Logical Reasoning I: The Skeleton Files

The test owns the costumes. The bones can be yours.

Here’s the section where your scholarship lives.

2 of your 3 scored sections are Logical Reasoning: around 50 of the roughly 75 to 80 questions that decide your number. The 4 questions worth $150,000 from page 1 of this book? Odds are they’re in here. And the weight comes with good news attached. LR is where skeletons (the recycled reasoning bones from chapter 4) repeat fastest, which makes it the most review-responsive real estate on the test. 50 questions a test, drawn from a short menu of bones.

2/3 of your score is 1 skill wearing 50 costumes.

So before we touch a single argument, point your hours at the math.

•••

The 2/3 Rule: Logical Reasoning is 2/3 of your scored test, so by default it gets about 2/3 of your study time, until your own miss ledger argues otherwise. Re-run the math monthly.

•••

The second half of that sentence matters as much as the first, so let me show you both halves working.

Why the default exists: with no data yet, you bet the base rate, and the base rate says 2 of every 3 points live in LR. There’s also a quiet asymmetry. LR training pays a commission into Reading Comprehension, because a passage is basically arguments stacked 4 high (chapter 9’s whole opening move). RC training pays almost nothing back into LR. When in doubt, fund the skill that subsidizes the other one.

Why the override exists: hours follow recoverable points, and your scoresheet outranks the base rate. Pull your last 2 practice tests and count misses by section. Maya drops 6 and 5 across her 2 LR sections and 5 in RC: 11 of her 16 recoverable points live in LR, the default holds, carry on. Noah drops 2 and 2 in LR and 11 in RC: 11 of his 15 live in RC, and the default would have him polishing his best skill while his cheapest points sit untouched. He flips RC-heavy for a month, then re-counts. The rule, complete: a default, a ledger, a monthly recount.

One wrinkle, because chapter 9 will tell you RC is the slowest section to move: recoverable points are how many you can win back times how fast they convert, and LR converts faster. Noah flips anyway; 11 versus 4 swamps the speed difference. At 7 and 7, the tiebreak goes to LR every time.

Now let’s go earn the points.

The anatomy, in 90 seconds

Every LR question is a short paragraph (the stimulus) followed by a question (the stem) and 5 choices. The stimulus is 1 of 2 things, and telling them apart is your first move every single time.

An argument is someone trying to sell you a conclusion using evidence. A fact set is just statements, nobody selling anything. The test: ask “is any sentence here leaning on the others for support?” If yes, the leaner is the conclusion and everything propping it up is evidence. If no, it’s a fact set, and you’ve got a different job (we’ll get there).

You already know what the gap is. Chapter 6 made you find one twice: the gap is everything the bridge between evidence and conclusion needs that isn’t on the page. What’s new is this: the size of the gap is set by the boldness of the conclusion. Watch 1 claim get bolder while the evidence stands still:

Dev wasn’t at the 9 a.m. meeting. A fact. Nothing to prove. Dev probably overslept. Now there’s a bridge to build, but a short one. “Probably” is a humble word; modest evidence might carry it.

Dev must have overslept. Same evidence, massive new gap. “Must” doesn’t just need oversleeping to be likely. It needs every other explanation (traffic, a sick kid, a calendar glitch, a better meeting) to be impossible.

One word changed. The proof burden tripled. The bolder the conclusion, the wider the gap, and the easier the argument is to attack. Top scorers read conclusions with their thumb on words like must, all, never, only, will, because those words are the argument writing checks. Your job is to ask whether the evidence can cash them.

The 2 jobs

Everything the LR section asks of you is 1 of 2 jobs.

Gap mode. The stimulus is an argument, and your job orbits its gap: name it, widen it, narrow it, or fill it. This is most of the section.

Build mode. The stimulus is a fact set (or an argument you’re told to treat as gospel), and your job is to find what those facts add up to. No attacking. The facts are bricks; you’re checking which choice is built from them.

2 jobs. That’s the section. Everything else is camera angles, and the next chapter proves it in full.

How to read a file

Chapter 4 promised the test recycles a short menu of bones. This chapter is the menu, written up the way a detective writes up a repeat offender: what the move is, an easy costume worked in front of you, a hard costume worked in front of you, the traps it likes to set, and a 4-question micro-drill so the bone moves off the page and into you before the file closes. 6 files, 5 from gap mode, 1 from build mode.

