Attacking-Channel Entropy

What it is

Information entropy applied to which vertical channels an attack uses, read off progressive pass/carry (PP) corridors. A perfectly one-sided attack has low entropy; a genuinely unpredictable one spreads across channels and scores high. Derived directly from information theory; needs a normalised-entropy (H_norm) benchmark and a tagging pipeline.

Built from

Grounded in existing KPIs/data: PP.

Development status

Draft — measurement method defined; benchmark and validation still to come.

2026 World Cup example

Approximate
A normalised entropy of line-break location across the pitch, computed from the PMSR zone breakdown rather than proper channel tagging, so treat it as a rough stand-in for this KPI. Quartiles: win rate 23.1% → 46.2% → 34.6% → 50.0% — directionally upward from bottom to top but not monotonic in between. Weaker and noisier than the report's other proxies; not strong enough to call this validated. Read the full report →