The LSAT Decision (Not the LSAT)
Nobody loses points for taking the test at the wrong time. They lose schools.
From the coaching file. Derek, 3.70, found me in April, which is the cruelest month to find me. He’d taken the LSAT once, in January, scored a 158 against a practice average of 161, declared himself “not really a test person,” and applied to 7 schools in March with a statement he’d written the same week. One admit, off the waitlist, in July. He went. Sticker price. Here’s the part that took us 20 minutes to reconstruct. He’d never run it as a single calculation. His January test was his first and only take, booked 6 weeks out, after a fall spent “kind of studying.” A September take with a real plan behind it, followed by October applications instead of March ones, was sitting right there in his calendar the whole time. Nobody had ever told him the cycle has a geometry. He’d treated “when” as logistics. When is strategy. Composite student, and I’d guess 20,000 real ones every cycle.
The last chapter established what the LSAT is in your file: the live number. This chapter is about the 4 decisions that surround it, none of which is “how do I get better at logic.” When to take it first. Whether and when to take it again. How to know when it’s done. And what to do in the 48 hours after a score lands. That’s when most of the unforced errors happen. If you want the number itself raised, that’s my other book, Anyone Can Beat the LSAT, and everything in it routes your actual practice to LawHub and the official PrepTests, the only test material worth your time. This chapter is the part that book leaves to this one: the score as a piece of a case, on a calendar, with money attached.
The last useful LSAT
Write this definition into the case file, because it runs every timing decision you’ll make.
Your last useful LSAT is the latest administration whose score arrives in time for your applications to be complete early in your target schools’ review season, not merely before their deadlines.
The second half of that sentence is where the money hides. Deadlines at most schools run deep into spring, but the deadline is a fact about the form. Review is rolling at most schools. Files get read roughly as they complete, all season, while seats and scholarship budgets get spent down the same way. (Chapter 1 taught you why; chapter 11 puts dates on it.) A January score followed by February applications is “on time” like arriving at a wedding during the toasts is on time. The doors were open. The good outcomes were largely distributed. Derek’s March applications didn’t fail because Derek was weak. They failed because they showed up to a mostly-eaten buffet, and then the waitlist admit in July, the one kind of admit that almost never comes with money, finished the job.
So run the geometry backward, on real dates, starting from your enrollment month. Applications complete by mid-fall the year before. Your last useful score, then, usually comes from a late-summer or early-fall administration. The take before that, your safety margin, lands in spring or early summer. And prep starts months earlier still. Notice what just happened: “I’ll figure out the LSAT this year” became 3 specific registrations, and the one I want circled is the buffer take. Schedule 2 administrations on purpose, the real one and the spare, and the spare stops being an emotional emergency and becomes a calendar line. If the first take delivers, cancel the spare and have a nice dinner instead. (Within the current limits, 5 takes in the reportable window, 7 lifetime, a planned spare is cheap. The thing that’s expensive is needing one you never booked.)
Abstract geometry is easy to nod at and easy to never do, so here’s the whole thing worked once, on real dates, for an invented reader. Jess wants to start law school in fall 2028. (Months below are the traditional shape of the testing year; specific dates move, so build yours from the current LSAC calendar, not from my snapshot.) Work backward with her. Applications complete and submitted: October 2027. So her last useful LSAT: the August or September 2027 administration, scores back with weeks to spare. That makes her real first take June 2027, with August already booked as the spare. Her prep, on a 4-to-6-month plan with a job in the way, starts January 2027. And the decision she has to make mindfully lands this December, 2026, when “I’ll deal with the LSAT next year” is technically true and quietly fatal. Written out as a list in her case file it’s 5 lines. The 5 lines are the difference between Jess and Derek, and notice the punchline: the whole plan got decided 22 months before a single class starts, by a person spending 20 minutes with a calendar. That’s the entire skill. It just has to happen on purpose, in writing, before the months get spent by default.
One more calendar fact, new as of this writing and easy to trip on. Beginning with the August 2026 administration, the LSAT is back to in-person testing at Prometric centers for nearly everyone, with narrow exceptions: certain medical accommodations that can’t be delivered at a center, and documented cases where the nearest center is more than 180 miles or 3 hours away. The bedroom test is over. For you this means seats at convenient centers and dates are now a real constraint, registration is worth doing early, and your practice conditions should look like a testing center, not your kitchen. It also means test day now has travel attached, so plan the morning like a court date: route, parking, buffer. Rules and exceptions drift; confirm current details at lsac.org rather than trusting any book’s snapshot, including mine.
If you’re starting from zero on prep itself and want the short version of my method while you decide whether to buy the long one, it’s free: unpluggedprep.com/cheatsheet.
The retake decision
Chapter 4 gave you the asymmetry: schools see your history, but the ABA counts only your highest, so the published-median machine treats your worst take as a rough draft. What that buys you is a retake market where the downside is mostly capped and the upside is not. What it doesn’t buy you is a free lunch, because a retake spends months, money, and one of your limited takes. So here’s the conditional verdict, all 3 conditions, and I mean all 3.