Don’t read them like chapters. Work them. Pen out, journal open to the LR page, log each handle as you go. The micro-drills are call-the-gap style: no answer choices, because the skill underneath everything is seeing the bone before anyone shows you options. Write your 1-line answer before reading the key. You know the law by now.

•••

File 1: after = because

The bone. This happened after that, so that caused this. The oldest move in the building, your acquaintance since chapter 4’s 9 invoices and chapter 6’s park lights, and the single most recycled skeleton on the test, because causal language is cheap to costume: any 2 events in time can wear it. Its weak points never change. Another suspect (something else caused it). The arrow backwards (the “effect” actually drove the “cause”). Or pure coincidence.

Easy costume, worked. The month after Greenfield installed security cameras in its parking garage, break-ins there fell by a third. The cameras are clearly working. The rebuild: evidence, break-ins dropped after installation; conclusion, the cameras caused the drop. Now the suspect lineup, which you should run on every causal claim: what else changed that month? Weather, a police patrol shift, the arrest of 1 busy thief, winter ending. Could the arrow reverse? Not here (break-ins can’t install cameras), which is worth pausing on: not every attack applies to every costume, and knowing which one fits is half the skill. Could it be coincidence? One month of data says maybe. Three live attacks, 0 of them addressed. That’s the bone in daylight.

Hard costume, worked. A 10-year study found that adults who volunteer regularly report significantly fewer health problems than adults who do not. Volunteering, the researchers concluded, protects health. Same bone, now wearing a lab coat, and the lab coat is the disguise: “study,” “10-year,” “significantly” all dress the leap in authority without shrinking it an inch. Run the lineup. Another suspect: people who volunteer tend to have more time, money, and social connection, all of which protect health on their own. The arrow backwards, and on this costume it’s the sharper blade: healthy people are the ones with the energy to volunteer. The health may be causing the volunteering. Prestige vocabulary doesn’t close gaps, it hides them. And the test knows you were raised to salute the word “study.”

Its signature traps. When this bone is the flaw, the right answer says some version of another cause was possible, cause and effect may be reversed, or the correlation may be coincidental. The wrong answers that hunt you: choices attacking the size of the effect (“the drop was only one third”) or the wisdom of the plan, both of which leave the logic untouched. You met that second one at Hartwell Park.

Journal handle: after = because. Forecast: any stimulus where the conclusion’s verb is caused, produced, led to, is responsible for, is working, fed by evidence that’s just a sequence or a correlation.

Micro-drill, file 1. One line each, your words, pen.

1. The cafe’s afternoon sales rose the week after it bought an espresso machine. The machine is paying for itself.
2. Runners who stretch before training get injured less often than runners who don’t. Stretching prevents injuries.
3. Vandalism reports dropped in the 2 months after the town enacted a teen curfew. The curfew is working.
4. Patients who keep their follow-up appointments recover faster than patients who skip them. Follow-up appointments speed recovery.

Key. 1: suspect lineup; what else happened that week (weather, a festival, a competitor closing)? One week is coincidence’s home turf. 2: the arrow, plus a hidden suspect; cautious, uninjured runners may be the kind who stretch, and the injury-prone may have quit. 3: suspects in bulk; school starting, lighting, patrols, and 2 months is a short ledger. 4: selection; patients healthy and organized enough to keep appointments differ from those who skip, before any appointment does anything.

•••

File 2: some = all

The bone. A handful becomes everybody. Evidence about a sample, a conclusion about a population. Every survey, poll, and study on this test deserves 2 questions, and the second one outranks the first: how many, and how chosen? Size cures noise. Nothing cures a sample that picked itself, which is the version this bone wears when it’s hunting good students; you watched it at the garden center in chapter 6.

Easy costume, worked. I asked 3 coworkers about the new schedule and all 3 hate it, so the whole office is against it. Rebuild: 3 opinions in, a whole office out. How many: 3, which is nothing. How chosen: you picked them, which is worse than nothing, because you asked the people you talk to, who are the people most like you. Both questions fail and the second failure would survive even if you’d asked 30.

Hard costume, worked. Of the 1,200 readers who emailed the newspaper about its redesign, 85 percent hated it. The redesign has clearly failed with readers. The disguise here is the number: 1,200 feels like a real sample, and 85 percent feels like a verdict. Run the senior question. How were these 1,200 chosen? They chose themselves, and the people moved to email a newspaper are overwhelmingly the angry ones; satisfied readers just keep reading. A self-selected sample of any size measures the intensity of the annoyed, not the distribution of opinion. n cures noise. n does not cure skew. Write that one down; it kills a whole family of costumes.

