Chemistry has a reputation as a memorization subject — all those reagents, all those named reactions, the flashcard decks that never end. But watch what a chemistry exam actually asks: not "name this reagent" but "what product forms, and why not the other one?" The exam tests mechanism and discrimination. Memorizing reagents is studying for a test that doesn't exist.

The students who drown in chemistry are the ones collecting named reactions like trivia. The ones who survive are the ones who can tell the electron-flow story — and, crucially, say why the story doesn't go the other way.

The method: stop collecting facts, start generating mechanisms. For each lecture, produce the electron-flow story from a blank page, then write one "why not the other product?" item. The memory for reagents attaches after the story is retrievable — because a reagent you understand is a label on a process; a reagent you memorized is a word that evaporates.

Why trivia-style studying fails chemistry

The reagent-flashcard approach — "LiAlH4: reduces esters to alcohols" — trains recognition of the card, not the decision the exam requires. The exam doesn't flash "LiAlH4" and wait for a fact; it shows you a substrate and conditions and asks which of two plausible products forms. Knowing the reagent's name is the easy half; knowing why it wins over the alternative is what's graded.

The discrimination is the point. A mechanism you can draw tells you a product; a mechanism you can defend against its lookalike tells you the product — and the defense is what exams test. This is why mixing similar reaction types (SN1/SN2/E1/E2) is such effective practice: it's a desirable difficulty that forces you to select the method, not just execute a memorized one.

The mechanism-discrimination workflow

Group by reaction family, not page order. The lecture's sequence is arbitrary; the mechanism families are the structure. SN1/SN2/E1/E2 share substrates and conditions — the exam's entire job is asking you to tell them apart. Grouping by family builds the discrimination in from the start.

Start the mechanism on a blank page. Electrons first — which arrows push, where the charge moves, what intermediate forms. The reagents are labels; the electron flow is the mechanism. If you can't produce it blank-page, you don't have the reaction, you have a picture of it.

Write the competing product. Every mechanism has a cousin — the other path the same substrate could take. Write it down. Naming the lookalike is what makes the decision visible; without it, "the product is X" is just an assertion.

Name the deciding condition. Why this path and not the cousin? Base strength, solvent, temperature, steric bulk — the condition that tips the mechanism one way is the answer the exam is looking for. "E2 because bulky base and heat" is a gradeable answer; "E2 because it's E2" is a shrug.

Swap a condition and re-predict. Same starting material, one condition changed — weaker base, different solvent, no heat. Predict the new major product. If you can predict the flip, you understand the mechanism; if the swap breaks you, you memorized a picture.

What it looks like in practice

The weak answer is circular — "it's E2 because it's E2" — a label, not a mechanism. The strong answer produces the deciding conditions (bulky base, heat) and explains why the alternative path (SN2) loses: the hindered nucleophile can't attack cleanly. That's the discrimination the exam grades.

Where reagent memory actually goes

Reagent memorization isn't useless — it's mistimed. Once you can run the mechanism, the reagent names become labels on a process you already understand: "bulky base" is what t-BuOK means in E2; "strong nucleophile, polar aprotic" is what NaN3 in DMSO means for SN2. The names attach to understanding and stick.

Memorized first, they're trivia — a list of associations unconnected to any mechanism, which is why they evaporate under exam pressure. The order matters: mechanism first, labels second. The labels are cheap once the mechanism is there; the mechanism is unreachable through the labels.