Using the Medians Forum

[pleasesendhelp]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

[20170322]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

This seems wrong. If you're above both 75ths, I would posit you have a much great chance of getting in than 75%. Same for the 25ths, but lower chance. Just use LSN for any value based judgement.

[blueapple]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

This seems wrong. If you're above both 75ths, I would posit you have a much great chance of getting in than 75%. Same for the 25ths, but lower chance. Just use LSN for any value based judgement.

Yeah. If you're below both 25ths (below both 50ths, really) you're unlikely to get in because you're not helping either median.

[cavalier1138]

Yeah, the medians aren't used the way you think they are. Medians just tell you what numbers are needed to be a lock at a school, and median trends can tell you what a school might favor (i.e. Berkeley favors reverse splitters, Northwestern favors splitters, etc.). But you cannot equate them with percentage chances of getting in the way that you're trying to.

[Clemenceau]

I think the fascination with the medians stems more from an interest in seeing the trends in applicant pool competitiveness, not because medians offer enormous predictive value wrt admission. All you can really take from the medians is that schools are going to want to hold/increase those numbers next year, and the applicants they admit will reflect that.

[R. Jeeves]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

your assumption is wrong. but you dont need to reevaluate your existence - well, not for this reason anyway.

[goldenbear2020]

I would assess chances as follows:

Below both medians = 1-5%
At least one median = 25-75% (with just median LSAT being better than just median GPA)
At least both medians = 60-100% (except for Yale/Stanford and YP cases)

[pleasesendhelp]

Yeah, the medians aren't used the way you think they are. Medians just tell you what numbers are needed to be a lock at a school, and median trends can tell you what a school might favor (i.e. Berkeley favors reverse splitters, Northwestern favors splitters, etc.). But you cannot equate them with percentage chances of getting in the way that you're trying to.

A lock is 100% or very close chance right? So then being at 75th percentile for both is a lock.

[pleasesendhelp]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

This seems wrong. If you're above both 75ths, I would posit you have a much great chance of getting in than 75%. Same for the 25ths, but lower chance. Just use LSN for any value based judgement.

LSN is great but I don't see the database alone as enough to make an accurare assumptionm If you've got a 170 and a 3.3 then there's maybe 5 people with the same on the LSN table. If they got in and I have the same number, do I take that as having 100% chance of getting in at those schools? How many people who apply/get in go back to report on LSN? Is it really enough to be representative? That's why I like the medians better. Larger sample.

[20170322]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

This seems wrong. If you're above both 75ths, I would posit you have a much great chance of getting in than 75%. Same for the 25ths, but lower chance. Just use LSN for any value based judgement.

LSN is great but I don't see the database alone as enough to make an accurare assumptionm If you've got a 170 and a 3.3 then there's maybe 5 people with the same on the LSN table. If they got in and I have the same number, do I take that as having 100% chance of getting in at those schools? How many people who apply/get in go back to report on LSN? Is it really enough to be representative? That's why I like the medians better. Larger sample.

Yeah, larger sample, but your underlying assumptions are incorrect so your conclusions will be as well.

If you're unsatisfied with the sample sIze, expand the range. It'll reduce accuracy in some ways but expand it in others.

[lymenheimer]

First off, thank you to those contributing to the LSAT/GPA Medians for 2019!

My question is, How does one REALLY use the given medians for the past year? I've been operating under the assumption that if you're in the 75%, you've got a 75% shot of getting in, 50% for 50%, 25% for 25%. If you're over/under then +/- to that I guess.

for splitters it's different obviously, if you're in the 75% for LSAT and 25% for GPA then I estimate usually at a 50%.

Seems REALLY arbitrary, but pretty much what 99% of what everyone on TLS goes by it seems. Any thoughts? I'm at/above most medians, so i don't worry too much about getting in, my concern is scholarships. Plus, I'd like to know if my assumption is wrong and I need to seriously reevaluate my entire existence.

This seems wrong. If you're above both 75ths, I would posit you have a much great chance of getting in than 75%. Same for the 25ths, but lower chance. Just use LSN for any value based judgement.

LSN is great but I don't see the database alone as enough to make an accurare assumptionm If you've got a 170 and a 3.3 then there's maybe 5 people with the same on the LSN table. If they got in and I have the same number, do I take that as having 100% chance of getting in at those schools? How many people who apply/get in go back to report on LSN? Is it really enough to be representative? That's why I like the medians better. Larger sample.

