LSN versus LSAC Search Forum

[Swimp]

Hi all. Anybody have any idea why LSN paints so much of a rosier picture of admissions chances than LSAC's search? It seems like LSAC's tool must be based on a much more complete set of data than LSN, but people around here rely so heavily on LSN and not at all on LSAC's search tool (that I've noticed, at least).


LSN: http://www.mylsn.info/app_checker.php
LSAC Search: https://officialguide.lsac.org/Release/ ... fault.aspx

[Dr. Review]

LSAC: Enter GPA/LSAT, see broad results that can be hard to interpret, but may also provide you with a somewhat realistic starting point for your cycle (targets/reaches/safeties).
LSN (http://www.lawschoolnumbers.com/): Look at a school, find trends including URM histories. Combined with --LinkRemoved-- allows for data mining from past cycles (I may also mention that you are conflating LSN with mylsn.info, which are based on the same data, but which are useful for slightly different functions).

I'm not really sure what you mean when you say that LSN paints a rosier picture.

[Swimp]

Okay, let's take 3.5/172 as a test.

If I go to LSAC's site and ask them to show me data for applicants with those numbers, it tells me that in 2012, those people would have had a 30-40% chance of getting into NYU, 13-25% at Duke, and 19-31% at Michigan.

MyLSN, using info only from the last two cycles (yeah I know, it's not the LSN site, but it's based on the exact same data) gives the following %ages:

NYU: 83%
Duke: 80%
Michigan: 50%

Pretty drastic difference. It just makes me think LSN's info is very strongly skewed toward applicants who get in. Maybe because they take the time to keep their profiles updated during their cycles, while less lucky applicants aren't interested in tracking their disappointments.

I mean, look, I'm applying next cycle and I would like nothing more than to believe LSN, but I can't seem to think of a reason why I should trust that data over the data LSAC provides.

[20130312]

That's because LSAC is notoriously bad at predicting applicants who have one of their numbers above the median and one below it.

[Swimp]

That's because LSAC is notoriously bad at predicting applicants who have one of their numbers above the median and one below it.

That's interesting. It seems like it would be so easy to just take all the data they have on the previous cycle and faithfully report the actual %ages of applicants who got in/didn't.