to convert everything to decimal odds first and then apply the 1 / odds formula to produce implied probabilities that are comparable across markets.

How do three-way markets differ from binary lines, and why must models match settlement rules?

Three-way markets allocate probability to home, draw, and away outcomes and are usually settled on regulation time; models must match that settlement horizon so implied probabilities and EV calculations align with the market's official rules.

Calculating implied probability and detecting overround

When you add the implied probabilities for the three selections in a 1X2 market they typically exceed 100 percent because bookmakers include a margin, known as overround. Detecting that overround is necessary before estimating your edge because raw implied probabilities are biased by operator pricing Understanding betting margins: how to calculate the bookmaker margin.

Normalization removes the overround and rescales the market probabilities so they sum to 100 percent, which produces the fair market probabilities you should compare with your model. The standard normalization divides each implied probability by the sum of all implied probabilities, then multiplies by 100 percent to recover a proper distribution.

Expected value in three-way markets: how to calculate and what it really means

EV formula applied to 3-way selections

Expected value for a single selection is computed from your estimated probability and the offered decimal odds as EV = (probability * decimal_odds) - 1, expressed per unit staked; a positive EV indicates that, given your probability estimate, the market price offers a favorable long-term return Expected Value (EV): Definition, Formula, and Examples.

Apply that formula separately to each of the three outcomes in a 1X2 market after you normalize the market probabilities. Only then can you judge whether the market price contains a long-run edge compared with your forecast.

Close up of calculator and notebook showing decimal odds and probability conversions on a tidy desk in Funded Plays palette for betting strategies

The Kelly criterion gives a formula to determine the growth-optimal fraction of your bankroll to wager when you believe you have an edge, and its appeal lies in maximizing long-run geometric growth under idealized assumptions. The exact formula for a single bet with decimal odds can be derived from the edge and odds and yields a stake fraction that is positive only when you have a positive expected value Kelly Criterion: Formula, Definition, and Uses.

Funded Plays Challenges

View FundedPlays Challenges

Why EV depends on model calibration

Expected value is meaningful only when your probability estimates are well calibrated; a model that systematically overstates the chance of an outcome will produce misleading EV figures even if the arithmetic is correct. Calibration and reliability checks are core prerequisites before you commit real bankroll to any selection.

Sizing stakes: Kelly, fractional Kelly, and safer alternatives for betting strategies

Full Kelly derivation and practical limits

a simple staking calculator to convert Kelly fraction into stake

Use as a sizing guide not a guarantee

Fractional Kelly and fixed-percentage staking rules

Full Kelly tends to be volatile in real-world sequences of bets because it assumes accurate, independent probability estimates and accepts large short-term variance; many practitioners therefore use fractional Kelly, for example half Kelly, or fixed-percentage staking to reduce drawdowns and psychological stress while retaining some of Kelly's growth properties.

Practical rules-of-thumb include capping the Kelly-derived fraction by a small fixed percentage of bankroll when uncertainty is high, and recalculating stakes only after updating model calibration and confidence rather than chasing short-term results.

Aligning your model with market rules and checking calibration

How to ensure model probabilities match settlement conventions

Before comparing your probabilities to market prices confirm that your model's event definitions match the market settlement, for example regulation-only 90 minutes for soccer or regulation-only for certain NHL bets; failing to do this is one of the most common and avoidable sources of error.

Document the horizon and any explicit exclusions in your model output so every EV calculation cites the exact settlement assumption used when comparing to a market price.

Simple calibration checks and goodness-of-fit

Basic calibration tools include reliability diagrams and summary metrics like the Brier score, which quantify how close predicted probabilities are to observed frequencies; using these checks on historical out-of-sample data helps ensure that EV signals are not artifacts of an overfit or miscalibrated model Expected Value (EV): Definition, Formula, and Examples.

If calibration shows systematic bias, adjust probability outputs or apply post-processing recalibration before using them for staking decisions and performance measurement.

Common errors and pitfalls when applying 3-way betting strategies

Mistakes in probability conversion and normalization

A frequent numerical mistake is mixing settlement horizons or failing to remove overround before estimating edge; using raw implied probabilities without normalization leads to biased EV estimates and poor decision-making in practice Understanding betting margins: how to calculate the bookmaker margin.

Other conversion errors include incorrect handling of fractional or American odds when converting to decimal form; standardizing to decimal odds first reduces the chance of a conversion error and keeps the arithmetic transparent.

Misapplying staking rules

Behavioral pitfalls include overbetting based on short-term winning streaks and treating Kelly outputs as guarantees rather than guidance. Technical pitfalls include applying full Kelly without accounting for model uncertainty or using Kelly on correlated bets without adjusting for dependence, both of which can increase risk.

Reduce these risks by setting conservative caps, using fractional Kelly, and logging each bet to review cumulative exposure and correlation effects over time.

Decision criteria: when a 1X2 selection is worth wagering according to betting strategies

Minimum EV thresholds and practical considerations

Turn an EV signal into a bet only after considering model confidence and bankroll limits; many practitioners set a minimum positive EV threshold and require acceptable recent calibration performance before acting, because small calculated edges can evaporate under model error or line movement Kelly Criterion: Formula, Definition, and Uses.

