how accurate are the admissions calculators?

[boilerguy84]

i have looked at the LSAC calculator and the chiasu one, and was just wondering if anyone knows how accurate these actually are? for some reason i just don't really trust them.

[Eveman in Ingmarland]

It soooooooooooooo depends on the school and your soft factors. Some schools weight things differently (Columbia seems to weigh the LSAT more, Boalt Hall the GPA), and some are obsessed with undefined soft factors (see Stanford Law School at Law School Numbers, they can reject someone with a 3.9+ and 175+ and admit multiple people with 3.8's and 170s? What are they looking for?). If you are a 'mainstream' candidate (non-URM, middle to upper class, 0-2 years work experience), then these things may be more relevant (altho not at all perfect) and can be a (very) rough guide to where to apply. But if you have something different to offer, then do not be afraid to aim higher than these calculations advise.

Honestly assess yourself and what you have to offer. If you have a 3.1 and a 163 and average soft factors, maybe Harvard is out of reach. But don't let the #'s control everything; if you have a special story to tell, something unique to offer, don't let these calculators prevent you from applying to somewhere you really want to go.

This isn't the reason the calculators are bad (and they are bad).

The Chiasu and LSAC things suck because the regression analysis they use seems to treat LSAT and GPA as independent variables--i.e. a 165/3.5 applicant's chances are estimated by multiplying the past performance of applicants with a 165 times the probability of applicants with a 3.5, which is obviously stupid.

In order for an admissions calculator to work effectively, it would have to use data based on the LSAT and GPA combinations of actual past applicants.

The best approximations of this sort of tool currently available are LSN and the admissions charts some schools publish online.

[boilerguy84]

It soooooooooooooo depends on the school and your soft factors. Some schools weight things differently (Columbia seems to weigh the LSAT more, Boalt Hall the GPA), and some are obsessed with undefined soft factors (see Stanford Law School at Law School Numbers, they can reject someone with a 3.9+ and 175+ and admit multiple people with 3.8's and 170s? What are they looking for?). If you are a 'mainstream' candidate (non-URM, middle to upper class, 0-2 years work experience), then these things may be more relevant (altho not at all perfect) and can be a (very) rough guide to where to apply. But if you have something different to offer, then do not be afraid to aim higher than these calculations advise.

Honestly assess yourself and what you have to offer. If you have a 3.1 and a 163 and average soft factors, maybe Harvard is out of reach. But don't let the #'s control everything; if you have a special story to tell, something unique to offer, don't let these calculators prevent you from applying to somewhere you really want to go.

This isn't the reason the calculators are bad (and they are bad).

The Chiasu and LSAC things suck because the regression analysis they use seems to treat LSAT and GPA as independent variables--i.e. a 165/3.5 applicant's chances are estimated by multiplying the past performance of applicants with a 165 times the probability of applicants with a 3.5, which is obviously stupid since LSAT and GPA do not correlate very strongly.

In order for an admissions calculator to work effectively, it would have to use data based on the LSAT and GPA combinations of actual past applicants.

The best approximations of this sort of tool currently available are LSN and the admissions charts some schools publish online.

the issue i have is that there are very few applicants on LSN with my numbers (horrible UGPA due to stupidity during freshman year, and decent enough LSAT[163] for the schools to which i want to apply...low T2 schools, mostly). i know that these things don't factor in things such as grade trends, soft factors, and whatnot, but i can't stand not knowing anything about my chances of getting into schools. i go insane having zero idea as to what's going to happen.

[devilishlyblue]

Well, but that's like basically every other process out there -- jobs, other graduate schools, undergrad, fellowships, etc.  Nobody else publishes calculators.

[rsr28]

It soooooooooooooo depends on the school and your soft factors. Some schools weight things differently (Columbia seems to weigh the LSAT more, Boalt Hall the GPA), and some are obsessed with undefined soft factors (see Stanford Law School at Law School Numbers, they can reject someone with a 3.9+ and 175+ and admit multiple people with 3.8's and 170s? What are they looking for?). If you are a 'mainstream' candidate (non-URM, middle to upper class, 0-2 years work experience), then these things may be more relevant (altho not at all perfect) and can be a (very) rough guide to where to apply. But if you have something different to offer, then do not be afraid to aim higher than these calculations advise.

