Accuracy of myLSN vs. Law School Predictor Forum

[Mr. Frodo]

Which of myLSN and the Law School Predictor is most accurate/reliable regarding chances of admission?

Not sure where LSP gets its data, and its blocked at work (no idea why) so I can't check currently.

Just curious.

[2014]

LSP is based in part on published admissions indexes and when those can't be found TLS and LSN information is substituted. This means that individuals don't skew the predictor as much, but changes from cycle to cycle are not picked up either. That is why for like this cycle LSP has been wildly inaccurate for some.

LSN is completely anecdotal and you have to trust that people are 1. entering accurate data and 2. that the sample size is sufficiently predictive. If you are good with both of those assumptions, LSN is probably the better source of information and Mylsn.info is just a different way to sort and express the data you see on the graphs on LSN.

[Mr. Frodo]

Alright, thanks. From reading that something like 87% of people were accepted if LSP said "admit," I thought this correlation tended to prove that LSP was pretty accurate. However, if people really do reflect their cycles correctly and truthfully on LSN, which I feel they do for the most part, it may be a more powerful tool. If I put in the numbers I think I'm capable of in LSP, I'm in to a lot of my targets, and strong consider at others. If I put in the same numbers to myLSN, I'm "green" for about half as many schools, although if I select the option to NOT count waitlists as denials, then I'm green in more schools on myLSN than LSP.

Does this just mean a lot of people really do get in off waitlists at a lot of the T-14?

This is where my curiosity as to which was more accurate stems from.

[2014]

It really depends on what school you are talking about. This cycle there has been heavy WL movement at CCN which trickles down to the rest of the T14. Next year their might be less WL movement but if that is the case you should expect more initial acceptances which benefits you the same.

[Mr. Frodo]

It really depends on what school you are talking about. This cycle there has been heavy WL movement at CCN which trickles down to the rest of the T14. Next year their might be less WL movement but if that is the case you should expect more initial acceptances which benefits you the same.

Cool. Thanks again. I've noticed Penn seems to WL many, many applicants with numbers that can get people straight into comparably ranked schools directly (i.e. 169-170 LSAT and 4+ LSDAS). Are they typically so selective in giving direct admission to applicants that are pretty much directly at or very close to their typical standards of admission?

And you're right, perhaps with less applications this year, this could change for the better, or just result in less waitlists, which would be good IMO.

[Yardbird]

LSN > myLSN.info > LSP

1. Law School Numbers
Allows you to look at the raw data and make inferences using charts, application information (softs) - taking into account URM status, work experience, residency, years out of undergrad, degree GPAs, reported trends, etc. You can look at trends for each individual year in the graphs or by looking at individual applications (i.e. GPA floors). You can also see how one person with similar numbers did at schools you are thinking of applying to.

2. myLSN.info
Takes the LSN raw date (specifically GPA, LSAT, application dates, early decision v regular decision, scholarship information and fee waiver information) and presents it in a statistical table looking at numbers of people being admitted with similar numbers. Doesn't account for any softs other than URM status. Take the percentages with a grain of salt (one output shows 6% of students getting a fee waiver from a school, but if you actually do out the math it's 1 student - average financial aid is for that small percentage receiving financial aid, not out of all the applicants). You can look at individual cycles from 2003-present or multiple cycles. It also offers a master graph of all cycles with "thermal hot spots" for areas of the chart with high admission rates.

3. Law School Predictor
Takes median and GPA information from LSAC/ABA data sheets and as reported on school websites to create indices for the 25th and 75th percentiles. It then takes your information and calculates chances at schools based on those indices. Not too helpful for splitters or URMs. Doesn't account for any softs and does not allow you to account for different cycles. If you are not a splitter, it can be used as a good baseline to find out what schools you are "in range for". A DENY on LSP is equivalent to a reach school, CONSIDER is a target school and ADMIT is a safety school (again, not useful for splitters since it gives silly results for schools with hard GPA floors).

Just note however, that LSN and myLSN are both skewed because the applicants using the sites are those who are more likely to be on this site (i.e. top performers). LSP is less biased but is inaccurate in that it doesn't account for GPA/LSAT floors and is terrible for URMs and normal/extreme/reverse splitters.

[Ti Malice]

OP, LSP is also not of great use for schools with “black box” admissions processes (Yale, Stanford, and Berkeley, among the top schools). It’s important not to confuse the percentages that LSP provides with one’s actual probabilities of admission. The percentage that LSP reports is the site’s assessment of how your LSAT/GPA combo (index) compares to those of other admitted/attending applicants. LSP’s calculations do not take note of the proportion of applicants with similar numbers that were not admitted.

For law schools where admissions outcomes correlate very strongly to applicant numbers, this is not such a big deal. For schools that place significant weight on other factors in addition to LSAT and GPA, treating LSP’s percentages as probability assessments is problematic. For example, you might well have numbers that, according to LSP, are better than those of ~75% of those admitted to Stanford in a given class, but it could still be the case that a large majority of applicants with your numbers were ultimately rejected. To see this, plug 173 and 3.9 into LSP, and then compare the results to these self-reported outcomes for applicants to Yale, Stanford, and Berkeley with very similar stats:

http://myLSN.info/dispresults.php?sk=y408p

[Mr. Frodo]

OP, LSP is also not of great use for schools with “black box” admissions processes (Yale, Stanford, and Berkeley, among the top schools). It’s important not to confuse the percentages that LSP provides with one’s actual probabilities of admission. The percentage that LSP reports is the site’s assessment of how your LSAT/GPA combo (index) compares to those of other admitted/attending applicants. LSP’s calculations do not take note of the proportion of applicants with similar numbers that were not admitted.

For law schools where admissions outcomes correlate very strongly to applicant numbers, this is
not such a big deal. For schools that place significant weight on other factors in addition to LSAT and GPA, treating LSP’s percentages as probability assessments is problematic. For example, you might well have numbers that, according to LSP, are better than those of ~75% of those admitted to Stanford in a given class, but it could still be the case that a large majority of applicants with your numbers were ultimately rejected. To see this, plug 173 and 3.9 into LSP, and then compare the results to these self-reported outcomes for applicants to Yale, Stanford, and Berkeley with very similar stats:

http://myLSN.info/dispresults.php?sk=y408p

Thanks a lot. I will check this out and let you know if I have any further questions. I think you covered them all though!