November 2026 LSAT Curve: How Many Can You Miss for a 170?
My November 2026 LSAT curve predictions are 7 to 9 missed questions for a 170 (scores run from 120 to 180), depending on which version of the test you got, and about 20 misses for a 160 on every version. LSAC, the Law School Admission Council that runs the LSAT, hasn't published any November curve, so every number on this page is my estimate.
The curve is the chart LSAC uses to turn the number of questions you got right into your score from 120 to 180. If you're counting the days until scores come out on December 2, this is the page I'd want open.
November 2026 LSAT curve by version
November had several versions of the test, with different questions, and each one gets its own curve. Find your version by the reading passages you remember, then read across. I name each version by its bolded passage. Each number is how many questions I think you could miss and still get that score.
| Reading passages you remember | 170 | 165 | 160 |
|---|---|---|---|
| Bacon and Shakespeare, gamifying work, Indian astronomy, state consent | 9 | 14 | 21 |
| Indus Valley script, Walden, children's stories, placebos | 8 or 9 | 14 | 20 |
| Vermeer, the fur trade, consumer shortcuts, the Fermi paradox | 8 | 14 | 20 |
| Miami English, arbitration vs. mediation, panoramas, autumn leaves | 7 or 8 | 13 | 20 |
Where a box shows 2 numbers, the evidence doesn't settle it. I walk through each row below.
Most people get one extra section that doesn't count toward their score. If you saw 2 reading sections, one of them was that extra one. If one of your 2 reading sections matches a row here, that one most likely counted. People who saw only one reading section, so theirs had to count, reported all 4 of these sets.
If you had different passages, or an older test given again in full, use about 8 misses for a 170, 14 for a 165, and 20 for a 160.
The other sections, the logic questions, also came in more than one version, and I can't match those to a reading set. So read every number here as plus or minus 1.
Don't open your score alone
On December 2, the day scores come out, I'll email you how to read your November score against your version's curve, and how I'd decide on a January retake.
Prefer to watch? I walk through every version's numbers in this 6-minute video.
In the video: 0:00 When scores come out, 0:45 My rule: retake, 1:00 What the curve is, 1:30 How many misses a 170 usually takes, 1:59 How I built the estimates, 3:02 Bacon version: 9, 3:26 Indus version: 8 or 9, 3:51 Vermeer version: 8, 4:03 Miami version: 7 or 8, 4:19 Don't remember? Use 8, 4:39 The full table, 4:51 Aiming for the 150s, 5:20 Where this could be wrong, 6:03 Retake and free January drills. Watch on YouTube.
How many questions can you miss for a 170?
On most November versions, my best guess is 8 misses for a 170. The Bacon version looks like 9, and the Miami version could be 7. For a 165, figure 13 to 15 misses. For a 160, figure 20 or 21. The table below sets those next to real curves anyone can check.
| Score | November 2026 prediction | LSAC's official practice tests | April 2025 test LSAC released (77 scored questions) | September 2025 test LSAC released (78 scored questions) |
|---|---|---|---|---|
| 175 | 3 to 5 misses, by version | 3 to 6 misses (about 4 on the recent ones) | 4 misses | 5 misses |
| 170 | 7 to 9 misses, by version | 6 to 10 misses (about 8 on the recent ones) | 8 misses | 9 misses |
| 165 | 13 to 15 misses | 11 to 16 misses (about 14 on the recent ones) | 14 misses | 14 misses |
| 160 | 20 to 21 misses | about 20 misses | 21 misses | 20 misses |
| 150 | 34 to 35 misses (every version) | 30 to 36 misses (about 35 on the recent ones) | 35 misses | 34 misses |
The practice test column covers LSAC's official practice tests, which are real past LSATs that LSAC puts out for practice, scored the way today's test is scored. The 2025 columns are the 2 most recent real tests LSAC has released. Each is a single version from its month, so other versions that month may have had a different curve.
Look at how little the bottom rows move. On past tests, when a harder version allows 1 more miss for a 170, it allows only about half a miss more for a 160, and for a 150 it doesn't change at all. So your version matters most if you're aiming near the top of the scale.
A 170 on the Bacon version and a 170 on the Miami version mean the same skill, even if the misses allowed differ. That's the point of having a curve for each version.
How I made these predictions
LSAC doesn't publish the curve for most tests. It only publishes it for the few tests it later releases as free PDFs, and those come out a year or more later. The September 2025 test booklet, for example, came out in fall 2026. So right now nobody outside LSAC knows the curve for any November 2026 version. That includes me.
What you can work with is how the test is built. Your raw score is the number of questions you got right on the sections that count, about 75 to 78 questions in all. Wrong answers carry no penalty. LSAC then turns that raw score into your 120 to 180 score with the curve.
LSAC sets each curve so the same score means the same skill on every version. On a harder version you can miss more questions for the same score, and on an easier one you can miss fewer. That's the whole reason curves differ.
