← Home

LSAT Cram Sheet

Logical Reasoning Question Types

Assumption (Necessary): find what MUST be true for argument to work; negate it — if argument collapses, it's necessary
Assumption (Sufficient): find what GUARANTEES the conclusion; use if-then logic
Strengthen: find answer that makes conclusion more likely; adds support to premise-conclusion link
Weaken: find answer that makes conclusion less likely; attacks the assumption or adds a counter-case
Flaw: identify the logical error in the argument (circular, ad hominem, false analogy, etc.)
Inference/Must Be True: find what MUST follow from the stimulus; cannot be only likely — must be certain
Main Point: identify the author's bottom-line conclusion; look for conclusion indicators (therefore, thus, so)
Method of Reasoning: describe HOW the author argues, not what they argue
Parallel Reasoning: find answer with same logical structure AND same validity as stimulus argument
Principle: find the general rule that justifies or matches the specific situation
Point at Issue: find what two speakers directly disagree about — both must have explicit views
Role of Statement: explain what logical function a specific sentence plays in the argument

Linear games: elements placed in sequence (1-7); draw a row and place or restrict elements
Grouping games: elements assigned to groups; track in/out or group membership
Hybrid games: combine ordering + grouping; requires dual-tracking diagram
If A then B → contrapositive: If NOT B then NOT A; always write the contrapositive immediately
Biconditional: A ↔ B means A→B AND B→A; they always go together or not at all
Chain rule: A→B and B→C therefore A→C; combine conditionals into chains
NOT-law: when a rule eliminates a position for an element, mark it with an X in the diagram
Local questions: add new condition to diagram; use fresh diagram, don't modify master
Global questions: ask what must/could/cannot be true with only base rules; work from deductions
Make inferences first: before answering questions, exhaust all forced deductions from rules
Most restricted element: find which element has most constraints; anchor diagram around it

Read for structure, not content: identify author's point, tone, and how paragraphs relate
Main Point questions: should match tone and scope of ENTIRE passage; not too narrow or broad
Inference questions: answer must be proven by passage; if not directly stated, it must logically follow
Author's Attitude: look for evaluative language (unfortunately, importantly, surprisingly)
Primary Purpose: what is the author TRYING TO DO? (argue, describe, evaluate, compare)
Comparative Reading: identify what both passages share in topic but differ on in perspective
Detail questions: go back to passage; eliminate answers that distort or add information
Passage structures: argument (author defends thesis), description (neutral survey), analysis (breaks down phenomenon)
Trap: answer that sounds good but goes slightly beyond what passage supports = wrong; must be proven

All A are B: A→B; contrapositive: NOT B → NOT A
No A are B: A→NOT B; contrapositive: B→NOT A
Some A are B: at least one A is also B; cannot be contrapositived
Most A are B: more than half of A are also B; does NOT guarantee any specific A is B
Only A are B = B→A (only introduces sufficient condition on left side)
Unless = if not; 'A unless B' = NOT B → A (contrapositive: NOT A → B)
Negating AND: NOT(A AND B) = NOT A OR NOT B (De Morgan's Law)
Negating OR: NOT(A OR B) = NOT A AND NOT B (De Morgan's Law)
Sufficient condition: if trigger occurs, result guaranteed; NOT sufficient if trigger absent
Necessary condition: must be present for result; result impossible without it
Combining quantifiers: All A→B + Some C are A = Some C are B (valid transitive inference)

Ad Hominem: attacks the person making the argument instead of the argument itself
Circular Reasoning: conclusion restates a premise without additional support
False Dichotomy: assumes only two options when more exist
Correlation vs. Causation: assumes that because A and B correlate, A causes B
Composition: assumes what's true of parts is true of the whole
Division: assumes what's true of the whole is true of each part
Equivocation: uses the same word with two different meanings
Hasty Generalization: draws broad conclusion from insufficient sample
Straw Man: misrepresents opponent's argument to make it easier to attack
Appeal to Authority: uses expert opinion as definitive proof in empirical question
Unrepresentative Sample: sample used doesn't reflect the larger population
Scope Shift: conclusion is about different population or concept than the premises

Format: 175 minutes total; 4 scored sections + unscored experimental section
Sections: 2 Logical Reasoning (52 questions) + 1 Analytical Reasoning (23 q) + 1 Reading Comprehension (27 q)
Scoring: 120-180 scale; raw score (# correct) converted to scaled score
Raw score of 100+ (out of 101-102) needed for 180; each wrong answer = 0 (no penalty)
Score of 170+ = 97th percentile; 160 = 80th percentile; 151 = median
Top schools: Yale ~174, Harvard ~173, Columbia ~172, Chicago ~171 median
No guessing penalty: always answer every question; guess if out of time
LG (Logic Games) most trainable: students gain most points here with practice
Time per section: ~35 minutes; approx 1:20 per question in LR, 8:45 per game in LG
LSAT-Flex (remote): same content, slightly different format; adaptive sections being rolled out

Identify conclusion FIRST in LR: read for 'therefore,' 'thus,' 'so,' 'hence,' 'consequently'
Premise indicators: 'because,' 'since,' 'given that,' 'as evidenced by'
Pre-phrase answers: before looking at choices, predict what the correct answer should say
Process of elimination: eliminate answers that are too strong, too broad, or distort the stimulus
Avoid extreme language: 'always,' 'never,' 'impossible,' 'proves' rarely correct in inference questions
Harder LR questions appear later in section; don't skip earlier questions to race ahead
In LG: build master diagram with all base rules before answering any questions
LG 'Could Be True' questions: one wrong answer per eliminated scenario; test each answer choice
LG 'Must Be True' questions: answer must hold in ALL valid scenarios; test with try-to-violate method
RC comparative: read Passage A fully; read Passage B with A in mind; note agreement/disagreement

Diagnostic: take full timed test first; identify which section has highest error rate
LG mastery path: understand game types → learn diagramming → drill by game type → timed sets
LR mastery path: identify question type → find conclusion → find assumption → evaluate answers
RC mastery path: low-res summary after each paragraph → predict main point before reading questions
Error log: for every wrong answer, record: question type, what you chose, what's right, WHY
Don't review correct answers: focus time on understanding why wrong answers were wrong
Timed practice: always do at least one timed section per day once past fundamentals
Score plateau? Drill your weakest question type exclusively for 2 weeks
PrepTest recommendation: use official LSAC PrepTests 52-92 for most realistic practice
2-3 month timeline for 10+ point improvement with 15-20 hours/week of focused practice