Retake if something about your preparation will actually be different (a method, a coach, 200 more hours, my book, anybody’s book, but a change, because rerunning the same prep and expecting new output is the LSAT version of reordering the meal that made you sick). And if the calendar has a slot at or before your last useful LSAT, because a point gained at the cost of a cycle’s timing is regularly a net loss, see Derek. And if your practice average on official tests under real conditions sits at least 2 points above your official score, because that gap is evidence the score underreports you, and without it you’re not retaking, you’re gambling with a fee attached.
All 3 true: retake, and do it like you mean it. 2 true: fixable, fix the third first. 1 or none: your energy belongs in the rest of the case now.
What do 2 or 3 points actually buy? The mechanism you already own: every school where the gain carries you across the median you were below flips your role from cost to lift, which moves you across piles in chapter 1’s room and across price brackets in chapter 13’s negotiation. The full arithmetic, score band by score band, dollars attached, lives in my LSAT book and I’m not re-deriving it here. The 1-sentence version for this book: across most of the scale, no document in your file can do for your outcomes what crossing 1 median does. Be suspicious of anyone who says the essays will make up for skipping it. Including the part of your own brain that’s tired.
When is the score done?
Done is a decision, not a feeling, and certainly not a verdict on your ceiling. The score is done when the conditions above stop being true: practice and official have converged, or the calendar has closed, or the next point now costs more than it’s worth to your actual list. That last clause matters and nobody applies it: a reader aimed at schools with 152 medians does not need the weeks it would take to push 158 to 161, and “I just want to see” is not a reason to spend a take; a take is only data if something changed. Declare the score done in writing, in the case file, with the date and the reason, and here’s what you get for that 2 minutes of ceremony: the entire LSAT tab of your brain closes, and Part 3 gets a writer instead of a hostage.
And if you’re reading this with deadlines 6 weeks out and no score at all: the introduction already told you, and now you’ve seen the machinery behind it. A rushed take feeding late applications buys the Derek outcome at the Derek price. Waiting a cycle isn’t falling behind. It’s declining to pay full price for the degraded version of the same product.
Cancel, keep, and the 48 hours after
Last decision pair, and then a protocol, because score week is when smart people do their dumbest things.
Keep beats cancel in almost every case. A delivered score is data and a floor; schools work off your highest anyway; and a cancel still spends one of your limited takes, a detail people learn at the worst time. The honest exceptions are catastrophic: you were ill, the room fell apart, something real derailed the administration. “It felt bad” is not a derailment; feel and result correlate worse than you think. If you use Score Preview to see the number before deciding, fine, but decide by the same rule. (Preview windows and fees change; lsac.org has current terms.)
Watch the rule sort 2 cases, because the difference is the whole lesson. Case 1: Mina’s practice average is 162. Test morning, her section 2 felt like wading through glue, and she walks out convinced she bombed. She wants to cancel. The rule says keep: “felt like glue” is a feeling, her history says 160ish even on a bad day, a 158 on the record costs her nothing at highest-score schools, and the data tells her retake plan exactly where the leak is. Case 2: Owen woke up with a fever of 102, took the test anyway because he’d paid for it, and spent section 3 deciding whether to leave. That’s not a feeling; that’s a derailment with a thermometer reading. Cancel defensibly, log why, and book the spare. The line between them isn’t how bad it felt. It’s whether something happened that you can write down in 1 factual sentence that isn’t about your feelings. Keep that test. This book will hand it back to you in chapters 6 and 10, wearing different names.
Then the score lands, and whatever it is, you run the Score-Day Protocol. For 48 hours: no forums, no list surgery, no announcements, no retake registration, nothing irreversible. The score you got an hour ago is loudest right now and you make property-line decisions about it while it’s shouting. Instead, 1 page in the case file: rewrite your chapter 4 position table with the new number, above, split, or below, school by school. That’s it. That’s the whole assignment. On hour 49, with positions on paper instead of feelings in your chest, make the 1 decision this chapter trained you for: done, or one more, by the 3 conditions. Either answer can be right. The protocol exists so the answer comes from the table and not from the shouting.
Do this now: 20 minutes
Build your version of Jess’s 5 lines, tonight, in the case file: your enrollment target, your application-complete month, your last useful LSAT, your real take and your spare, your prep start date. Pull the actual administration dates from lsac.org while you’re there, and register for anything whose registration window is open. 20 minutes. If you already hold a score, run the retake verdict instead: the 3 conditions, in writing, and either “done, because…” with today’s date, or a plan with a test date on it. Either way, the LSAT tab in your brain gets to close tonight.
Evidence secured, either way. Now we find out what it’s evidence of. Part 3 is the center of this book and of your file: the 1 sentence everything else exists to prove.