Its signature traps. When this bone is the flaw, the right answer names the unrepresentative sample. The classic bait is its near-twin: a choice accusing the argument of generalizing from only one case when the evidence already covered the field, a crime that didn’t happen. The next chapter’s bakery question runs that bait, so consider yourself warned.

Journal handle: some = all. Forecast: any stimulus with surveyed, polled, sampled, respondents, the customers we asked, and any conclusion with generally, typically, customers as a whole, the public.

Micro-drill, file 2.

1. In January, the gym surveyed members who had joined that month, and 88 percent rated the facility excellent. Member satisfaction is clearly high.
2. Every house on my block has a dog. This is a dog town.
3. All 6 of the city’s award-winning restaurants are downtown. Downtown plainly has the city’s best food.
4. The product’s page shows hundreds of 5-star reviews, so buyers love it.

Key. 1: how chosen; brand-new January members are the honeymoon population, sampled at peak enthusiasm. 2: how many and how chosen; 1 block is neither random nor big. 3: selection upstream; awards cluster where critics dine, so the sample of “award-winning” was picked by geography before quality got a vote. 4: self-selection plus curation; reviewers volunteer, and pages can sort. n does not cure skew.

•••

File 3: the word swap

The bone. The evidence proves something about X. The conclusion quietly talks about X-prime, a close cousin. Best-selling becomes best.Most talked about becomes finest.Fewer complaints filed becomes fewer problems occurring. This is the sneakiest skeleton on the menu, because both sentences sound like they’re about the same thing, and it’s the reason chapter 6 made you rebuild arguments in your own words: the swap survives skimming and dies in a rewrite.

Easy costume, worked. Harbor House sells more copies than any novel this year, so it’s the year’s best novel. Rebuild in 2 columns: evidence is about copies moving; conclusion is about quality. Underline the noun that walked in (sales) and the noun that walked out (best). They’re cousins, not twins; marketing budgets, movie tie-ins, and airport shelf space all move copies without touching quality. The rewrite test takes 10 seconds and the swap cannot survive it.

Hard costume, worked. Hospital readmission rates fell 12 percent in the year after the new discharge protocol was adopted. Patients are clearly recovering better. Two bones are in this one, and the louder one (after = because, tally it) isn’t the one I want. Look at the nouns. Evidence: readmissions, a measurement of who comes back. Conclusion: recovery, a fact about bodies. Those can come apart in every direction: readmission criteria may have tightened, patients may be going to other hospitals, or going home sicker and staying there. A measurement and the thing it measures are 2 different nouns, and the test’s favorite hard swaps live right there: complaints filed and harassment occurring, reported incidents and actual incidents, scores and learning. The bureaucratic register is the costume.

Its signature traps. When the swap is the flaw, the right answer reads like presumes that X establishes Y or takes high sales to indicate high quality, the gap named as an equation. The bait is usually a scope crime that didn’t happen (“concludes about all from some”) sitting right next door, hoping your fresh file 2 tally is itching.

Journal handle: the word swap. Forecast: superlatives and stand-ins. Any conclusion whose key noun you cannot find, verbatim or in a true synonym, anywhere in the evidence.

Micro-drill, file 3.

1. Glide is the most downloaded app in its category, so it’s the most useful.
2. Calls to the city’s noise hotline fell by half this year. The city is getting quieter.
3. Brightside is the state’s fastest-growing supermarket chain, so it will soon be the biggest.
4. After the workshop, 95 percent of students said they understood the material. The workshop clearly taught it well.

Key. 1: downloaded and useful are cousins; free apps and ad campaigns drive downloads. 2: calls and noise; the hotline may have gotten harder to find, or callers may have given up. 3: growth rate and size; the fastest-growing chain can be the smallest one, growing from 2 stores to 4. 4: saying you understand and understanding; the swap plus a survey, 2 files in 1 costume.

•••

File 4: the one-way door

The bone. Rules and conditionals. If it rains, the deck gets wet is a door that swings 1 way: rain in, wet deck out. There are exactly 2 illegal walks, and the test sells both daily. Backwards: the deck’s wet, so it rained (the sprinkler says hello). Negated: no rain, so the deck’s dry (the sprinkler, again). The door image carries you a long way; the full machinery (translations, chains, the 1 legal reverse) lives in the next chapter, where build mode needs it. This file’s job is smaller and comes first: spotting an illegal walk inside an ordinary-looking argument.

