Using ChatGPT as your hyper-intelligent LSAT tutor sounds like a dream come true. But is it too good to be true? As it turns out, yes.
While ChatGPT has its uses, as an LSAT study tool, it’s far from perfect. Here are the five worst LSAT study tips I got from ChatGPT, and my best advice about what to do instead.
Worst LSAT Study Tips from ChatGPT
1. The Problem of Outdated Information
I asked ChatGPT to give me a bunch of LSAT study tips. One of the first five it gave was:
“Master Diagramming: Create clear, consistent symbols for rules. Get good at seeing what’s ‘in play’ and what’s locked down.”
I’ll admit, the phrases “in play” and “locked down” threw me for a moment. But then I realized: “It’s talking about logic games!”
This is great advice for logic games. The problem, of course, is that logic games no longer exist on the LSAT.
Alas, this is the problem of outdated information. Large Language Models like ChatGPT are based on years of training data. These tools sometimes confuse what was true in the past with what’s currently true. The LSAT eliminated Logic Games in August 2024, replacing them with a second Logical Reasoning section.
To be fair, diagramming is still of limited importance today. You might find a small handful of diagrammable Logical Reasoning questions, maybe three to six per exam, but it’s no longer the LSAT “survival skill” it once was.
💡 What to do instead: Focus your energy on the sections that actually exist! Master the current LSAT format: Two Logical Reasoning sections and one Reading Comprehension section. Definitely still learn how to diagram a conditional statement for Logical Reasoning purposes, but save the intensive diagramming practice for when you’ve already conquered the fundamentals.
2. The Problem of Agreeableness
Here’s a recreation of several conversations I’ve had with ChatGPT over Logical Reasoning questions:
Me: “Is the answer B or C?”
ChatGPT: “The answer is definitely B. Here’s why: [detailed explanation].”
Me: “Actually, I think it’s C because [weak reasoning I’m not even sure I agree with].”
ChatGPT: “You’re absolutely right! I apologize for the confusion. The answer is C. Here’s the correct reasoning: [completely different explanation].”
In many contexts, this agreeableness is a feature, not a bug. But when you’re studying for the LSAT, you need pushback. You need to be challenged when you’re wrong. ChatGPT’s tendency to agree with you can reinforce incorrect reasoning patterns and give you false confidence.
💡 What to do instead: First off, verify correct answers against official answer keys and explanations from a trusted source. If you disagree with an explanation, argue with a human teacher, not with an AI that might just agree with you to be polite!
Curious about what makes a fabulous LSAT question explanation? Take our free LSAT diagnostic exam, review your results, and explore our detailed, easy-to-understand explanations!
3. The Problem of Fake Practice Questions
When I asked ChatGPT to write me a flaw question, here’s what I got:
“Recent studies show that people who drink green tea daily have lower rates of heart disease. Therefore, green tea prevents heart disease.”
What is the flaw in this reasoning?
(A) The argument assumes correlation implies causation
(B) The argument uses outdated evidence
(C) The argument appeals to false authority
(D) The argument commits a hasty generalization
(E) The argument creates a false dilemma
This isn’t terrible, but it’s not exactly LSAT-quality either. Real LSAT questions are more nuanced, with trickier wrong answers and more sophisticated reasoning patterns. The language is also too straightforward—real LSAT passages are dense and often convoluted.
More importantly, practicing with fake questions can actually hurt your performance. You’ll develop pattern recognition for AI-generated questions rather than actual LSAT questions, potentially leading you astray on test day.
I’ve written more about my misadventures with AI-generated LSAT questions here.
💡 What to do instead: Stick to real LSAT questions from actual prep tests. There are decades of released LSATs available on LawHub. Use them! And remember, be wary of ChatGPT’s explanations. As evidenced by the diagramming error from earlier, it can be prone to detail misfires and hallucinations.
Sign up to get expert tips and exclusive invites to free LSAT classes and law school admissions workshops!
4. The Problem of Precise Details
ChatGPT often fumbles crucial details. Case in point, I asked it to diagram a simple statement: “X unless Y.” This is what ChatGPT said:
🔍 Step-by-step: What does “X unless Y” mean?
Let’s say:
“X unless Y”
This means: “X happens only if Y happens.”
That is:
If Y doesn’t happen, then X doesn’t happen either.
→ In logic:
~Y → ~X
If you’re up on your conditional diagramming, you’ll have noticed that ChatGPT got this wrong. In LSAT-speak, “unless” translates to “if not.” So “X unless Y” means “X if not Y,” and would be diagrammed as:
~Y → X
Yes, ChatGPT was close, but as my old math teacher, Mr. McClellan, used to say, “close only counts in horseshoes and hand grenades.”
💡 What to do instead: Look, I’ve seen this before. ChatGPT gets the gist but gets a wire crossed on the details. Don’t trust a chatbot for diagramming logical statements. Double-check any conditional logic explanations against trusted LSAT prep materials!
5. The Problem of Context Blindness and Generic Advice
This was number two on ChatGPT’s “Top 5 LSAT practice tips”:
“Blind-review every practice set. After finishing a section, flag any answer you’re not 100% certain about, then re-do those questions untimed without looking at the key. Compare your original and revised answers to see where your reasoning slipped. That side-by-side analysis is where the real learning happens.”
This sounds reasonable in theory, but it’s problematic advice for most students. If you’re like most students starting out, you could be missing half or more of the questions in your initial practice sets. Blind reviewing every last question like this becomes a recipe for frustration. The time you spend blind reviewing for the umpteenth time could be better spent doing five to ten new questions and mastering the strategies and test patterns.
ChatGPT gave this advice because it is context blind. It doesn’t know if you’re a high-scoring or low-scoring student. It doesn’t know if you have the patience and wherewithal to blind review every practice set or if you’ve missed 14 straight questions and are at your wits’ end. Because ChatGPT is context blind, it often defaults to generic blanket advice that sounds good for the fictional “average” student, but makes sense only in certain cases.
I don’t even think “blind review everything” is good advice for anyone. Don’t get me wrong — if you’re a 170+ scorer striving for perfection, scrutinizing every detail might serve you well. But if you’re lucky enough to be at this point, you’re only missing a small handful of questions over an entire exam. My advice: Don’t let perfect be the enemy of the good. We don’t need to blind review every missed question, and we don’t want perfectionism to compromise our ability to complete a healthy volume of practice questions.
💡 What to do instead: Tailor your review strategy to your current skill level. If you’re scoring below 160, the priority is practice. Focus on volume and pattern recognition among practice questions. Save intensive blind review for practice exams, or for the real “head scratcher” questions.
The Bottom Line
Look, I’m not telling you to avoid ChatGPT entirely. It has its uses. For example, it can define basic concepts when you’re stuck.
But as an LSAT study tool, it’s more of a liability than an asset. It gives outdated advice, fumbles crucial details, ignores your individual needs, creates subpar practice questions, and worst of all, can reinforce your mistakes instead of correcting them.
The hard honest truth is that there’s no shortcut to LSAT success, and no AI chatbot can replace good old-fashioned practice and review. So stick to the fundamentals, use real practice tests, and get feedback from sources that will actually challenge you when you’re wrong.
And if you need help reaching your full LSAT potential, we’re here to help! Blueprint LSAT students increase their LSAT scores by 15 points on average. Whether you want the flexibility of a Self-Paced Course, prefer to navigate the LSAT with instructors in a Live Course or 170+ Course, or even private LSAT tutoring, we have the study method that fits your learning style.
Get access to tons of LSAT prep resources when you create a free Blueprint LSAT account!