Also assess market liquidity, typical line movement patterns, and whether the edge remains after reasonable slippage assumptions; if you cannot execute the stake at the quoted price the realized EV will be lower than the calculation.

Risk-adjusted decision rules

Incorporate drawdown tolerance into your decision rule: aggressive players may accept higher variance while cautious participants prefer smaller fixed-percentage stakes despite similar EVs. Document the rationale for any deviation from the model's recommended stake to maintain discipline and reproducibility.

Keep a rolling review of performance and adjust minimum EV thresholds if calibration or market conditions change materially.

Practical example A: Evaluating a soccer 90-minute 1X2 market

Step-by-step implied probability and EV calculation

Start with the decimal prices for home, draw, away and convert to implied probabilities with 1 / decimal_odds. Remember to treat this as a 90-minute market if the listing does not include extra time, since Laws of the Game specify that regulation time is the default settlement for normal match markets IFAB Law 7: The Duration of the Match.

$Minimalist vector spreadsheet showing normalized probabilities EV columns and a highlighted Kelly fraction cell in dark Funded Plays brand colors for betting strategies$

Next, sum the three implied probabilities to compute the market overround and normalize each implied probability by dividing by that sum, yielding fair market probabilities that you can compare to your model's forecast. Then calculate EV for any selection using the formula EV = (model_prob * decimal_odds) - 1 and decide whether the EV exceeds your minimum actionable threshold.

Interpreting results and choosing a stake

When you obtain a positive EV, convert the calculated Kelly fraction into a practical stake using fractional Kelly or a capped percentage of bankroll to control drawdown. Keep the logic transparent: record the model probability, normalized market probability, EV, chosen fraction, and the reason for any manual adjustment when you place the wager.

Consistently logging these steps helps with later calibration checks and ensures you are not confusing short-term variance with true model-driven edges.

Practical example B: NHL regulation-only versus OT-inclusive markets

How to treat regulation-only lines

NHL markets can be explicitly labelled as regulation-only; when so labeled those markets exclude overtime and shootout outcomes and must be matched to a model that predicts regulation results. Using the wrong settlement assumption will produce incorrect implied probabilities and EV calculations Official Rules of the National Hockey League 2024-25.

One standard workflow is to produce full-game probabilities and then map them to regulation-only probabilities by removing the probability mass attributable to overtime and shootouts and renormalizing the remaining outcomes for comparison to the regulation market.

Adjusting probabilities if your model includes overtime outcomes

If your model naturally returns full-game probabilities, explicitly document and apply the conversion method you use to derive regulation-only estimates, and show this mapping in your pre-bet checklist so any reviewer can reproduce the adjustment and its implications for EV.

Preferably match your model type to the market offered. If you must convert, be conservative about the uncertainty introduced and reduce stake sizes to account for that additional model risk.

A short checklist to apply 3-way betting strategies reliably

Pre-bet checklist

Before staking confirm the market settlement convention, convert odds to implied probabilities and normalize for overround, compute EV using your calibrated model probabilities, and select a stake using a pre-defined staking rule that accounts for model uncertainty.

Also check market liquidity and likely line movement, and ensure your planned execution method can capture the quoted price or a realistic slippage-adjusted substitute Implied Probability: Definition, Formula, and How to Calculate.

Post-bet record-keeping

After resolution log the outcome, stake, payoff, model probability, normalized market probability, and any deviations from plan. This record fuels calibration reviews and helps detect persistent biases or execution losses that can erode EV over time.

Use a simple spreadsheet to timestamp entries and tag bets by model version so you can measure performance across strategy iterations and review how our evaluations work.

Conclusion: how to use three-way markets within responsible betting strategies

Key takeaways

Three-way 1X2 markets allocate outcome probability across home, draw, and away and are usually settled on regulation time, so the most reliable process is to match your model horizon to the market and normalize implied probabilities to remove bookmaker margin before computing EV Understanding betting margins: how to calculate the bookmaker margin.

EV is a useful decision metric only when forecast probabilities are calibrated, and staking rules like fractional Kelly help manage the high variance that can follow from full Kelly recommendations.

Maintain disciplined record-keeping at Funded Plays, run routine calibration diagnostics, and treat any positive EV signal as an opportunity that still requires risk management rather than a guarantee of profit.

Next steps for practice and model testing

Start by applying the normalization and EV steps on a historical sample, keep thorough logs, and iterate on calibration using resources on our blog before committing significant capital. Where appropriate, prefer simulated or small-scale execution while you refine model assumptions and the mapping between full-game and regulation horizons.

What is a three-way bet?

A three-way bet, often called 1X2, is a market with three mutually exclusive outcomes: home win, draw, or away win, typically settled on regulation time unless stated otherwise.

Why normalize implied probabilities?

Normalization removes the bookmaker margin so the probabilities sum to 100 percent, allowing fair comparison between market prices and your model forecasts.

When should I use fractional Kelly?

Fractional Kelly is appropriate when you want to reduce variance and drawdown risk relative to full Kelly, especially if model uncertainty is material.

Apply the steps here first in simulation or with small stakes until your model calibration and execution are proven. Keep detailed records and treat every positive EV signal as conditional on model accuracy and execution quality. Responsible participation and disciplined risk control matter more than any single edge; use these methods to build consistent, testable workflows rather than chase quick wins.