Easy costume, worked. Gym members park free in the members’ lot. Lena parked free in the members’ lot on Tuesday, so she must have joined the gym. The rule: member, therefore free parking. The walk: free parking, therefore member. Backwards through the door. Open-house days, broken gates, and generous attendants all park free without joining. Note the tell, because it never changes: the conclusion sits on the if side of the rule, and the evidence sits on the then side. When the argument travels then-to-if, it walked backwards.

Hard costume, worked. The festival will be cancelled unless the permits arrive by Friday. The permits arrived Thursday, so the festival is on. It feels airtight. Translate the rule anyway: unless X, Y means if not X, then Y. If the permits don’t arrive by Friday, cancellation. The permits arrived, which removes that 1 trigger for cancellation, and proves nothing more. The rule never said permits, therefore festival. Rain, funding, a headliner with the flu: the festival has a dozen other ways to die that this rule is silent about. That’s the negated walk wearing the word “unless,” and “unless” is where good students get robbed, which is why the next chapter gives it a full translation kit.

Its signature traps. When the illegal walk is the flaw, the right answer says takes a condition that is sufficient for an outcome to be necessary for it (or the reverse), which is the door in formal dress. The bait: choices fussing about whether the rule itself is true. In these arguments the rule is a gift; the crime is the walk.

Journal handle: walked the door backwards (or negated it). Forecast: if, only if, unless, requires, guarantees, whenever. Any of those words in the stimulus puts your hand on the doorframe.

Micro-drill, file 4.

1. Everyone who passed the lifeguard exam can swim 500 meters. Ria can swim 500 meters, so Ria passed the exam.
2. If today were a holiday, the bank would be closed. Today isn’t a holiday, so the bank is open.
3. Members receive a door code, and anyone with a door code can enter before 9. Lena entered at 8:30, so she’s a member.
4. Applicants are admitted only if their file is complete. Theo’s file is complete, so Theo will be admitted.

Key. 1: backwards; swimmers aren’t all lifeguards. 2: negated; banks close for renovations, weekends, and a hundred unholidays. 3: backwards through a 2-link chain; the code can leak, and the rule only ever promised member-to-code-to-entry, not the return trip. 4: “only if” marks a requirement, not a guarantee; a complete file is the ticket to the lottery, not the prize. (If items 3 and 4 made your eyes narrow, good. The next chapter’s machinery section is built for you.)

•••

File 5: percent vs. count

The bone. Shares and totals impersonating each other. A percentage is a fraction with 2 moving parts, and arguments that hop between a share and a raw number have quietly assumed the denominator sat still. It rarely did. Three ways a share can rise: the top grew, the bottom shrank, or both. The test hides the second one and bets you’ll never look.

Easy costume, worked.Bikes made up a bigger share of rush-hour traffic this year than last, so there must be more bikes on the road. Rebuild with the fraction visible: bikes over total traffic went up. More bikes is 1 explanation. Fewer cars is another, and it requires not a single additional bicycle. Gas prices spike, a highway closes, remote work spreads: the share climbs while the bike count sleeps. Always ask where the denominator went.

Hard costume, worked. A decade ago, parks received 4 percent of the city budget; today they receive 8 percent. The city now spends far more on parks than it did then. The disguise is direction: the share doubled, which sounds like growth twice over. But 8 percent of a shrunken budget can be fewer dollars than 4 percent of a fat one. Recessions cut budgets in half all the time. The conclusion needs a count; the evidence offers a share; the swap between them is the entire argument. And note the bone runs both directions: “spending rose, so its share rose” fails the same way, with the denominator growing instead of shrinking.

Its signature traps. When this bone is the flaw, the right answer says some version of fails to consider that the total may have changed. The bait: choices quibbling with the exact figures or their source. The numbers are fine. The fraction is the crime scene.

Journal handle: percent vs. count. Forecast:share, proportion, rate, percentage in the evidence with more, fewer, increased, declined in the conclusion, or the reverse. The moment an argument changes units mid-stride, the gap is open.

Micro-drill, file 5.

1. The unemployment rate fell this quarter, so more people are working.
2. Dana has the team’s best free-throw percentage, so she scores the most free throws.
3. Trucks were involved in a smaller share of highway crashes this year. Trucks are getting safer.
4. Average donations per member doubled this year, so the charity raised more money than last year.

Key. 1: the rate’s denominator is the labor force; people who stop looking leave the denominator, and the rate falls with nobody hired. 2: percentage and volume; a 95 percent shooter who attempts 4 a season scores fewer than a 70 percent shooter who attempts 200. 3: share of crashes, plus a stowaway word swap; fewer trucks on the road shrinks their share, and “involved in crashes” was never “at fault” anyway. 4: per-member average doubling while membership falls by two-thirds is a smaller total. The denominator, always the denominator.

