how useful is Law School Numbers? Forum

[thelawz09]

There are 768 U Chicago applicants with LSN profiles. Last year U Chicago had 5,400 applicants.

For Georgetown, the ratio is 1,296 to 11,649.

You get the picture. For most schools, the sample size is about 10% of the total number of applicants.

Two questions:

1) How is the pool of applicants who post their stuff on LSN different than the overall pool of applicants--in general? Younger? More obsessive? Worse grades? Better chances?
2) Does Law School Numbers provide a reasonably accurate measurement of what your chances are?

[D. H2Oman]

1) Doesn't matter
2) Yes

[jks289]

There are 768 U Chicago applicants with LSN profiles. Last year U Chicago had 5,400 applicants.

For Georgetown, the ratio is 1,296 to 11,649.

You get the picture. For most schools, the sample size is about 10% of the total number of applicants.

Two questions:

1) How is the pool of applicants who post their stuff on LSN different than the overall pool of applicants--in general? Younger? More obsessive? Worse grades? Better chances?
2) Does Law School Numbers provide a reasonably accurate measurement of what your chances are?

I think it gives you a general idea of admission trends. Of course it will miss the outliers, which is what I think lots of people are looking for. The one long shot that got in and means maybe they get in too. If anything LSN reaffirms what most of us already know; Above both medians is safety, between is target, and if you are below one 25% you need to be above the other 75% to have a fighting chance, and URM get different thresholds. I think for this cycle especially, last years graphs are going to be less useful, due to the massive increase in applications.

[dynomite]

1) How is the pool of applicants who post their stuff on LSN different than the overall pool of applicants--in general? Younger? More obsessive? Worse grades? Better chances?
2) Does Law School Numbers provide a reasonably accurate measurement of what your chances are?

Answering these questions would take some time and you'd need a lot of data. Basically, the problem is that this is not a representative sample -- it's a sample of the 10% who choose to self-report.

Ultimately, there will never be a sure-fire answer to the rhetoical question that most applicants ask: "Can I get in here?"

[09042014]

There are 768 U Chicago applicants with LSN profiles. Last year U Chicago had 5,400 applicants.

For Georgetown, the ratio is 1,296 to 11,649.

You get the picture. For most schools, the sample size is about 10% of the total number of applicants.

Two questions:

1) How is the pool of applicants who post their stuff on LSN different than the overall pool of applicants--in general? Younger? More obsessive? Worse grades? Better chances?
2) Does Law School Numbers provide a reasonably accurate measurement of what your chances are?

Polls that fairly accurately predict the presidential election only poll 1200 people out of 120,000,000 voters.

That being said, LSN is not extremely accurate. There are false data points, and it totally ignores most soft factors.

What LSN is useful for is determining where a schools cut offs are. You can look at UChi and see that below 3.4 there be dragons. You can also tell that below 3.7 and below 169 you are almost definitely out.

So look at the chart, and if you see a bunch of green near your number, you are probably in. If you see all red, you are probably out.

Where LSN is bad is the areas where there are acceptances waitlists, and rejections all in one area. If you see 5/10 applicants with your numbers getting in, that doesn't mean you have a 50-50 shot.

[Grizz]

2) Does Law School Numbers provide a reasonably accurate measurement of what your chances are?

Yes, but in general it works better for schools with more applicants.

[ArthurEdens]

LSN is accurate. I applied to 17 schools, and my cycle has gone exactly as the charts would indicate. This is despite having what many would consider to be a fantastic soft. It's all about the numbers.

[sundevil77]

I think LSN has been pretty accurate for me. The one area I would be wary of looking at is scholarship numbers. So far, I would say I have underperformed with scholarships based on LSN. So much more goes into scholarship decisions than numbers, apparently.

[r6_philly]

Polls that fairly accurately predict the presidential election only poll 1200 people out of 120,000,000 voters.

This is misleading. If you use LSN to predict what percentage of applicants get admitted/denied vs the whole pool then LSN would be fairly representative. Because election polls only have 2 outcomes regardless of factors. It does not show the chance of being elected for any candidate who isn't being polled about. It isn't a good comparison.

Sorry for nitpicking

[thelawz09]

There are 768 U Chicago applicants with LSN profiles. Last year U Chicago had 5,400 applicants.

For Georgetown, the ratio is 1,296 to 11,649.

