What 7th-grade pre-algebra actually looks like (and what to do when the variables show up).

The 30-second version.

7th-grade math is mostly four ideas, in this rough order: integers and the rules for negatives, expressions with variables, solving simple equations, and ratios and proportional reasoning. The cognitive shift underneath all four is the same: a number stops being a count of objects and becomes a thing you can manipulate by symbolic rules. That shift is harder than it looks, and it's where a lot of previously-strong math kids get quietly lost.

Where it actually stalls.

1. Negative numbers.

The hardest part of 7th-grade math, by a wide margin, is the rule that −3 × −4 = 12. Most kids accept it; few internalize it. The reason is that there's no clean physical analog. You can act out 3 × 4 with apples; you can't act out −3 × −4 with apples. We pile on metaphors (debt, temperature, elevators) but the kid mostly memorizes the sign rules and hopes for the best.

The slip: the kid signs the answer correctly half the time. Not because they don't know the rules but because under any cognitive load — a word problem, a multi-step equation, a clock running — the sign tracking falls off first. The arithmetic was fine; the sign got lost.

2. Variables and expressions.

3x + 5is not a number. It's a recipe for a number once you know x. That distinction lands at different ages for different kids. Until it lands, the symbol x reads as “something I should solve for, immediately.” Asked to evaluate 3x + 5 at x = 2, the kid writes 11 — fine. Asked to simplify 3x + 5 + 2x, the kid panics, because there's nothing to solve for.

The slip:“3x + 5 + 2x = 5x + 5” — correct. “3x + 5 = 8x” — the kid added the constant to the variable. The mental model fused the 5 into the x rather than seeing them as separate objects. The fix is the like-terms picture (boxes, color-coding, anything physical that separates the x's from the constants).

3. Solving equations.

The dominant instructional move at this age is “do the same thing to both sides.” It works. It's also where the kid stops thinking about what an equation means and starts manipulating symbols by ritual. Asked to solve 2x + 3 = 11, the kid runs through subtract three, divide by two, x = 4 — correct. Asked whythey subtracted three, the kid says “because you do the same thing to both sides.”

The slip:the kid loses track of what the equation is about. When the equation gets longer, the ritual breaks. The fix is the “undo” framing — “the equation says do this to x and you get 11; what's the un-do?” — paired with the “same thing to both sides” algorithm.

4. Ratios and proportional reasoning.

If 3 cookies cost 75¢, how much do 12 cookies cost? Easy. If 4 painters paint a room in 6 hours, how long does it take 8 painters? Easy. If 4 painters paint a room in 6 hours, how long does it take 1 painter? Stops being easy.The shift between “more is more” and “more is less” is one of the canonical pre-algebra cliffs.

The slip:the kid applies the proportional move to a non-proportional situation, or vice versa. The fix is asking, every time, “does it make sense that more of X gives you more of Y, or less?” before the kid picks an arithmetic strategy. The intuition has to come before the algorithm.

Three things to do at the kitchen table.

Move 1 — Don't hand them the calculator.

Calculators are fine for arithmetic. They are not fine for the conceptual moves of pre-algebra, because they short-circuit the work that's actually being learned. If your child reaches for the calculator on −3 × −4, ask them to predict the sign first. The prediction is the thing they're learning; the calculation isn't.

Move 2 — Draw the equation.

Pre-algebra is the last grade where every problem can be drawn. Equations as balance scales. Variables as mystery boxes. Negative numbers as positions on a number line. If your child is stuck on 2x + 3 = 11, draw it as a balance with two boxes-of-x and a 3 on the left and an 11 on the right. The picture makes “subtract 3 from both sides” mean something. The algorithm runs faster after the picture lands.

Move 3 — Read it back.

For any equation your child solves, ask them to read it back to you in plain English before they touch it. 2x + 3 = 11becomes “some number, doubled, plus three, gives eleven.” Now solve it as a story: the doubled-number-plus-three is eleven, so the doubled number is eight, so the number is four. The plain English builds the meaning the symbols were always pointing at.

What Koda does at this age.

Koda's pre-algebra coverage is about the verifier doing what the calculator can't. Every step your child writes — sign work, like-term combination, equation manipulation — is checked exactly by a deterministic symbolic-algebra engine. Stuck doesn't have to mean lost; the verifier knows precisely where the slip happened. The hint ladder hands the next move back, never the answer.

Pre-algebra explainer videos are the next major catalog expansion (math is deepest at grades 2-5 today; 6-8 ramps up over the next year). What ships at launch is the verifier and the hint ladder for every problem your child writes — same software, fully on-device, same on-device privacy as the rest.

What this is not a substitute for.

A real teacher with a whiteboard. The hardest conceptual moves at this age — what a variable is, what the equation means, why negative-times-negative is positive — land best when a fluent human walks through them at a whiteboard. Koda is the regular practice partner for after the lesson. It's not the lesson.

Math anxiety. If your child is stuck not because they can't but because they've decided they can't, the fix isn't more practice. It's a calmer adult, lower stakes, smaller chunks, and a long time. We can help with the lower stakes and the smaller chunks; the calmer adult is you.

If your child is in middle school.

The waitlist gets you launch details and early-bird pricing. Math expands to grades 6-8 over the year after launch — verifier and hint ladder work for any problem from day one. Drop your email and we'll write when there's something to ship.

For the engineering version of why Koda runs a deterministic verifier instead of asking the language model to grade the math: we don't trust the LLM to grade arithmetic.