Method
Consequence is the curriculum.
Every FenriX game runs the same loop. The design is old, tested learning science; the arena makes it stick.
The loop
Decide. Pay. Debrief. Replay.
01 / Decide
An authentic decision with real trade-offs: price against margin, capacity against payroll, argument against clock.
02 / Pay
The consequence lands in the shared world, visibly, and it persists. Rivals feel it and answer.
03 / Debrief
The match ends in a readout: what each choice cost, what it earned, and what to try differently.
04 / Replay
Round-to-round adjustment inside the match is the replay. Five rounds means five chances to act on the last debrief.
Inside the AI
Not everything here is an LLM. That is deliberate.
| Game | Engine |
|---|---|
| Bakery Bash | Deterministic shared demand model. Math, not an LLM. |
| Trading Floor | Order book fed by historical price series and simulated traders. |
| Front Office | Samples real NBA performance distributions. |
| Holdings | Prices set from comparable-sales data; demand shifts each round. |
| Alpha Hunt | License-free and simulated market data. Scoring runs out of sample. |
Limits first. Front Office samples real performance distributions, not a scouting verdict. Trading Floor runs on historical price series and simulated order flow, never a real-money feed. Alpha Hunt ships license-free and simulated market data, and every strategy is judged out of sample. Where a system is still being built, its card says in development.
No validated skill-lift metrics exist yet. Pilots are how that evidence gets built, openly, with partners who can check the work.
Reading
Designed in line with.
Kolb, Experiential Learning (1984)
the decide-consequence-reflect cycle is the spine of every match
Fanning & Gaba, The Role of Debriefing in Simulation-Based Learning (2007)
why every match ends in a structured debrief, not a score alone
Ericsson, Krampe & Tesch-Romer, The Role of Deliberate Practice (1993)
tight feedback loops on hard decisions, repeated round after round
Roediger & Karpicke, Test-Enhanced Learning (2006)
retrieval under pressure beats re-reading; playing beats reviewing
Perkins & Salomon, Teaching for Transfer (1988)
varied scenarios across rounds so the lesson leaves the game
Freeman et al., Active Learning Increases Student Performance in STEM (PNAS, 2014)
doing beats listening across 225 studies; an arena is active learning at full volume
Sailer & Homner, The Gamification of Learning: a Meta-analysis (Educ. Psychol. Rev., 2020)
game elements move cognition and motivation; effects are real and modest, so ours carry a whole match, not points and badges
Sinha & Kapur, When Problem Solving Followed by Instruction Works: Evidence for Productive Failure (Rev. Educ. Res., 2021)
struggling before instruction beats instruction first; losing a match is part of the syllabus
Designed in line with, never proven by. The proof gets built in pilots.
Your data
Plain words.
Games log in-game decisions to power scoring and debriefs. Players see their own debrief; instructors and facilitators see the round-by-round data of their own cohort. Nothing is sold, nothing is profiled.