We've rebuilt our admissions predictor from the ground up. The interface didn't change, but it uses a completely new algorithm, now trained on the most recent cycles. Expect lower but more accurate predictions.
What changed
- Newer data. It's now trained on the three most recent cycles, including 2025. The old version was learning from an era that was meaningfully less competitive than the one you're applying into.
- Stats are measured against current medians. Your GPA and LSAT are now evaluated relative to each school's medians rather than in absolute terms, so median drift is built in instead of ignored.
- Your numbers are treated as a package, not a checklist. The old model added up LSAT, GPA, and URM status as independent boosts. The new one learns how those factors interact with each other and with the tier of school you're applying to — the same LSAT can mean very different things depending on the rest of your profile and where you're applying.
- Fewer extreme predictions. The old model was happy to tell you 95% at a T14 on numbers alone. It shouldn't have, and the new one won't.
What that means for you
Most people will see lower numbers, and the drop will be largest at reach schools and for profiles that were previously getting very high admit odds. Nothing about your candidacy changed — the old figures were just too generous.
How we know it's better
Tested against actual 2025 outcomes, log loss (a standard measure that punishes confident wrong answers) fell from 1.35 to 0.85, and predicted probabilities now track observed results closely across the whole range. For scale: applicants the old model gave a ~90% admit chance were actually admitted roughly half the time. That gap is largely gone.
The usual caveat
It still only sees GPA, LSAT, and URM status. Essays, softs, work experience, letters, and school-specific priorities are invisible to it. Treat the output as a well-calibrated starting point for building a school list, not a verdict.
61 comments
damn i'm cooked
ha well there went my odds at every school I was looking at...
Awww my chances are so much lower for the schools I had on my list, but at least I have a more realistic sense of which schools I have a good chance at.
Could you release a video or article explaining in depth the methodology behind the admissions predictor?
Does this give significantly more weight to GPA? For example I put a 3.7 GPA and a 180 LSAT and essentially everything in the top 25 is labeled a reach
@KiranPoudel seems a little weird for me too. Michigan turned from a "Target" to a "Reach" after I added 5 points to my hypothetical LSAT. At the end of the day, I guess this shows that everything is getting much more competitive...
@Mjamail honestly as weird as it sounds/seems that lines up with the actual admissions results from the last cycle anecdotally at least. People with 3.9 gpa, 178 lsat, still rejected, it’s tough
@KiranPoudel The GPA can go up to 4.33, so give it your gpa on a 4.33 scale.
Can you release an in-depth overview of your methodology? I only ask because you have a pretty strong financial incentive to claim "lower but more accurate predictions" at this point in the cycle where people are willing to throw their money at you in desperation.
@PhillyRed19 We are releasing a podcast soon with more details on the methodology. We also put some details in this blog post: https://7sage.com/blog/admissions-predictor-update-same-look-new-engine
I'm with another commenter that it seems that either the current version is or the previous version was way off. I went from a good mix of reaches, targets and safeties to all super-reaches...this is extraordinarily disheartening.
@McSec Because of how unusual last cycle was I really wouldnt put too much stock in these changes. This now takes into account the last admission cycle which was historically difficult. People with 3.9's and 175+s were getting waitlisted. They are more or less just educated guesses at best anyways.
@McSec Well, the man said that the previous version was less accurate since the current version is more accurate.
FWIW any predictor should be taken with a gigantic grain of salt. All of these data are backward looking, are now considering one of the hardest and most unusual admissions cycles in history.
Would having a feature for in-state v.s out of state be helpful? Do schools increase your chances if you want to stay in your area to practice law?
@Kihla333 also maybe early decision
Hi there,
I love this feature so much and appreciate all the stuff you’ve been coming out with as a company lately.
If you could include Canadian law schools as well, I would be jumping for joy :D We don’t have that many, and it’s a pretty simple formula to convert the GPA..
@Lillia555 seconding this! I'm looking to apply to both US and Canadian schools, and this would be super helpful :)
I know there's probably no way to really predict/quantify this, but is there a way to add 'non-traditional' as a factor (like the URM toggle)?
