Platform: BBC micro:bit v2 + Cutebot Pro (MakeCode). Team roles: Coder, Fabricator, Analyst — rotate weekly. Daily structure: Hook (20–25 min) → Challenge Block (45–70 min) → Snack / Lunch (20 min at Crosstown 8:00–12:00; 60 min at NYC sites 10:00–3:00) → Match-Build-Make (75–90 min) → Open Lab (20–40 min) → ENDS Closing Ritual (10 min).
If these happen consistently, the program still works. Everything else strengthens the experience — these are the floor.
Confusion is useful only when students are still moving, trying, writing, testing, or discussing. If students freeze, withdraw, or wait passively, intervene.
This program is not hard because the math is hard. It is hard because students are building, competing, losing, recovering, and explaining in public.
These commitments protect the program's learning sequence. When any of them slips consistently, student thinking degrades faster than most facilitators expect.
Cut extended debrief first. Do not cut prediction, match logging, loss recovery, or closure.
Seven seconds activates thinking across all confidence levels. Under 3 seconds, only high-confidence students respond. Protect this whenever possible. If a student is genuinely lost rather than processing, respond to them directly.
Default expectation: every losing team completes all three steps before their next match begins. This is the single highest-dropout-risk moment in the program. If time collapses, preserve the core step — Step 2 (locate the decision in the data).
Under normal conditions, every match gets a prediction logged before it starts and an actual logged immediately after. A prediction entered after seeing the result is worthless as data. Circulate and verify during every post-match window.
"Did we do this right?" → "What would a correct answer look like? Is that what you see?" The match result is the teacher. Your job is to keep students using their own data, not to validate in real time.
Before the meeting, confirm proposals include two data points and name one math concept. During the meeting, redirect accuracy challenges to the room. Weak proposals go to the floor — the debate is the learning.
Post the four categories before Day 1 and leave them up. Students who struggle to start a gap sentence need the scaffold visible throughout, not just mentioned once at the beginning of the program.
Protect ENDS daily. It is not a wind-down — it is the day's consolidation moment. Each type targets a different reflection goal. If time collapses, run a 3-minute version rather than skipping it entirely.
You will make judgment calls every day. Make them in the spirit of the framework: keep thinking in students' hands, use data as the authority, treat confusion as information. The tools are yours to use flexibly — the principles are not.
Frame the diagnostic before distributing it using the exact script in the Day 1 risk box. Use non-bot contexts (speed, distance, ordered pairs) so no student feels disadvantaged by not yet knowing the stat system. On Day 15, use the "comparing you to yourself" frame explicitly.
Memorize the 3-step script before Day 1. Step 1: acknowledge the emotion (genuinely, briefly). Step 2: locate the decision point in the data. Step 3: restore student ownership with a hypothesis.
Your only intervention during the Commissioner's Meeting should be redirecting accuracy checks to the room: "Does this proposal use the data correctly? Can anyone verify the math?" Clarifying how the budget formula works is fine. Improving what proposals say is not.
These are not decorations. They create multiple ways to matter besides winning.
Ten specific things to know, make, or practice before Day 1. Each represents a predictable new-facilitator failure when absent.
Post before Day 1. Leave them up. Students who cannot start a gap sentence need the scaffold visible, not just mentioned once.
This section explains why specific design choices were made. Not required reading before Day 1.
Improvised emotional responses during losses tend to cut students off from their data — either by rushing past the emotion, or by dwelling in it without redirecting to action. The 3-step script creates a reliable container: acknowledge the emotion briefly, locate a specific decision in the data, restore student ownership with a hypothesis. Facilitators who internalize this structure handle the highest-stress moments with consistent quality.
Research on classroom wait time consistently shows that under 3 seconds, only high-confidence students respond. Seven seconds is a specific, noticeable target — long enough to feel uncomfortable the first time, which signals to facilitators that they are doing it correctly. "Just wait" produces 2–3 seconds in practice.
Each stat introduces a new mathematical relationship: Speed (linear proportion), Endurance (rate of change and decay), Turning (coordinate geometry), Power (non-linear scaling). Introducing all four on Day 1 eliminates the learning sequence — students would have no lived experience of each stat's behavior before being asked to allocate across all four.
Student ownership peaks at the Commissioner's Meeting and the pre-match pitch. Both moments require students to make and defend a public mathematical argument without a safety net. Facilitators who improve proposals before they reach the floor remove the productive struggle. A weak proposal going to the floor is not a failure — it is the design.
The final award signals what the program values: mathematical communication, not winning. Students who finished last in the competition can win this award. It reframes the entire four-week arc in the final ten minutes — from "who built the best bot" to "who understands this well enough to teach it."