Gotham Tech · Day 11 · Facilitator note
Two hooks for Day 11 — no match data needed
The scheduled hook (Human Number Line) asks students to test a belief against Season 1 dashboard data. With no matches run yet, that second half can't happen. Pick one of the two options below — both need only bodies, tape, and a whiteboard.
What the original hook was actually doing: make a public claim → surface disagreement → let something settle it. Only the settler has to change. In both options below, the morning's proportional scaling lesson becomes the verdict instead of the dashboard — which lands Day 11's big idea harder, not softer.
Option A
Double One Stat
15–20 min · closest to the original
Keeps the Human Number Line format exactly — students physically commit to a belief — but swaps the question to one the morning's math will answer.
You need
- Four signs on the walls: SPEED · POWER · ENDURANCE · TURNING
- Nothing else. No bots, no dashboard.
Run it (10 min)
- Post the four signs before students arrive.
- Give the prompt, then stop talking and let them move:
"You get to double exactly one stat on your bot. Which one gives you the biggest jump in performance? No math, no talking to your neighbor — go stand under your pick."
- Once everyone has committed, read the room aloud: where the crowd is, which corner is empty.
- Take two or three one-sentence reasons. Do not correct anyone, and do not hint.
- Then hold the tension and move on:
"Remember where you're standing. By lunch you'll be able to prove who's right."
The payoff (5 min, at the end of the scaling block)
Bring them back to the signs and let anyone who changed their mind move. Then show why:
| Stat | Formula | Double it | Result |
| Speed | motor power = Speed × 5 | 4 → 8 | 20 → 40 (×2) |
| Power | charge = Power² | 4 → 8 | 16 → 64 (×4) |
Power wins, and it isn't close. Most students pick Speed — watching the room migrate to Power is the lesson. Speed is proportional (double in, double out); Power is not (double in, quadruple out).
Bridge to the lesson: the hook uses doubling because it's intuitive; the lesson formalizes the same idea at +50% (Speed +50% → +50% output; Power +50% → +125% charge speed). Say that out loud so they connect the two.
Exit ticket (replaces the data-citing one):
"I first thought ______ would give the biggest jump. The math shows ______, because ______."
Watch for
- Students migrating to whichever corner their friends chose — call for silent movement.
- Anyone arguing "Endurance, because the bot lasts longer." That's a good answer about matches, not about output — park it, don't kill it.
- Do not resolve it early. The whole payoff depends on the answer staying open until after the math.
Option B
The Growth Race
10–12 min · most physical, least prep
Two students become the two formulas and walk it out. Best if the room is restless or you want the shortest possible opener.
You need
- An open lane (hallway works), a start line of masking tape.
- If space is tight: run it on the whiteboard with two magnets on a number line instead.
Run it
- Pick two volunteers. One is Speed Bot, one is Power Bot. Both start on the line.
- Each round the stat goes up by 1. Each bot moves to their new total position, in heel-to-toe steps:
Speed Bot = stat × 5 · Power Bot = stat × stat
- Before each round, the class predicts who will be ahead. That prediction is the whole activity.
| Stat | Speed Bot (×5) | Power Bot (²) | Who's ahead? |
| 1 | 5 | 1 | Speed |
| 2 | 10 | 4 | Speed |
| 3 | 15 | 9 | Speed — ask them to predict the end here |
| 4 | 20 | 16 | Speed, barely |
| 5 | 25 | 25 | Dead tie — the crossover |
| 6 | 30 | 36 | Power, and pulling away |
Most classes call it for Speed at round 3 — it's been winning the whole time. The tie at 5 and the flip at 6 is the moment. Land it:
"Speed was ahead the entire time and still lost. Speed adds the same amount every round. Power multiplies by itself — it starts slow and then it runs away. That's the difference we're studying this morning."
Exit ticket:
"Speed Bot was winning at stat 3. Power Bot won by stat 6. In one sentence, why?"
Watch for
- Counting drift — have the class count the steps out loud together.
- Students predicting the flip early: ask them to say which round it happens on and hold them to it.
- Keep steps small (heel-to-toe). Power Bot needs 36 of them.