Youre reading it incorrectly. It's not that you have a 100% chance of getting in. But of those people that reported, 100% have gotten in. Better than assuming you know what the medians mean and if they're a predictive numerical measurement of your chances.

[a_j_l]

There's also the LSAC UGPA/LSAT search if you're worried about sample size.

https://officialguide.lsac.org/release/ ... alsat.aspx

[CPAlawHopefu]

In my experience, assuming you are around the median GPA (less than 0.05 deviation from the median), then having 1-2 point below the LSAT median gives you a reasonable shot at admission (assuming you apply early, meaning by the end of October), 2-3 below is a waitlist, and any further deviation leads to a rejection. Of course, this is a very general statement under the assumption that you are not a URM and have very average softs.

[lymenheimer]

In my experience, assuming you are around the median GPA (less than 0.05 deviation from the median), then having 1-2 point below the LSAT median gives you a reasonable shot at admission (assuming you apply early, meaning by the end of October), 2-3 below is a waitlist, and any further deviation leads to a rejection. Of course, this is a very general statement under the assumption that you are not a URM and have very average softs.

What experience is this? Anecdotal based on a sample of one (Yours)?

[CPAlawHopefu]

Mine, few TLSers, and several of my friends. The pattern seems obvious.

[pleasesendhelp]

Hrm.. I despise the LSAC index because I believe they do go by the medians but VERY strictly without accounting for even slight splitter situations (at median LSAT, slight below median GPA). Maybe in that way they're actually incredibly accurate.. Not according to most on TLS though.

Still, I want to be able to derive something from the medians..it makes no sense to not use them as predictors.

[bmathers]

If you are above both 75 percentiles, like stated earlier... you are looking at a VERY substantial scholarship (read: probably a full ride)

[poptart123]

Hrm.. I despise the LSAC index because I believe they do go by the medians but VERY strictly without accounting for even slight splitter situations (at median LSAT, slight below median GPA). Maybe in that way they're actually incredibly accurate.. Not according to most on TLS though.

Still, I want to be able to derive something from the medians..it makes no sense to not use them as predictors.

Before it shows your chances LSAC states they have a 95% confidence level.

[SlippinJimmy]

LSN is great but I don't see the database alone as enough to make an accurare assumptionm If you've got a 170 and a 3.3 then there's maybe 5 people with the same on the LSN table. If they got in and I have the same number, do I take that as having 100% chance of getting in at those schools? How many people who apply/get in go back to report on LSN? Is it really enough to be representative? That's why I like the medians better. Larger sample.

Yeah, larger sample, but your underlying assumptions are incorrect so your conclusions will be as well.

If you're unsatisfied with the sample sIze, expand the range. It'll reduce accuracy in some ways but expand it in others.

Well, he/she has a point about self-selection re: reporting on LSN. People tend to keep better track of their cycle when decisions are made earlier in the process, and acceptances tend to come earlier than rejections. That or they flat out are more likely to report acceptances than waitlist/rejections. I've noticed quite a few profiles where almost every school is accounted for, except for the borderline schools where their decision is not reported. I assume those decisions came much later and were either waitlist or rejections. My understanding is that those profiles are NOT counted in MyLSN, but it would still skew the MyLSN % predictions by reporting a higher rate of acceptance than in reality.

[RamTitan]

Hrm.. I despise the LSAC index because I believe they do go by the medians but VERY strictly without accounting for even slight splitter situations (at median LSAT, slight below median GPA). Maybe in that way they're actually incredibly accurate.. Not according to most on TLS though.

Still, I want to be able to derive something from the medians..it makes no sense to not use them as predictors.

I was wondering that very same thing....I'm at median LSAT for most of the t14, but between median and 25th gpa. The LSAC index put me under a 50% chance of admission for all of the t14.

[pleasesendhelp]

Hrm.. I despise the LSAC index because I believe they do go by the medians but VERY strictly without accounting for even slight splitter situations (at median LSAT, slight below median GPA). Maybe in that way they're actually incredibly accurate.. Not according to most on TLS though.

Still, I want to be able to derive something from the medians..it makes no sense to not use them as predictors.

I was wondering that very same thing....I'm at median LSAT for most of the t14, but between median and 25th gpa. The LSAC index put me under a 50% chance of admission for all of the t14.

