Cashout API (cashout-build)

The Cashout API (cashout-build) calculates the recommended cashout value for an open ticket. It combines live odds, real-time probabilities, market volatility, bettor profile, and contextual risk factors to produce dynamic, fair, and profitable cashout offers in real time. This API is part of Sportradar's broader Cashout Engine and is intended for integration in MTS and hybrid MTS/Insight Tech Services environments.

What it does

The cashout-build API evaluates the value of a bet using multiple calculation strategies, tailored to both the current market state and configurable operator logic. At its core, the API provides not just the computed cashout offer, but also the key inputs that influenced it, enabling full transparency and validation flexibility.

When selection-level probabilities are available from the live offer, the API calculates the fair cashout value by evaluating each live combination in the ticket independently. The calculation follows the formula:

Fair Cashout Value = Σ [Combination Stake × Combination Odds × Combination Probability]

This method accounts for standard win/loss logic as well as more complex market resolutions such as Asian handicaps or over/under lines with half-win/half-lose or voided outcomes. Each selection's probability distribution (win, refund, half-win, half-lose) is parsed and weighted according to the market's resolution rules. These weights are used to determine the expected value of each valid combination, which are then aggregated to form the total fair cashout value.

If probability data is missing or incomplete (due to market suspension, limited coverage, or delayed feeds) the engine can automatically query the Probability API to fill in the gaps. This fallback mechanism expands model applicability to long-tail sports and improves resilience during volatile market conditions.

In addition to the fair cashout value, the API supports margin-adjusted cashout calculations, which are configurable by the operator through the Insight Tech Services UI (coming soon!). The margin logic modifies the fair value according to dynamic conditions, such as:

• Time remaining until event start

• Pre-match and live event ratings

• Market-specific factors (volatility, risk profile)

• Customer Confidence Factor (CCF)

• Original ticket odds and compensation parameters

The final recommended cashout offer, returned in the cashout field, reflects these adjustments, providing a balance between attractiveness to the bettor and margin preservation for the operator.

Beyond pricing, the API also surfaces additional decision-support data:

• Live Time Delay recommendation and applied LTD

• Selection-level confidence factors, event ratings, and applied LTD offsets

• Bettor profile metrics (confidence, bot score, marker score, late bet score)

• Bet-level settlement status and market result context

This means the cashout-build response contains both the "ingredients" and the "cake": a complete basis for decisioning, auditing, and optional override logic. Whether used to automate cashout decisions or power bettor-facing widgets, the API enables fair, risk-aware, and customisable cashout strategies at scale.

When it's triggered

This API is typically triggered whenever a cashout value needs to be computed or updated in real time. Operators may call the API:

• When rendering a live cashout offer in the user interface

• At regular polling intervals during live event progression

• As part of a backend validation or automated auto-cashout flow

The calculation is dynamic and reflects the latest available data, which may change frequently based on match events, odds updates, and customer profile inputs.

Input requirements

The request must contain a cashout-build operation payload with the following fields:

• A valid operatorId identifying the client in the system.

• A cashoutId to uniquely identify the request instance.

• A ticketId and corresponding ticketSignature referencing the original ticket.

• Payout details as known at cashout time.

• Operation metadata such as correlation ID, timestamp, and version.

This request structure ensures that the engine can match the cashout request to the ticket context and retrieve selection-level probabilities from offer data or external APIs. More details available in the Cash-out Build Request and Response section.

Output schema

The API responds with a detailed structure containing the recommended cashout values and contextual information that supports offer validation and display logic. The output includes:

• The cashout section with three sub-values: the maximum possible payout at the time of the cashout placement, the fair cashout value (based on current probabilities), and the adjusted cashout value (based on operator configuration).

• The Live Time Delay (LTD) recommendation, including configured, suggested, and applied delays and any dynamic offsets.

• Bettor profile suggestions and scores, such as CCF, Marker Score, Bot Score, and Late Bet Score.

• Selection-level information across all bets, including market identifiers, applied LTDs, current probabilities, market resolution states, and result indicators.

• Exchange rate data for multi-currency support.

The response schema is designed to serve both display and validation use cases. More details available in the Cash-out Build Request and Response section.

Interpretation guidance

The cashout field in the response contains the primary recommendation that can be offered to the end user. Operators should interpret this value as a model-generated suggestion that already accounts for live odds, volatility, bettor profile, and operator-specific configuration.

The fairCashout represents the calculated fair value of the ticket assuming neutral margins, while the maxPayout reflects the theoretical maximum payout of the ticket, taking into account any settled selections. The adjusted cashout value reflects configured margins or compensation factors applied by the operator.

LTD recommendations can guide UI-level locking mechanisms or be used to override pre-configured values in cases of profile-based risk. The bettor profile scores can support offer segmentation or fraud detection.

Operators are encouraged to use the model's components as a reliable benchmark for validating internal logic, defining thresholds for deviation handling, or powering automated offer acceptance workflows. The combination of transparent inputs and calculated outputs makes this API suitable for both bettor-facing UI layers and internal validation pipelines.

The returned cashout value serves as the suggested operator benchmark and may be used directly or fine-tuned based on operator-specific policies. When margin-adjusted configurations are in place, the cashout value reflects a strategic balance between risk mitigation and bettor engagement.

Operators may choose to lower the offer value from the fair calculation using fixed or dynamic margin strategies, or uplift it through compensation factors for high-odds multiples or early-stage bets. Where applicable, high Late Bet Scores, low customer confidence, or system configurations may restrict the ability to skip live time delay validation.