And honestly, they don't differ much. On LSAC's official practice tests, a 170 took between 6 and 10 misses every time, and most of the time it was 7, 8, or 9. The recent ones average 8. The 2 real tests LSAC released from 2025 fit right in, at 8 and 9. So my starting guess for every November version is 8 misses for a 170.
Next I checked each version against what its own test takers said. We read the main LSAT forum on Reddit from the first test day through November 16 (about 4,800 comments and 500 posts) and pulled every one that named a passage from a single version, 383 in all. We read each one twice, separately, and marked it hard, easy, or neither for the passage it named. The 2 reads agreed on 94% of the hard and easy calls, and I only used the ones where they agreed.
One brutal passage can fool you, though. It draws a pile of complaints even when the rest of the section was gentle. So we went through again for what people said about the whole reading section. I move a version 1 miss off that 8 only when both checks point the same way. When they split, it gets 2 numbers.
- Bacon version: 9. 24 test takers called its passages hard and 2 called them easy. Of the 10 who described the whole section, 9 called it hard. Both checks agree, so it moves.
- Indus version: 8 or 9. The passage calls ran 27 hard to 3 easy, but 20 of those 27 people complained only about the Indus passage. Of the 15 who described the whole section, 5 called it hard and 10 said the rest was fine or easy. A passage that eats your time still costs points, so I lean toward 9.
- Vermeer version: 8. Close to an even split on both checks. Nothing here moves it.
- Miami version: 7 or 8. The passage calls ran 5 hard to 9 easy, and 9 of the 11 who described the whole section found it easy or manageable. That's the kindest review of the 4, from a small group.
No version moves by more than 1 miss. Reading is only 1 of the 3 sections that count, and by my math on past tests, one hard section is usually worth about half a miss for a 170. So a full move is the exception, and I make it only when both checks agree.
These predictions have weak spots, and you should know them. People who struggled post more. Many Bacon test takers also had a second, easier reading section to compare it with, and that can make the one that counted sound worse than it was. Some Miami test takers had a harder second reading section, which can make Miami sound easier than it was. The whole-section check was done once, not twice, and covers only 46 people. And nobody can check a read like this against real curves yet. I've been coaching the LSAT since 2005. I trust the long-run average to do most of the work here, and Reddit only nudges it.
Was the November LSAT hard?
It depends on your version. Going by what test takers posted on Reddit, the Bacon version was the rough one. Nearly everyone who described that reading section called it hard, and the Bacon passage was the most talked-about passage of the week. The Indus version had one passage people hated and 3 they mostly liked. The Vermeer version drew mixed reviews. The Miami version drew the kindest ones.
How a version felt can't tell you your score, though. If a single section went sideways, don't assume the worst. If you had an extra section that didn't count, it might have been that one.
And a harder version just lets you miss more questions for the same score, so it isn't bad news. If you want to look back at what we posted before test week, it's on our November 2026 LSAT predictions page.
Frequently asked questions
Did everyone get the same test?
No. November had several versions of the test. This comes from what test takers posted on Reddit, not from LSAC. Test takers named at least 4 different reading sets that counted, and the logic sections came in more than one version too. At least 1 person got an older test given again in full, and 1 person in Canada reported a different set. LSAC sets every version's curve so a 170 means the same skill on each one. The number of misses for that 170 can still differ, which is why the table at the top gives each version its own prediction.
When do November 2026 LSAT scores come out?
December 2, 2026. If you'd like a note from me that day on how to read your score, leave your email in the box near the top.
Was my reading section scored or experimental?
Most people get one extra section that doesn't count toward their score, often called the experimental section. If you saw 2 reading sections, one of them was that extra one. People who saw reading only once reported each of the 4 sets in the table at the top, so each of those sets counted for at least some people. If one of your 2 reading sections is in that table, it was most likely the one that counted. That comes from test takers' reports, not from LSAC. And if the one that counted was the rough one, remember that a harder version lets you miss more questions for the same score.
How many questions can you miss for a 165 or a 160?
My prediction for a 165 is 13 to 15 misses, and for a 160 it's 20 or 21. Those numbers change less from version to version than the 170 number does. On the 2 real tests LSAC released from 2025, a 165 took 14 misses on both, and a 160 took 21 and 20.
My test felt easy. Does that mean I did badly?
Not because it felt easy. An easier version just allows fewer misses. Your score depends on how many you got right on your version, and the curve already adjusts for how hard it was.
Will you check these predictions against the real curve?
Yes. If LSAC ever releases the November 2026 test, we'll check our predictions against the real chart and say so on this page.
If you might retake in January, our January 2027 LSAT curve page covers what to expect there. Start with the free January 2027 LSAT predictions and practice sets.
Try one from the January LR set
Loading the question...
Get the rest of the January LR set
The set has more like this one, each with an explanation. Enter your email to keep going.
Or open the setYou're in. The full January LR set is ready.
Start the setI've written free explanations for every question in LSAT PrepTest 159, 158, 157, 156, 155 and 154, with the right answer and why each wrong answer fails. All LSAT explanations are here.
