Gotham Tech · 4 teams · Days 12–15
Match Schedule + How to Enter the Data
Part 1 is the round-by-round schedule — print it and put it on the wall. Part 2 is the dashboard, in the fewest steps possible, with one worked example.
Which day runs which round is now set by the v3 plan. Use
Part 2 (dashboard entry) and the pairings below exactly as written — but for the calendar, follow v3:
Monday runs the 3-round cycle twice (Rounds 1–6),
Tuesday afternoon runs it a third time as the Proving Rounds (7–9), and
Wednesday is the championship.
The four teams: Goon Bros · The Minions · Triple T · Liquify.
Both mats run at the same time — two matches per round, all four teams playing. A full cycle is 3 rounds, and by the end every team has played every other team exactly once (three matches each).
With both mats live, there's no bye team to ref — so each team staffs its own mat. Send the
driver to the controller and a
ball-runner to the mat edge. The two ball-runners share the mat work while their teammates drive:
- Put the ball back on the field whenever it leaves.
- Put rogue bots back on the field and press B to get them running again.
You float between the two mats — call "go" to start both, keep the clock, and settle any scoring dispute. One adult can run two mats this way as long as the ball-runners are handling their own field.
Part 1 The schedule
Day 12 · Thu Jul 23 — Match Day 1 (first ever matches — expect it to run slow)
| Round | Mat 1 | Mat 2 |
|---|
| 1 | Goon Bros v The Minions | Triple T v Liquify |
| 2 | Goon Bros v Triple T | The Minions v Liquify |
Only two rounds today on purpose. The first match day is always slow — bots fail, the protocol is new, two mats at once takes a beat to coordinate. If it's dragging, run just Round 1 and pick up Monday. Clean matches with real predictions beat rushed ones.
Day 13 · Mon Jul 27 — Match Day 2 (completes the cycle)
| Round | Mat 1 | Mat 2 |
|---|
| 3 | Goon Bros v Liquify | The Minions v Triple T |
After Round 3 the cycle is complete — every team has played all three opponents. Post the standings; this is the first honest leaderboard of the season. Day 13 is also the stat-reallocation window, so keep the match load light and give that the time.
Day 14 · Tue Jul 28 — Match Day 3 (rematches — same pairings, new predictions)
| Round | Mat 1 | Mat 2 |
|---|
| 4 | Goon Bros v The Minions | Triple T v Liquify |
| 5 | Goon Bros v Triple T | The Minions v Liquify |
| 6 | Goon Bros v Liquify | The Minions v Triple T |
A full second cycle of rematches: the same opponents, but now with a revised loadout and an equation-based prediction. That contrast is what makes the residual work on Day 14 mean something. If you're short on time, stop after Round 5 — don't skip the recording.
Day 15 · Wed Jul 29 — Championship
Seed #1–#4 by record across Days 12–14 (wins first, then goal difference).
| Match | Who plays | | Refs |
|---|
| Semis | Mat 1: #1 v #4 · Mat 2: #2 v #3 (both mats at once) | ball-runners |
| 3rd place | Semi loser v Semi loser | the two finalists |
| FINAL | Semi winner v Semi winner | everyone watches |
Give it a real ending. Run both semis on the two mats at once, then one mat for the final so the whole room watches. Announce the bracket at the start of the day so teams know what they're playing for, and crown a champion before the Exhibition on Thursday. (The 3rd-place match is optional — drop it if you're short on time.)
How long a round takes
| Step | Time | Notes |
|---|
| Write predictions | 2 min | All four teams write a goal number before hearing the opponent's. |
| Both matches | 3 min | Mat 1 and Mat 2 run at the same time. You call "go" for both; ball-runners return the ball and reset any rogue bot (press B). |
| Record + loss recovery | 5 min | Both losing teams run loss recovery now, not at the end of the day. |
| Total per round | ~10 min | Two matches every round · a full three-round cycle ≈ 30 min. |
Part 2 Entering data in the dashboard
The single most important rule: the prediction goes in BEFORE the match, the score goes in AFTER. A prediction typed in after seeing the result is worthless as data — it's the whole point of the exercise.
Once, at the start
- Open the dashboard and pick your teamTeam tabs run across the top. If a team isn't there, click the + to add it. Each team uses their own tab — four teams, four sets of data.
- Enter the Stat LoadoutType the four stats. They must total exactly 20. Do this once — the loadout is then locked for the season. The only edit is at the Day 13 reallocation window.
Before each match (takes 30 seconds)
- Fill in the top five boxes, then click "Add match row"Round · Day · Opponent · Prediction source · Predicted goals.
Leave Goals For and Goals Against empty.
- "Prediction source" is requiredThe Goals boxes stay grayed out until you pick one. It's just how the team got their number — early on choose
Estimate or All stat formulas; from Day 13 on, Scatter plot equation or Student equation.
After the match (takes 30 seconds)
- Type the score into that rowFind the row you just made and enter Goals For and Goals Against. Error and W/L fill themselves in — nobody calculates those by hand.
- Add the two sentencesWhat happened? — one sentence, not the score (the score is already in the boxes). What are you testing next round? — one change to strategy or the launch sequence, not to their stats.
Worked example — Goon Bros, Day 12, Round 1 vs The Minions
Before the match, Goon Bros fill in:
| Box | What they type |
|---|
| Round | 1 |
| Day | 12 |
| Opponent | The Minions |
| Prediction source | All stat formulas |
| Predicted goals | 3 |
| Goals For / Against | — leave blank — |
They click Add match row and go play. After the match (they won 2–1) they type into that same row:
| Box | What they type |
|---|
| Goals For | 2 |
| Goals Against | 1 |
| Error (automatic) | +1 — predicted 3, scored 2 |
| W/L (automatic) | W — 2 > 1 |
| What happened? | Our launch sequence got us to the ball first every time. |
| Testing next round? | Start wide left instead of straight at the ball — we lost the tip-off both times. |
Stats are locked once the season starts. "Testing next round" means a change to strategy or the launch sequence — approach angle, which side they attack from, when they kick, re-tuning the launch code. It does not mean changing Speed / Power / Endurance / Turning.
The only time stats change is the reallocation window on Day 13 — one change, recorded before and after. That before/after is exactly what percent change runs on during Day 14, so a team that quietly edits stats mid-season breaks their own data.
That +1 error is the number the whole week runs on. On Day 14 they'll look at a column of these and ask: are we consistently over-predicting? Is the model getting better? That only works if the prediction went in first.
The five mistakes to watch for
- Filling in the score before the match. The one that ruins the data. Predictions first, always.
- Skipping "Prediction source." The row won't save and the Goals boxes stay grayed out — this is the "it's broken" complaint you'll hear most.
- Logging into the wrong team's tab. Each team owns their tab. Assign one Data Owner per team.
- Writing the score in "What happened?" That box is for the reason, not the result.
- Leaving blanks "to fill in later." Later never comes. Record between rounds, while the match is fresh.
All four teams record every round — with both mats live, nobody sits out. Each team's Data Owner adds one row per round for their own match.