Graf(x) reverses the usual direction of function problems. One player draws a card with a linear function—say, "Multiply by 3 and Add 10"—and keeps it hidden. Teammates call out numbers; the player applies the rule and announces only the output. Input: 4, output: 22. Input: 7, output: 31. From these pairs, the team reconstructs the function.
This setup creates what's called a "black box" problem: students can probe the function's behavior but can't see its internal structure. They're working inductively, building the rule from observations rather than evaluating a given rule—closer to how mathematicians actually discover patterns than how algebra is typically taught.
Students reconstruct functions from behavior, not just evaluate them.
The game works exclusively with linear functions in slope-intercept form: f(x) = mx + b. Each function multiplies the input by some value m, then adds (or subtracts) some constant b. Simple structure, but figuring out both parameters from limited trials requires systematic thinking about how linear functions behave.
Students quickly learn that certain inputs reveal specific information. Testing zero isolates the constant term. Testing consecutive integers reveals the slope through constant differences in outputs. These strategies aren't taught—they emerge from the problem structure itself.
The m coefficient represents slope—how much the output changes per unit input. Students develop intuition about rate of change through direct experience: a function with m = 8 produces outputs that grow much faster than one with m = 2. The b constant shifts everything vertically without changing the rate of growth.
Each Graph Card shows the function's visual representation alongside the rule, so after discovering a function, students see what it looks like graphed. This connects their algebraic reasoning to geometric intuition—another link that formal instruction sometimes skips over too quickly.
Teams balance speed with systematic data collection.
The four-minute timer creates an interesting tension. Test too few values and you're guessing; test too many and you run out of time. Students learn to judge when they have sufficient evidence—a useful metacognitive skill that extends well beyond this specific game.
The collaborative structure matters pedagogically. One student might notice "it's going up by threes," while another tests the hypothesis deductively: "If it's times three, then seven should give twenty-one, but we got twenty-three, so something's being added." Different cognitive approaches strengthen each other.
Because one player holds the card and evaluates the function while others discover it, everyone practices both skills: careful function evaluation under time pressure, and pattern recognition from limited data. Roles rotate each round.