Everything in this phase is about closing gaps in your own readiness before you stand in front of students. Three of the four resources here are not things students ever see — they're scaffolding for you.
Ten specific items: every formula calculated by memory in under 10 seconds, the four-step loss-recovery script practiced aloud, the room laid out with the wall reserved for Driving Question + Bot Names + Leaderboard + Rivalry Cards.
Not open Day 1 with anything on this list unfinished. The list exists because the ten most common new-facilitator failures correspond exactly to skipping one of these ten items.
The four program goals with named evidence statements, the six-block daily structure, room setup, and a day-by-day arc with risk moments flagged.
Recognize success in real time. The evidence statements ("a residual: predicted vs. actual, named", "an explanation revised after a visitor got stuck") are the look-fors you'll use every block.
An 8-question, 12-minute baseline covering proportional reasoning, slope, distance, and quadratic thinking — no bot framing. Results land per-concept on the dashboard.
Decide which Day 1–5 mini-blocks need extra scaffolding for which students — and have a baseline you can compare against on Day 15 to show actual growth, not vibes.
Offline MakeCode install, WebUSB pairing flow, and a reference launch-sequence MakeCode block teams will adapt for Day 6.
Catch hardware problems before students do. A bot that won't pair on Day 1 burns the trust the Failure Bow is supposed to build.
Every day has the same six-block shape (Hook → Challenge → Snack → Match-Build-Make → Open Lab → ENDS). These are the resources you reach for inside the loop, not the ones you read once before the program starts.
The day's six blocks with durations and the per-activity facilitator description — the same description that's now mirrored across days-default.js, days-tn.js, and facilitator-moves.js.
Know where to stand at each transition and what the next 30 seconds of the activity must accomplish — instead of reading the script mid-block.
Each activity links a teacher-only quick card (pink pill) and a full facilitator script (Google Doc). Quick cards are scripted lines, common-error watchlist, and exit-ticket expectations. Cheat sheet collapses all four formulas onto one page.
Run the activity from the front of the room without paper — and intervene with the right talk-move when a team is stuck, not whatever comes to mind.
Every activity in the Teacher Schedule lists the worksheet, the interactive lesson page, and (where it exists) the MakeCode link students will use. The Student day view is the same page students see — no teacher-only fields rendered.
Pace-check against the lesson, not against your gut. If you can't predict where a team will be at minute 20, re-read the lesson — students may be on a section you haven't seen.
For most activities, three real student-work samples at Approaching / Meets / Exceeds, with a one-line explanation of what distinguishes each level. Also surfaced inline in the Teacher Schedule as purple Exemplar pills next to each activity.
Push a team using a specific language move: "Your work looks like the Meets sample — what's missing to make it Exceeds?" Better than a generic "add more detail."
Pre-Season guide walks through the Days 1–5 workflow: stat cards, the first official allocation, dashboard loadout entry, and the three-part gap sentence (formula predicted / bot did / gap). Intro Script is the 12–15 min Day 5 walkthrough that lands all teams in the dashboard the same way.
Lock in the data-entry standard before Season 1 begins. Every team enters loadouts the same way and writes gap sentences in the same format — so when Season 1 matches start, residuals are comparable instead of each team using its own shorthand.
Thirteen browser-based concept pages — Unit Rate, Ratio, Proportional Scaling, Linear Decay, Coordinate Geometry, Arc Approximation, Slope, Budget, Percent Change, Graphs, Linear Equations, Data & Residuals — plus a Playground that runs all eight core visualizations on one canvas for board demos.
Hand a stuck team a self-contained interactive instead of a re-explanation — they manipulate it, you keep moving. Use Playground when you need a quick whole-room visualization (e.g. "watch what happens to Charge% when I drag Power from 4 to 8").
The match itself isn't the assessment — the loop around it is. Prediction before the bots move, residual calculation after, and a structured recovery for the team that lost.