Don't trust LSAC. With some numbers you have a better shot of getting into HLS than UVA (based on 175 LSAT, 3.5 GPA) according to them.
I think they take GPA and LSAT and weigh them to be equal, when most people assume (rightly?) that LSAT is more important.

Thus LSAT medians should be a better indicator than GPA medians.

[RamTitan]

Hrm.. I despise the LSAC index because I believe they do go by the medians but VERY strictly without accounting for even slight splitter situations (at median LSAT, slight below median GPA). Maybe in that way they're actually incredibly accurate.. Not according to most on TLS though.

Still, I want to be able to derive something from the medians..it makes no sense to not use them as predictors.

I was wondering that very same thing....I'm at median LSAT for most of the t14, but between median and 25th gpa. The LSAC index put me under a 50% chance of admission for all of the t14.

Don't trust LSAC. With some numbers you have a better shot of getting into HLS than UVA (based on 175 LSAT, 3.5 GPA) according to them.

That's what I figured; it said I had a significantly greater chance of getting into NYU than Cornell

[pleasesendhelp]

LSN is great but I don't see the database alone as enough to make an accurare assumptionm If you've got a 170 and a 3.3 then there's maybe 5 people with the same on the LSN table. If they got in and I have the same number, do I take that as having 100% chance of getting in at those schools? How many people who apply/get in go back to report on LSN? Is it really enough to be representative? That's why I like the medians better. Larger sample.

Yeah, larger sample, but your underlying assumptions are incorrect so your conclusions will be as well.

If you're unsatisfied with the sample sIze, expand the range. It'll reduce accuracy in some ways but expand it in others.

Well, he/she has a point about self-selection re: reporting on LSN. People tend to keep better track of their cycle when decisions are made earlier in the process, and acceptances tend to come earlier than rejections. That or they flat out are more likely to report acceptances than waitlist/rejections. I've noticed quite a few profiles where almost every school is accounted for, except for the borderline schools where their decision is not reported. I assume those decisions came much later and were either waitlist or rejections. My understanding is that those profiles are NOT counted in MyLSN, but it would still skew the MyLSN % predictions by reporting a higher rate of acceptance than in reality.

My underlying assumption is really that somehow using 2 numbers I can accurately predict where i'll be accepted, but seems like everyone has this assumption. And it's not really about the size but the method. Slippinjimmy points it out really. 100 people with a 170 and a 3.3 apply. 10 accepted, 30 waitlisted, 15 waitlisted -> accepted, 15 waitlisted -> rejected, 1 accepted reports their results on LSN. That's entirely possible, but for some reason people will tell me that it's unlikely. Why? The medians? What's in the box?!

Mainly I'm worried that my plans will go to shit because of flawed expectations. Yeah, yeah, I know. Don't put all your eggs in one basket. Sadly, I only have one fucking egg and all the baskets have centers from SpaceJam waiting to cockblock me from my dreams.

[RamTitan]

In my experience, assuming you are around the median GPA (less than 0.05 deviation from the median), then having 1-2 point below the LSAT median gives you a reasonable shot at admission (assuming you apply early, meaning by the end of October), 2-3 below is a waitlist, and any further deviation leads to a rejection. Of course, this is a very general statement under the assumption that you are not a URM and have very average softs.

What experience is this? Anecdotal based on a sample of one (Yours)?

I would say this is extremely anecdotal. I was waitlisted at a t14 school where my GPA was between 25 and 50, but my LSAT was median. If CPA's hypothetical was the case, then it would seem schools valued GPA more than the LSAT, which is contrary to everything else these boards espouse. But then again, maybe we're all deluded

[pleasesendhelp]

In my experience, assuming you are around the median GPA (less than 0.05 deviation from the median), then having 1-2 point below the LSAT median gives you a reasonable shot at admission (assuming you apply early, meaning by the end of October), 2-3 below is a waitlist, and any further deviation leads to a rejection. Of course, this is a very general statement under the assumption that you are not a URM and have very average softs.

What experience is this? Anecdotal based on a sample of one (Yours)?

I would say this is extremely anecdotal. I was waitlisted at a t14 school where my GPA was between 25 and 50, but my LSAT was median. If CPA's hypothetical was the case, then it would seem schools valued GPA more than the LSAT, which is contrary to everything else these boards espouse. But then again, maybe we're all deluded

I think, for my particularly case as an international applicant (though U.S. citizen), that LSAT is definitely weighted more. Do you mind sharing which t14? or at least upper or lower?