1. The LSAT is deterministic.

Every Logical Reasoning question is one of 13 types. Every flaw is one of 17 patterns. Every wrong answer is engineered against a small, knowable set of trap families. Reading Comprehension obeys VIEWSTAMP, Viewpoint, Inflections, Evidence, Worldview, Structure, Tone, Argument, Main point. Logic Games are gone, but the moves they trained transfer cleanly to the new conditional-reasoning questions.

This isn't a vibe. It's the canonical taxonomy used by every elite tutor, PowerScore, Manhattan Prep, the LSAT Trainer. We codify it explicitly and teach it directly, instead of hoping you absorb it through 3,000 questions of trial and error.

2. The Algorithm

The fastest way to a higher LSAT score is to stop treating each question as a puzzle and start recognizing the move.

Once you can name the question type in under three seconds, name the conclusion before reading the answer choices, and predict the trap pattern before it appears, your average accuracy and your average pace both jump. That's the lever.

• 13 LR question types, Strengthen, Weaken, Necessary Assumption, Sufficient Assumption, Flaw, Method, Parallel Reasoning, Principle, Resolve, Inference, Most Strongly Supported, Point at Issue, Role.
• 17 flaw families, circular reasoning, ad hominem, equivocation, sufficient/necessary swap, hasty generalization, correlation/causation, etc.
• 10 trap patterns, out-of-scope, half-right, opposite, partial fix, extreme language, reverses the relationship, etc.

3. The Drill Engine

You don't get a generic playlist. Every drill set is generated for you, in the moment, by an engine that watches every attempt:

• 50% of each set targets your weakest question types (sample size ≥ 5 attempts).
• 30% reinforces types you're improving on but haven't mastered.
• 15% samples your mastered types so we can verify mastery hasn't decayed.
• 5% introduces new patterns you haven't seen yet.

The engine respects two recency rules: you don't see the same question within 14 days, and you don't see a question you got wrong within 7 days (so the right answer doesn't bleed into your memory). Difficulty is capped against your rolling accuracy, no death-spirals, no easy-mode coddling.

4. Coaching Signals

After every drill set, your coaching dashboard updates:

• Your three weakest move types, ranked by gap-from-mean accuracy.
• Your most-frequent trap pattern (the one you keep falling for).
• Mastered types, the ones you can stop drilling.
• A natural-language coaching message: "You're falling for half-right traps on Strengthen, they look right because they fix one premise but ignore the conclusion. Slow down on the second sentence of the choice."

5. Score Prediction (Pro+)

On the Pro tier, we surface a real-time predicted score with a confidence interval. The model uses your accuracy by question type, your timing under pressure, and your performance on mock-test conditions, calibrated against actual LSAC scoring distributions. It's directional, not crystal-ball, but it gets you off the "am I ready?" treadmill.

6. The Score Guarantee

If you complete 80% of the curriculum, log at least 30 drill attempts, and take a real LSAT, and your score doesn't improve by at least 5 points, we refund you. The terms are short and on the guarantee page. We're confident because the algorithm + the engine + your effort actually works.

What you don't get from us

We don't sell textbooks. We don't sell "LSAT mindset coaching." We don't have a 12-step funnel that upsells you a $500 retake course. We have one product, three tiers, and a simple promise: stop memorizing tricks, start seeing the test.