Player Profiling

The Player Profiling API OpenAPI specification is available for download here. Explore the API endpoints here.

The Player Profiling API is an endpoint that returns a set of customer-level scores, generated by multiple behavioural and risk detection models. It is designed to provide sportsbook operators with a consolidated view of customer risk profiles, combining several ML-based insights into a single response.

By calling this API, operators can receive actionable scores such as the Suggested Customer Confidence Factor (sCCF), Customer Confidence Factor (CCF), Bot Score, Late Bet Score (LBS), and Marker Score, each of which reflects a different dimension of player behaviour. These scores can then be used to drive dynamic trading strategies, personalised validation rules, bonus eligibility checks, or responsible gaming workflows.

This model is particularly valuable for sportsbooks that want to automate parts of their risk management without relying on a fully managed trading service. It enables decisioning at the time of bet placement, based on a player’s historical and current behaviour.

When it’s triggered

The Player Profiling API is typically called at the point of ticket placement, cashout placement or during validation. Operators invoke it using a customer identifier and bookmaker context. The API then retrieves the latest scores associated with that customer, based on the most recent activity and behavioural signals available.

Since the API operates in real time, it can be embedded directly into pre-bet flows, segmentation jobs, bonus engines, or alerting systems.

Input requirements

The API expects a request that includes the following key fields:

• bookmaker: Your assigned MTS or Insight Tech bookmaker ID

• sub-bookmaker: Optional for clients with hierarchical structures

• account: The customer’s unique account ID

These inputs ensure that the API can accurately locate and return the relevant model outputs for the requested customer.

A full breakdown of all accepted parameters and filtering options is available in the Endpoints section.

Output schema

The API returns a JSON object with an items array containing one or more customer objects, each populated with the relevant model scores. Each score is returned as a numeric value, typically floating-point, that can be used for thresholding, ranking, or segmentation logic.

Interpretation guidance

Each field in the response corresponds to a different behavioural or risk model:

• sccf: Suggested Customer Confidence Factor – generated by Sportradar’s machine learning model, this score is designed to power risk-sensitive validation logic, particularly within Insight Tech workflows.

• ccf: Customer Confidence Factor – reflects the current CCF value in use within the MTS Console. This score is applied during MTS ticket validation and is relevant for clients using MTS or a hybrid MTS/Insight Tech setup.

• botScore: Bot Score – Indicates the likelihood that the account is controlled by automation or scripted tools rather than a human user.

• lbs: Late Bet Score – highlights potential timing abuse by measuring how closely a bet was placed to the moment of relevant event change (e.g. goal, red card).

• markerScore: Marker Score – Flags accounts that exhibit behavioural patterns commonly associated with sharp betting, fraud attempts, or high-risk segments.

These values are meant to be interpreted relative to your own historical player base. There are no strict thresholds, but operators are encouraged to calibrate actions (e.g. delays, restrictions, bonus suppression) based on internal tolerances and compliance policies.

You can map these scores into existing rule engines or visualise them through the Insight Tech Console to support human-in-the-loop decisions.

For schema details and request/response validation, refer to the Endpoints page.