@EEC_517 This data is not publically available!
I think something is wrong with the calculator; at 170, Columbia and NYU are Super Reaches, and at 166 they are Reaches.
@mitchmcconnellburner From my understanding it's based on real admission information from the last cycle so there may still be some noise in the data, i.e. an uncommon amount of really good 166 scorers applied to Columbia and NYU last cycle, or some combination of yield protection (colleges denying applicants they think will chose another school, remember it benefits a law school to deny you) - so I bet they are ironing stuff like that out but I bet theres where it comes from
Penn State = the impossible school
@ElijahWhite yeah cuz it doesnt exist anymore :)
great, thanks for softening the blow ig. "here lies squidward's hopes and dreams"
I get the methodology but it still doesn't make sense. Lowering LSAT from 178 to 177 turns Berkely from a reach to a target
This algorithm makes no sense at all. I've played around with it and even when I bump my score from a 160 to a 170 it only adds like a 5% better chance. That's ridiculous. So even improving your LSAT score doesn't make your chances any higher?
@HeidiLaCoste If your GPA is poor, then potentially yes. How strongly GPA is weighted compared to LSAT depends on the school
@Lillia555 my gpa is 3.9
This is super helpful. Just add onto what everyone else has noticed, some schools should be safety's or targets but are listed as super reaches. This definitely scared the crap out of me, but glad to see I'm not the only one experiencing this.
@sgdressel We're fixing it.
I think something is wrong because with my GPA and predicted LSAT, some schools should be safety's or targets but now even low ranking schools are super reach...
I can tell you for certain something is wrong with it, for example, when I put in a 178 and 3.75 GPA, Texas A&M is a "target school" but when I change the metrics to a 180 and 3.77, Texas A&M becomes a "reach."
@binks @mitchmcconnellburner the predictor is capturing some real jitter in the data, and its results are better fitted to last cycle's results than the old predictor's. Finally, there is a real effect here: law schools protect their yield. It's possible, for example that your chances at Texas A&M really are worse with a 180/3.77 than a 178/3.75, because the admissions officers might think the 180/3.77 candidate is less likely to accept.
All that said, thank you for pointing out the weirdness. We're looking into it.
Thank you this seems much more accurate it previously had my 174 as a "safety" at uchicago lmao
I might be crazy but this post feels like it was written by AI.
@CarterPoniatowski It has the cadence, I agree. Though hard to tell these days.
Part time programs?
I'm so happy that you gave us realistic predictions, and opt for giving me a lower anticipated LSAT score / admissions chance rather than an inflated one. Thank you for the courage of being honest. We aren't all going to Harvard.
It looks like the gpa is on a 4.33 scale, a small change could make that more noticeable to other users.
Rather than feeling disheartened, I feel that the information 7sage is providing me is reliable and realistic. I like that you don't hide the room we all have to grow and we can see the impact of our continued studying and improvement.
@Lillia555 Why is gpa on a 4.33 scale? my undergrad is a 4.0 scale, and an A- is worth the same as an A+, do i seriously have to compete on the GPA with people whos undergrad GPA system allowed them to go above 4.0?
@binks I do not know why, 7sage will have to update this and clarify.
Either they made a mistake or everyone’s GPA is scaled up to meet 4.33, I do not know
@Lillia555 @binks 7sage does that to be in line with LSAC. LSAC recalculates your GPA when you apply, so your school's grading system may be different from theirs and thus your GPA may be different. LSAC's CAS GPA is on a 4.33 scale.
It absolutely disadvantages students whose schools do not have A+ in their system, because it is literally impossible to get above a 4.0 under the CAS system.
You can find the CAS general grading system here: https://www.lsac.org/applying-law-school/jd-application-process/cas/requesting/transcript-summarization#GradeConversionTable. You can look up how your specific school's grading system is interpreted by LSAC here: https://members.lsac.org/MenuItems/IGUGS/IGUGS-Search.aspx
love how it just pushed all predictions down