Skip to main content

Release Notes – Insight in an Instant: AI-Powered Match Summaries and more

AI Match Summaries and more

This release marks an exciting step forward in how match insights are delivered across the SciSports platform. It introduces new AI-powered summaries in both Team Match Reports and Tactical Team Reports, bringing faster and more intuitive interpretation of match dynamics. In addition, XML files for event-only and tracking-data matches are now fully aligned.

AI Match Summaries

BETA release - the new functionalities are provided in an early-access state and will continue to be refined as we gather insights from daily use and feedback.

PDF Team Match Report:

AI summaries are now available in all Team Match Report PDF's. These summaries provide an automated narrative of key match themes based on event data, offering an immediate interpretation of passing behavior, build-up patterns, chance creation and more based on event data only. You can find the Team Match Report PDF:

  1. Data center → Data sets

  2. Open a match

  3. Select Reports

For a full overview of the PDF Match Report structure, refer to the Knowledge Base article: Match Reports

The AI summaries appear as SciSports AI Match Analyst. The summaries support match review by highlighting performance patterns before diving into the detailed statistics and visualizations.

You can use the AI Match Analyst to quickly see where your team exceeded or fell below its benchmark. By comparing these insights with the benchmark chart, it becomes clear which behaviors were unusual for your team. This helps decide what to prioritize in the next training session.

(Tracking Data only) PDF Tactical Match Reports:

The PDF Tactical Match Report now contains a Match Team Summary generated by AI. It provides an automated tactical interpretation of the team’s behavior using both positional tracking data and event data, highlighting in possession and out of possession. These appear as the SciSports Tactical AI Assistant and can be found on the second page of the Tactical Match Report.

On this page a shape per Game-Phase is added as a graphic as well. This ties back to the in or out of possession phases. This gives you a high-level overview of the game shape.

For a full overview of the PDF Tactical Match Report structure, refer to the Knowledge Base article: Tactical Match Report

Please note: The summaries are generated automatically using AI and rely on the underlying match data. They offer a quick interpretive layer but may occasionally contain inaccuracies or phrasing that does not fully match a team’s preferred terminology. These outputs should therefore be viewed as supportive insights alongside the detailed tactical analysis in the report.

Other Updates

Tracking Data: First and second half XML’s added

To align tracking-data XML’s with event-only XML’s, separate first half and second half XML files are now available for matches containing tracking data.

With this release, both event-only and tracking-data matches now produce:

  • First Half SciSportsEventsXml

  • Second Half SciSportsEventsXml

  • Full Match SciSportsEventsXml

This ensures consistent file availability regardless of the match’s data sources.

Sequences added to Events-only XML exports

Sequences were previously included only in the XML exports for matches with tracking data. With this update, sequences are now also included in the Events-only XML export as well.

A sequence is a continuous period of play in which a team is in possession of the ball. A sequence can be differentiated with its start and end result. In this article you can find more information about sequences.

Please note:

  • When tracking data is available, sequences in the XML follow the same definitions used in tracking-based analysis (Sequence feature)

  • When tracking data is not available, sequences are generated based solely on event data. These differ in definition or context compared to when tracking data is available. The sequences in the events-only XML align with the sequences shown in the Sequences feature inside the platform.

In this article you can find the differences between the sequences for events-only and tracking data.

This ensures that sequences are now consistently accessible, regardless of whether the match contains positional tracking data.

For questions regarding these updates, contact the SciSports Support Team at [email protected]

Did this answer your question?