FAQs

❓ What is the Marker Score API?

The Marker Score API is a machine-learning-powered service that identifies accounts likely to be high-performing, sharp bettors (referred to as "markers"). These bettors consistently place winning bets before market close, outperforming expected outcomes and posing increased financial risk.

❓ What does the Marker Score represent?

The Marker Score is a probability value between 0.01 and 1.00 that estimates how likely a customer is to exhibit “marker-like” behaviour:

• Pre-match betting

• Beating the closing odds

• High long-term win ratios

The closer the score is to 1.00, the more confident the system is that the account behaves like a marker.

❓ How is the Marker Score calculated?

The score is derived from historical ticket data and behavioural indicators, including:

• Bet timing relative to market close

• Bet types and market patterns

• Win rates compared to market efficiency

• Frequency of flagged outcomes

This model runs daily, continuously refining scores for active accounts.

❓ What is the purpose of using the Marker Score?

Operators use the Marker Score to:

• Monitor accounts with a high predictive edge

• Apply stake restrictions or enhanced verification

• Support segmentation and CRM tagging

• Combine with other scores (e.g., Bot Score, sCCF, LBS) to form a holistic risk profile

❓ What input parameters are required?

To retrieve Marker Scores via the API, you need:

• bookmaker – your assigned MTS/Insight Tech bookmaker ID

• sub-bookmaker – sub-entity under the main bookmaker (use * for all)

• customer – customer ID (for single or batch queries)

Some endpoints also support lower, higher, startKey, and pagination controls.

❓ What is included in the API response?

Each response typically includes:

• id – the customer ID

• mscore – the Marker Score (between 0.01 and 1.00)

• srcId – internal compound source identifier for tracking

If the request returns multiple pages, a lastEvaluatedKey is included for pagination.

❓ How often is the Marker Score updated?

Marker Scores are updated daily for accounts that meet the data threshold. If a previously flagged account drops below the threshold, the “Marker” tag may be automatically removed.

❓ Why is a Marker Score missing for some users?

Marker Scores are only available if the account:

• Has sufficient betting history

• Has been active recently

• Meets the model's profiling thresholds

New or inactive users may not be included in the model yet.

❓ How should high Marker Scores be handled?

A high Marker Score indicates potential sharp behaviour. Operators may choose to:

• Increase monitoring

• Restrict bet limits

• Tag the account for manual review

• Trigger segmentation or verification processes

Note: The API does not enforce any action — decisions are at the operator's discretion.

❓ What are the API rate limits?

Limit Type	Value
Requests/second	10
Daily quota	200,000 requests
Minute quota	500 requests

If limits are exceeded, the API returns 429 Too Many Requests. Implement exponential backoff for retries.

❓ Can I adjust thresholds or scoring rules?

No, the scoring logic is handled centrally by Sportradar's AI systems. However, you can:

• Adjust how you react to specific scores

• Combine Marker Score with internal KPIs

• Set your own thresholds for triggering interventions

❓ Who should I contact for support?

For access issues, integration help, or data questions, contact:

• Your OAM

• Your assigned CI engineer

• Or submit a ticket via Sportradar support.