Honestly assess yourself and what you have to offer. If you have a 3.1 and a 163 and average soft factors, maybe Harvard is out of reach. But don't let the #'s control everything; if you have a special story to tell, something unique to offer, don't let these calculators prevent you from applying to somewhere you really want to go.

This isn't the reason the calculators are bad (and they are bad).

The Chiasu and LSAC things suck because the regression analysis they use seems to treat LSAT and GPA as independent variables--i.e. a 165/3.5 applicant's chances are estimated by multiplying the past performance of applicants with a 165 times the probability of applicants with a 3.5, which is obviously stupid.

In order for an admissions calculator to work effectively, it would have to use data based on the LSAT and GPA combinations of actual past applicants.

The best approximations of this sort of tool currently available are LSN and the admissions charts some schools publish online.

This is the correct answer.  Whoever said the fact that schools weigh things differently being the reason for bad calculators is way, way off.  These calculators account for that.  As Eveman said, their regression analyses are piss-poor.  I'm not saying it's easy to model these carefully, but if you gave me a good 40-hour work week I could come up with something far, far better than what exists now.  Alas, I don't have that luxury of time.

[Eveman in Ingmarland]

This is the correct answer.  Whoever said the fact that schools weigh things differently being the reason for bad calculators is way, way off.  These calculators account for that.  As Eveman said, their regression analyses are piss-poor.  I'm not saying it's easy to model these carefully, but if you gave me a good 40-hour work week I could come up with something far, far better than what exists now.  Alas, I don't have that luxury of time.

It would be very easy to create useful calculators. All it would require is assembling enough data to improve the admissions grids. Right now, the grids schools publish use really thick integers (i.e. all applicants with LSATs between 165-169 are included in the same set). This works well for giving you a general idea of where you might be competitive, but is insufficient for estimating probabilities (all of the applicants in that LSAT range with <167, could have been rejected, and all applicants with >/=167 could have been accepted, for example). If schools narrowed the integers substantially--say to every 1 LSAT point and every .1 GPA point--I think the grids would be very, very useful for estimating your chances, and consequently giving you a clearer idea of where you should apply.

Of course, that would probably mean fewer people paying application fees to schools at which they are not really competitive, and applying to fewer safeties, but I'm sure that has nothing to do with why it hasn't happened... 

[Eveman in Ingmarland]

Well, but that's like basically every other process out there -- jobs, other graduate schools, undergrad, fellowships, etc.  Nobody else publishes calculators.

Yeah, but it doesn't have to be that way with law school applications. From what I can tell this is enough of a numbers game that with access to the right data, most applicants could compute fairly accurate estimates of their chances for admission at a given school.

[bloomlaw]

the issue i have is that there are very few applicants on LSN with my numbers (horrible UGPA due to stupidity during freshman year, and decent enough LSAT[163] for the schools to which i want to apply...low T2 schools, mostly). i know that these things don't factor in things such as grade trends, soft factors, and whatnot, but i can't stand not knowing anything about my chances of getting into schools. i go insane having zero idea as to what's going to happen.

what you can do is use the lsn index calculator... this will find the index(ration between lsat and gpa) you would have with the school, and compare this to the indexes that people have that are getting accepted and rejected.. imo, the index is more important than the actual numbers of lsat and gpa

[DerekShiHarvard]

Calculators are a good starting point to get a rough idea of where you would get in. The next step would be to scrutinize the schools' 25%-25% LSAT ranges, and see if you are between those ranges or not too far off from the 25% mark. Next check lawschoolnumbers.com. Good luck!

[boilerguy84]

thanks for the responses. i guess i'm just going to have to wait and see. my numbers aren't common enough for me to determine much of anything (as i can't seem to come across anyone with a gpa as low as mine with an lsat at 163). it's frustrating, but whatever.