Walks through the match-day workflow across Seasons 1 and 2: equation-based prediction before the match, real-time match logging, residual calculation, and the gap-sentence entry that ties each match back to the team's model.
Require a written prediction before any bot moves — the rule that converts a match from "drama" into "data."
Cutebot Pro Soccer simulation: enter two stat loadouts, run a match, see goal/time/residual outcomes. Useful in Open Lab and Commissioner's Week reallocation.
Direct teams toward bench-testing instead of live testing when stakes are high (Season 1 final, Championship), and surface the loadouts most worth running on the field.
A four-step script. Step 1 is 30 seconds of silence after naming the feeling. Step 2 is the dashboard. Step 3 is the next match. The script is the resource — say it the way it's written.
Hand a frustrated team a specific math move instead of generic reassurance. Run it for every losing team, every time — not just the ones who visibly need it.
Three sample residual analyses per activity. The Exceeds samples typically name which equation parameter (slope vs intercept) the residual implicates; the Approaching samples stop at "the bot was off."
Pick the team to call on for share-out — and know in advance whether their finding will move the room forward or stall it.
Every Friday you make the same call: who's behind, on what concept, and what fits inside next week to address it. The dashboard is the system that tells you, and Cold Recall (Day 10) is the formal check.
Per-team match history, prediction accuracy, residual size and direction. The Best Prediction column tracks model accuracy independent of win/loss, so a team that loses with tight residuals is visible.
Pick the one or two teams to target in next week's Open Labs — and decide whether the gap is a model problem, a code problem, or a hardware problem.
Five questions, individual, no notes, no devices, no grade. The point isn't to score it — it's to tally who can still write the Endurance decay or Power² formula cold. When the tally shows a class-wide gap, the matching interactive concept page (Linear Decay, Proportional Scaling, etc.) is the re-teach.
Fold the re-teach into the next Challenge Block with specific names attached — and send students to the concept page in Open Lab instead of re-lecturing.
3-minute presentation format, math-verification protocol, the rule that an objection only counts if it names a number and a calculation showing differential impact. The reference doc (teacher-only) holds the worked examples.
Dismiss "I don't like it" arguments without a number and hold the room to the standard the rest of the program has built.
Two summative measures: the post-task diagnostic (math you can measure) and the visitor exhibition (math students can teach). Both produce reportable evidence.
Parallel-form post-diagnostic on Day 15. The Growth Dashboard shows per-concept percent change vs. Day 1 baseline, anonymized.
Report concrete growth to stakeholders — and identify which concept(s) underperformed for the next cohort's redesign.
Each visitor attempts three mini-challenges (unit rate, coordinate distance, proportional scaling) after the student's explanation. A Visitor Completion Tracker logs first-try completions per team.
Score Goal 4 ("the ability to teach") as a binary observable — did a stranger complete all three challenges on the first try? — rather than a judgment call.
Every artifact from four weeks on the wall — bot names, leaderboard, Commissioner rules, scouting reports. The Awards Ceremony script names each award with its data criterion (e.g. "Most Improved — the highest Season 1 → Season 2 win-rate percent change") so the math stays visible in the celebration.
Close on what the program produced — not a single winner, but a room of students who can explain algebra using their own data.
Three reflective toolkit pages aimed at the facilitator, not the student. Failure Modes catalogs the moves that quietly sink a session (rescuing too early, dropping the exit ticket, narrating the gallery walk). Experience Map traces the emotional arc across the 16 days so you can see which moment a struggle actually belongs to. Battlefield Cards are the talk-moves and recovery scripts in print-and-keep-in-pocket form.
Name what didn't work and pick one change for next time — instead of relying on vague "I should run that differently." Between-cohort reflection is the only place these resources land; during a session you reach for the scripts instead.
The reviewer asked us to walk five resource categories through the lens of implementation. Each category is represented in the page above — woven through the arc, tagged on every card. Counts are tag occurrences across all decision moments.