•••

File 6: silence = evidence (the build-mode bone)

The bone. The other 5 files attack bridges. This one builds them wrong. In build mode, your job is to find what the facts force, and the facts are all you have: what the stimulus doesn’t say proves nothing about what’s false. The bone is treating silence as testimony, and it usually works by letting your picture of a normal world fill the gaps. The stimulus says a man sells flowers; your brain quietly adds “and nothing else,” because that’s how flower vendors usually work. The test charges full price for “usually.”

Easy costume, worked.Nothing in the course syllabus prohibits laptops. So the professor must allow them. The fact: the syllabus is silent. The conclusion: a policy exists. Between them, a bridge made entirely of assumption: that the syllabus is the complete record of the rules. Policies get announced out loud, posted on doors, invented on Tuesday. Silence in 1 document is silence, not permission.

Hard costume, worked. The publisher’s spring catalog lists 40 titles, of which 12 are mysteries. Therefore mystery is the catalog’s largest category. Every fact is clean, the arithmetic is real, and the conclusion still isn’t bought, because the stimulus is silent about the other 28 titles. They could be 28 separate categories of 1 book each, which makes mystery the giant. They could include 15 romances, which doesn’t. The tempting conclusion needs the unlisted books to be spread normally, and “normally” is not a brick on the page. In build mode, every right answer is assembled entirely from bricks on the page, and the most seductive wrong answers are 1 plausible brick short. That sentence is the whole job. Its signature traps. In must-be-true questions, the bait is the choice that merely fits the facts (consistent, reasonable, probably true in real life) sitting beside the answer the facts force. Fits is not forced. The full-dress version of this trap, a 5-choice question with a vendor named Dmitri, is waiting in the next chapter, and you’ll be ready for it.

Journal handle: silence = evidence. Forecast: any build-mode question where the tempting answer needs the world to be normal.

Micro-drill, file 6. Different format, because build mode is a different job. Facts, then a candidate conclusion: write proven or not proven, plus why, 1 line.

1. Facts: All staff badge in at the side door. All visitors sign the front-desk log. Dana didn’t sign the log. Candidate: Dana is staff.
2. Facts: The shelf holds 30 books; 18 are paperbacks. Candidate: at least 1 book on the shelf is a hardcover.
3. Facts: The museum is free on Sundays. Today’s admission cost $14. Candidate: today is not Sunday.
4. Facts: Most attendees at the conference drove there. Candidate: most attendees own cars.

Key. 1: not proven; the facts never said everyone is staff or visitor (a contractor walks through both rules untouched). Silence about the categories is silence. 2: not proven, and the trip wire is a word swap hiding inside build mode; the other 12 books are non-paperbacks, but non-paperback isn’t hardcover (audiobook cases, spiral-bound, journals). 3: proven, and worth savoring; if Sunday, then free, and it isn’t free, so it isn’t Sunday. That’s the door’s 1 legal reverse walk, and the next chapter makes it a tool. 4: not proven; drove is not own (borrowed, rented, carpooled). The bricks run out exactly 1 word short.

•••

Capstone: the files in disguise

One more drill before the cabinet closes, and it’s the one that proves the files have moved into you. The 6 drills above each told you which bone to expect. The test never will. So here are 10 arguments with the labels stripped off, mixed, in no order, the way they’ll actually arrive. For each, do 2 things in writing: name the file, then name the gap in 1 line. 2 of these carry more than 1 bone, because real questions do. Pen. Key after, no peeking.

1. Since the new principal arrived, test scores at Lincoln Middle School have climbed 3 years running. The principal’s leadership is clearly raising achievement.
2. Of the 50 alumni who returned for the reunion, 90 percent said their degree was worth the cost. Graduates overwhelmingly feel the program paid off.
3. Anyone admitted to the honor society has a GPA above 3.8. Nina’s GPA is above 3.8, so she’ll be admitted.
4. Streaming now accounts for a larger share of the studio’s revenue than ticket sales do. So the studio earns more from streaming than from tickets.
5. The Tribune was named the region’s most-read newspaper this year, so it’s the region’s most trusted source of news.
6. Facts: every car in the company fleet is serviced monthly; the sedan was not serviced this month. Candidate: the sedan is not part of the fleet.
7. The clinic that prescribes the most antibiotics also has the lowest rate of untreated infections, so prescribing more antibiotics reduces untreated infections.
8. A survey on the city’s website found that 80 percent of respondents oppose the new parking meters. Residents clearly reject the plan.
9. Unless the shipment clears customs by Thursday, the launch is delayed. The shipment cleared customs Wednesday, so the launch is on schedule.
10. The percentage of graduates entering public-interest law has fallen over the past decade, so fewer graduates are taking public-interest jobs than before.