You get the picture. For most schools, the sample size is about 10% of the total number of applicants.

Two questions:

1) How is the pool of applicants who post their stuff on LSN different than the overall pool of applicants--in general? Younger? More obsessive? Worse grades? Better chances?
2) Does Law School Numbers provide a reasonably accurate measurement of what your chances are?

I think it gives you a general idea of admission trends. Of course it will miss the outliers, which is what I think lots of people are looking for. The one long shot that got in and means maybe they get in too. If anything LSN reaffirms what most of us already know; Above both medians is safety, between is target, and if you are below one 25% you need to be above the other 75% to have a fighting chance, and URM get different thresholds. I think for this cycle especially, last years graphs are going to be less useful, due to the massive increase in applications.

I think I'm ready to agree with you, jks. It sucks with outliers, it doesn't take in account softs. Other than that, it tells us what we already know.

In regard to who gets in where the answer isn't obvious, I think a lot of us forget how few seats we are talking about. We talk in percentages and percentiles, but when we talk about the "25th percentile" at U Chicago, for example, we are literally talking about 48 people. That's tiny.

[beesknees]

.

[PlugInBaby]

Also I think its useful in the fact that the numbers illustrate which schools YP and which don't (I'm looking at you American). That way if you get a shocking waitlisting, you can get an idea whether you screwed up or if the adcomms are just playing games.

[09042014]

Polls that fairly accurately predict the presidential election only poll 1200 people out of 120,000,000 voters.

This is misleading. If you use LSN to predict what percentage of applicants get admitted/denied vs the whole pool then LSN would be fairly representative. Because election polls only have 2 outcomes regardless of factors. It does not show the chance of being elected for any candidate who isn't being polled about. It isn't a good comparison.

Sorry for nitpicking

True, but I don't think it would even be very accurate at predicting how many get admitted vs denied because of heavy selection bias. The kind of person who finds and posts on LSN, knows which schools to apply to, and probably has a halfway good idea on how to put an app together.

[09042014]

I think LSN is a good tool for applicants to determine what their targets are (around medians), what their safeties are (above medians), and what their reaches are (below medians). Of course its not a sure-fire way to predict someone's cycle, but from my own experience, its pretty damn accurate (in that my cycle played out exactly as I expected it to). It also gives you a good ballpark for how much money you ***could*** get with your numbers (not that you will, because I know there are a lot of bitter people out on the Michigan thread).

Whoever brought up the soft factors, I'd just like to say that LSN doesn't reflect it because, unless you won a Nobel Peace Prize or something, neither will your cycle reflect your "awesome" softs that you think you have. Yes there are exceptions, but those people tend to be, well, exceptional. For the average applicant who was involved in college, LSN gives a good idea of what to expect.

It is a mistake to assume the only softs that matter are epic softs. A well written personal statement is often times the difference between a WL and an admit. It won't be a game changer, but at the margins those softs do matter. Because a school will pick 1/3 of those students and the margins, and since they have the same numbers there is something that is causing them to choose.

[DeSilentio2728]

It would be a whole lot more useful, at first glance, if they had one graph for urm's and one graph for everyone else.

[Rawlsian]

LSN has been completely accurate for me, at least after I update decisions.

[runn3rs]

LSN has been dead-on accurate for me. In my experience most, but not all, outliers are instate/outofstate differences or URM.

[yeff]

.
That being said, LSN is not extremely accurate. There are false data points, and it totally ignores most soft factors.

What LSN is useful for is determining where a schools cut offs are. You can look at UChi and see that below 3.4 there be dragons. You can also tell that below 3.7 and below 169 you are almost definitely out.

So look at the chart, and if you see a bunch of green near your number, you are probably in. If you see all red, you are probably out.

Where LSN is bad is the areas where there are acceptances waitlists, and rejections all in one area. If you see 5/10 applicants with your numbers getting in, that doesn't mean you have a 50-50 shot.

This.

It would be nice if the site could weed out obviously fake/joke profiles, and offer the ability to zoom in.

[ram jam]

That and the data points are not aligned correctly with Y axis for GPA... for example, a 3.1 will be displayed as a 3.05, 3.05 will be displayed as a 3.0, etc. Not a huge mistake though. I find LSN to be very accurate. Although, it is extremely biased in its sampling.