Collective selection
1,284 votesFormation preference
4-3-3
48%
4-2-3-1
30%
3-5-2
22%
Interactive recorded sandbox
FET Fan IntelligenceThe crowd sees what data misses.
Turn your match knowledge into a visible collective signal. Pick a team, read the moment, rate performance, and compare fan instinct with an explainable AI model.
FET Invitational (Sandbox)
Harbor City
Home · HAR
Full-time recording
2–1
15 January 2026Harbor Sandbox Ground
Northern Athletic
Away · NOR
Fictional clubs and a recorded match created for product demonstration.
One match, three moments
Contribute in context.
Move through the recorded match timeline. Each phase captures a different kind of fan knowledge without presenting it as a live feed.
Your pre-match selection
Experiment freely. Formation changes preserve your current eleven.
Your starting eleven
Build the Harbor City lineup
Width from both forwards with three central passing lanes.
Northern Athletic goalHarbor City goal
Select any player, then choose an eligible substitute.
Nothing is added to the aggregate until the server accepts the signal. A sandbox success confirms storage behavior, not live match data.
Fan wisdom, made visible
See the crowd without losing the individual.
Every bar below is an aggregate of recorded signals. Your choices remain separate until the contribution API confirms them.
Recorded sandbox aggregate 1,284 participants
No live match or voting claimChecking for a connected fan-signal snapshot…
16,842
Signals in this snapshot
4-3-3
Preferred shape · 48% of formation votes
68%
Confident or encouraged
8.4/10
Malik Haddad · highest fan rating
What the crowd is feeling
Sentiment describes fan perception, not an objective match-state measurement.
Confident397 · 31%
Encouraged474 · 37%
Concerned287 · 22%
Frustrated126 · 10%
Reading the result
The aggregate leans positive, but the concerned and frustrated minorities remain visible rather than being averaged away.
Recorded AI comparison
Sandbox model outputFan instinct vs. tactical model
Illustrative 0–100 priority scores for the same fictional match. The AI series is a fixed sandbox model output, not a live recommendation.
Where they diverge
Wide overloads13 pt gap
Fans weight repeated wide attacks more strongly than the model.
Central compactness17 pt gap
The model places greater value on protecting central transitions.
High pressing11 pt gap
Both agree on pressure, while the model flags its energy cost.
Every voice stays visible
Distributions preserve agreement and disagreement, so minority readings are not hidden by one score.
Context travels with the signal
Match phase and recorded moment accompany each contribution, keeping reactions interpretable later.
Continue with FET AI
Ask the personal assistant to explain the shapes, roles, and trade-offs behind your selection.
Sandbox match recorded 15 January 2026 · API connection is optional for exploring the interface.