How much is each LSAT point worth, GPA-wise? Forum

[MysticalWheel]

I'm thinking each LSAT point is worth about 0.075 GPA points, so hypothetically, a 175/3.5 would be roughly as attractive as a 177/3.35? Hmm, doesn't seem right.

MW

[pwyoung]

There's no point in making this sort of equation.

[DieAntwoord]

I'm sure the worth changes depending on what range you are in.
For example the one point that seperates 179 and 180 is probably worth nothing.

[capitalacq]

There's no point in making this sort of equation.

this.

There's no point in making this sort of equation.

this.

this.

There's no point in making this sort of equation.

this.

this.

this.

[kazu]

There's no point in making this sort of equation.

this.

There's no point in making this sort of equation.

this.

this.

There's no point in making this sort of equation.

this.

this.

this.

this.

[motiontodismiss]

I'm thinking 169->170 is worth like 1 point for splitters (very slim shot at T14 vs. in with ED). But 179->180 is worth nothing.

[CG614]

Don't worry about the haters. It is an interesting question to think about, especially since around admission time, you only have control (kinda) of one variable. I'd be intrigued to see an empirical study done, but I am a 1L and don't have the time. ha

[tallboone]

The above answers are actually incorrect and it depends on the school. Some law schools (GWU, others I can't think of off the top of my head) pretty much use an index system to admit people primarily based on numbers alone. So, you plug in X.XX GPA and multiply it by 1XX LSAT to get a number. If you are above that number, you get it barring negative factors.

Other schools use a holistic admissions process and and do not give you an index number.

For schools in the T-14, you should be above at least one median to have a shot.

[ahduth]

There's no point in making this sort of equation.

This is actually pretty wrong, both because, as tallboone points out, schools do this all the time to determine who they're even going to look at, but also because you would use something like this to determine reach/safety schools.

For schools where we have detailed information on admit profiles, like was posted in this thread:

http://www.top-law-schools.com/forums/v ... 2&t=138084

it would be pretty simplistic to construct an formula that would tell you how much a certain shift in LSAT score would "worth" percentage-wise compared with a certain shift in UGPA.

Edit: for splitters in particular this type of analysis is implicit in whether or not they even bother applying. All the time we see threads from people saying, "I got a 171 but I have a 3.1 GPA, should I bother applying to NYU?"

[YCrevolution]

..

[AreJay711]

Here's an arbitrary guess assuming it is linear (which it clearly isn't). A better guess would be to look at the equations for the schools in LSP. I messed up my terms at some point bc I'm at work and kept having to actually do work in the middle so if I missed any when I went back and fixed it please forgive me.

Using the numbers on LSP for Michigan, the 25-75 LSAT is 167- 170, a difference of 4, and for the GPA it is 3.55 - 3.84, a difference of .29.

0.29 (GPA) = 4 (LSAT)

(LSAT/GPA)= 13.7 ; (GPA/LSAT)= 1/13.7


So .1 difference in GPA= 1.37 difference in LSAT according to our model


Test it out:

My GPA/LSAT: 3.93/168. What LSAT would I need to compensate a 0.2 drop in my GPA at Michigan?

.2/LSAT = 1/13.7

0.2 = (1/13.7) * LSAT

0.2 / (1/ 13.7) = LSAT

LSAT = 2.74

3.93/168 =?= 3.73/171 (round up)


3.93/168 % at/below according to LSP: 66%

3.73/171 5 at/below: 74%


Conclusion:

So probably not right but maybe If we could incorporate the relative importance in available from LSP we could get a better prediction.

By plug and chug it looks like 0.1 GPA = 1.0 LSAT at Michigan but of course this is not linear (so it depends where you start and how far away from that you go) amd at some point either GPA (or LSAT) will be too low or you will become an attractive splitter that could be balanced off by someone with a better GPA than LSAT.... like me!