A wrong answer on the MCAT is not a random wrong fact. It is engineered: built to be picked by someone who read a little too fast, remembered a term a little too fondly, or trusted a claim one notch stronger than the passage. Our tutors see the same construction patterns reused across sections, and Atlas, the AI tutor inside every question on our platform, is built to name the pattern out loud whenever you miss. Here are the five it calls out, and the counter-move for each.
1. Too extreme
The claim you almost want, dialed past what anything supports: always, never, only, proves, eliminates. Science sections use it on mechanisms, CARS uses it on the author's position. It works because the core of the sentence matches your understanding, so the absolutes slide by.
The counter-move: read answer choices with a finger on the quantifiers. When two options say nearly the same thing, the calmer one is usually the credited one, and the loud one is usually this trap.
2. Out of scope
A statement that is true, reasonable, even interesting, and answers nothing the question asked. In a passage-based question it imports material the passage never touched; in an experiment question it reaches past what the data can support. It works because your brain rewards recognition: true feels like right.
The counter-move: ask "could I point to the line or the data column that supports this". If the support lives in your coursework instead of on the screen, it is not the answer.
3. Right words, wrong claim
The passage's own vocabulary, rearranged into a sentence the passage never says. Cause and effect swapped, a comparison inverted, a finding attached to the wrong group. It is the most mechanical of the five, and under time pressure it is brutal, because every word passes the recognition check.
The counter-move: slow down exactly once per question, at the relationship. Who acts on whom, what increases what. The nouns being familiar is not evidence; the verb is where this trap lives.
4. True but irrelevant
The cousin of out-of-scope: an answer that correctly describes something, just not the thing asked. Classic on experiment passages, where a choice restates the setup instead of interpreting the result, and on "which finding would weaken" questions, where it strengthens instead. It works when you hold the topic in mind but let the actual question blur.
The counter-move: re-read the question stem after reading the choices, not just before. Ten seconds of "wait, what was asked" kills this pattern almost completely.
5. The famous-concept bait
A big-name concept from your coursework, placed where it does not belong. Our QBank demo carries a clean example: asked which amino acid sits in a transmembrane region, the options offer famous charged residues, hoping you will recall a name instead of checking hydrophobicity. The concept is real; the fit is wrong.
The counter-move: when an option feels like a reunion with an old friend, that is precisely the moment to check the mechanism. Try it live on the QBank demo and watch the reveal name the trap.
Make the patterns part of your review
Knowing the taxonomy is worth little on its own. The move that changes scores is tallying your own misses against it: after every practice set, one tick per trap. Within a couple of weeks the tally shows your personal pattern, and a trap you can name mid-passage is a trap that has mostly stopped working. That tally habit matters most on CARS, where the traps do nearly all the work; the section's full training loop is in CARS practice.