Key. 1: after = because; 3 years of “after” is still after, and a new math curriculum, a changed student body, or easier tests are all unaddressed suspects. 2: some = all; the alumni who return for a reunion are the ones who feel good about the place, a sample that picked itself, and 50 satisfied returnees say nothing about the dropouts who didn’t come. 3: the one-way door, walked backwards; the rule runs admitted-to-high-GPA, and Nina’s walking it in reverse (high GPA doesn’t force admission; the society may cap its numbers or require more). 4: percent vs. count; a bigger share of revenue can mean ticket sales collapsed, not that streaming dollars exceed ticket dollars in raw terms. Wait, read it again. It actually does compare 2 shares of the same revenue, so that comparison is fair. But “larger share than” still isn’t “earns more from,” unless those are the only 2 sources. A third revenue stream changes everything. Denominator, always. 5: the word swap; most-read and most-trusted are cousins, not twins, and a tabloid outsells a broadsheet daily. 6: silence = evidence run as a door contrapositive; fleet cars are serviced monthly, the sedan wasn’t, so by contrapositive it’s not a fleet car. Proven, actually. The trap is assuming you can’t conclude; you can, because this one’s a legal contrapositive walk. 7: after = because with the arrow begging to reverse; clinics in sicker areas prescribe more and still have infections, or the causation runs the other way. 8: some = all; a website poll is self-selected twice over (you visit the site, you choose to answer), and the angry are the ones who answer. 9: the one-way door, negated; “unless X, Y” means no-X-then-delay, and clearing customs removes 1 trigger for delay without promising the launch survives the dozen others. 10: percent vs. count; the share entering public-interest law can fall while the raw number rises, if the graduating class grew. Score yourself loosely, mark the rematch for 1 week out, and rework any file where 2 or more items fooled you. The standard, one last time: not done, easy.

Closing the cabinet

6 files. The test will show you others, and when it does, your journal catches them; a bone you name yourself sticks harder than any of mine. But these 6 are the repeat offenders, and something important just changed: you now know them by name, with worked costumes and a written tally, before your first timed section. Most students meet these bones for the first time at full speed, under a clock, with a proctor watching. You’ve now met each one twice, slowly, with a pen.

What you can’t do yet is cash them at speed, and there’s exactly 1 skill that converts file knowledge into fast points. Zach used it to call answers before reading a single choice. It gets the whole next chapter.

•••

This week’s assignment, 3 parts.

Tonight: copy all 6 handles into your journal’s LR page, with a tally column. This week: work the micro-drills, 2 files a day, every answer in writing before the key, and rework any file where you went 2 for 4 or worse. End of week: take 10 untimed official LR questions from LawHub, and for every troubled one, run the full 20 minutes; your step 5 forecast should now name a file.

•••

Free companion #7: the Skeleton Files card. All 6 files on 1 reference card, tells and killing questions included, with a journal starter page attached. The card walks all 6 files, 1 bone at a time. Free at unpluggedprep.com/books.

•••

KEEP THESE 3

The 2/3 Rule: LR gets 2/3 of your hours by default, but hours follow recoverable points, so your miss ledger can override the default. Recount monthly.

The section has 2 jobs (attack a gap, build from facts) and 6 repeat offenders with files: after = because, some = all, the word swap, the one-way door, percent vs. count, and silence = evidence. The bolder the conclusion, the wider the gap.

Files are for working, not reading. 24 micro-drills back there. Pen.

•••

Next chapter, the skill that cashes the files: written predictions before you touch a single answer choice, the translator that collapses the question-type zoo into 1 move, and the door machinery that build mode runs on. Zach’s trick. Your turn to learn it.

Watch: How to Destroy LSAT Logical Reasoning (Even If You Blank Out on Every Argument)

How to Destroy LSAT Logical Reasoning (Even If You Blank Out on Every Argument)

This chapter is from Anyone Can Beat the LSAT by Steve Schwartz, LSAT Coach and Founder of LSAT Unplugged. I've been coaching the LSAT since 2005.

Published July 28, 2026.