Understanding the Engine
lsatly uses an Item Response Theory (IRT) engine to track your skill level and adaptively serve you questions that match your exact ability.
The Rating System
Your skill is measured using an Elo rating, similar to chess. You start at 1200 Elo. When you answer a question correctly, your rating goes up. When you answer incorrectly, it goes down. The amount it changes depends on the difficulty of the question.
Your rating uses Glicko-2, which tracks both ability and confidence. The LSAT estimate is a projection from your rating across the difficulty distribution of our question bank, not a guaranteed official score. As you answer more questions, the confidence range narrows.
Skill Bands & Estimated LSAT Range
These are rounded illustrative ranges from the current conversion model; neighboring bands may meet at a boundary score.
Elite command of logical structures and difficult argumentation.
Consistently handles subtle assumptions, flaws, and trap answers.
Strong grasp of conditional logic and argument structure.
Solid fundamentals with room to sharpen timing and consistency.
Developing core reasoning skills and learning to avoid common traps.
Building foundational reading comprehension and logical mapping skills.
Percentile data sourced from the official LSAC 2022–2025 three-year rolling dataset.
Join the lsatly community
Share feedback, report issues, and talk through Logical Reasoning with other beta users.
Join the Discord →Legal & Disclaimers
LSAT® is a registered trademark of the Law School Admission Council, Inc. (LSAC), which is not affiliated with, and does not endorse, this product or service. All platform content is provided for educational purposes only. lsatly makes no warranties as to the accuracy or completeness of any content.
View Full Legal & Privacy Policy →