Running the program — 16 days, decision by decision.

If you only remember one thing from this page: the bots are a vehicle for the math, not the other way around. Students who leave with a working bot but no math means the program failed.

• It's a math program in a robotics costume. Proportional reasoning, slope, distance, quadratic, percent change, residuals, systems of inequalities. The 16-day arc walks students from "formulas are recipes" to "formulas are predictions I revise from data."
• One driving question stays on the wall for 16 days: "How can you use data, equations, and code to build a robot athlete that wins on math, not luck?"
• The four Program Goals — Transfer · Ownership · Failure · Teach — are evidence statements, not slogans. Goal 4 (a stranger completes 3 mini-challenges after a single student explanation) is the summative on Day 16. See the Toolkit Guide.
• The math destination lives in the Math Pre-Brief — read it once before Week 1. Five mathematical relationships, four understandings to leave with.

You can't run this program from cold. Three hours of preparation in the days before Week 1 saves the program from the most common failure modes.

• Read the four walkthroughs in order: Math Pre-Brief → Progression in Context → Evidence in Context → Resources in Context. The order matters — destination, journey, what to look for, what to reach for.
• Memorize the Loss Recovery Protocol word-for-word. You'll use it from Day 7 on. The script only works if you can say it calmly under pressure without improvising.
• Calculate all four stat formulas for every point value 1–8 by hand, in under 10 seconds each. Motor% = Speed × 5 · Charge% = Power² · decay rate = (20−E) × 1.5 · turn ratio = 0.2 + T × 0.02. If you hesitate when a student asks mid-match, you become unreliable at the moment that matters most.
• Lay out the room. Reserve wall space for: Driving Question · Commissioner's Rules · Bot Names · Leaderboard · Rivalry Cards. Tape the field down with black vinyl or black duct tape.
• Confirm your site's data infrastructure works — open the Spreadsheets Reference and check that you can see your per-site sheets. If you can't, escalate now, not on Day 1.

The math changes day to day. The structure of the day does not. The predictability is the point — it lowers cognitive load so students spend it on the math, not on figuring out what's next.

Match days are when the modeling cycle becomes visible. These three rules are how a match stops being drama and starts being data. Hold them like they're load-bearing — because they are.

These aren't optional context. Make sure every student can state them by the end of Week 1. The first is the math; the other two are how the game actually works.

• The Budget equation: S + E + T + P = 20. The = means balance, not arithmetic. Every point given to one stat is a point taken from another. This frame drives every allocation for four weeks.
• Own goals count. A ball crossing either goal line scores for that net's attacker. (Surface this on Day 6 before Match Day 1 — otherwise the first own-goal is a long argument instead of a learning moment.) See the Match Rules page.
• Robot Off the Field: facilitator places the bot at the nearest edge, at the center line, facing inward. Referee presses Button B to try to reset the match loop.

What students produce today is what you teach tomorrow. The Evidence in Context walkthrough goes deeper; this is the short list of what to read every evening.

Every pilot site has a code that drives data routing. You don't edit URLs or sheets — but you do need to know what's connected to what, and where your data lives.

• Your site code: NY1 = Gotham Tech · NY2 = Claremont · NY3 = South Bronx Community Charter · TN = Tennessee. The code shows up in dashboard URLs as cohortTag=NY1 etc. If a URL is missing the tag, the data won't route.
• Five spreadsheets to know exist (all linked from the Spreadsheets Reference — gated):
  • • Loadouts — shared across all sites. Where stat-card lesson submissions land.
  • • Dashboard (per-site) — where match data lands.
  • • Pre-Task Diagnostic (per-state) — Day 1 baseline.
  • • Post-Task Diagnostic (per-state) — Day 15 growth measure.
  • • What-If Analysis — shared across all sites.
• If data isn't landing, it's not your problem to fix. Note what's missing, when, and from whom — and escalate. Don't try to edit a sheet ID, redeploy anything, or change a URL.

The student-facing site lives at the same URL minus /teachers. Share the student URL with students; never share teachers.html. Everything below is gated by the same password.

The hardest part of running this program is what you don't say. These are four moments where over-helping breaks the learning.

Each of these has burned a program before. They're worth memorizing — not just reading.

Read these every time you prep. They're the kind of thing that's correct in this cohort but might be stale by the next one — and the kind of thing that's hardest to catch by re-reading the curriculum.

• Follow-Through formula changed 2026-05-30 from P × 80 to 120 + (Power × 35). The new ms values run from 155 (P=1) to 540 (P=12). The lesson pages were updated to match — if you find any doc that still says Power × 80, it's stale and worth flagging.
• Underdog Mechanic (Day 11): the last-place team gets 2 bonus stat points, so their four stats sum to 22, not 20. Announce publicly; check the math.
• The cohortTag in dashboard URLs determines which site's sheet receives data. If a teacher emails you a dashboard URL missing the tag, data won't route — check this before Day 6.
• Power² sweet spot is currently unconfirmed — the old "P9 overshoot" rule was tied to the previous Power × 80 formula. Frame it as a Season 1 question students can investigate, not a known answer.

These are the moments where you stop trying and ping the developer. They're all infrastructure problems, not pedagogy problems — different toolset, different person.

• A submission says "Sent" but no row appears in the expected sheet.
• The diagnostic dashboard shows empty data after the post-task, even though students submitted.
• Apps Script returns an error in the browser console when a student submits.
• A page that worked yesterday is now 404.
• You want to add a new pilot site (e.g., NY4) — requires code changes in the API + new env vars + new sheet setup.
• You think a formula is wrong on a student-facing page (like the Follow-Through bug that landed wrong values in the Loadouts sheet for months).