myLSN.info Forum

[LSATSCORES2012]

Popping in to say that I like what you're doing with the links on top of the myLSN page. It provides easy access to some of TLS's most useful information to people who might not otherwise be able to find it. As usual, great work bro.

Thanks If you (or anyone else) finds any useful links, please let me know and I'll add them.

[Granfalloon12]

Between recently stumbling upon this site, --LinkRemoved--, and http://www.lstscorereports.com/, I've become so inundated with and engrossed in law school admittance and statistics lately...

[LSATSCORES2012]

So, I recently added the quantifying boosts page, which calculated boosts based on a few data points which I had done by hand...

Since then, I integrated an automated calculator into the graphing script, and got values from thousands of data points in the T14 to try to approximate boosts. I've updated the page, which you can find at --LinkRemoved--

I explain the methodology there... it's worth noting that some of the sample sizes are very small. But I think that, using each school, the overall value should be relatively reasonable. But there are some values which are clearly false (ex, UVA definitely does not apply a 10 point subtraction to military members... this is probably just the result of too little data points)

I may see what I can do to refine the methods further... but I thought I'd provide the information now since it's interesting and may prove useful. Of course, take it with a grain of salt.

[LSATSCORES2012]

I was wondering the same thing... the problem, of course, is that to check that I would need to at least test a few groups together, and then the sample size is just going to be too small. For most of these groups the sample size already is too small.

But it's an interesting question that I wish I could answer...

ETA: For anyone wondering how both Male and Female applicants can have a 1 point boost, that's comparing them to the non-altered group - so it's basically just saying that people who specified their sex on LSN have this "boost" over those who don't. (Really, it's probably just a strange fluctuation... though perhaps those people who report their sex on LSN are generally more thorough and so had better applications? Dunno...)

[LSATSCORES2012]

Another update...

I just added --LinkRemoved--

Basically, it graphs the change in admission standards between the 10-11 and 11-12 cycle (just simply subtracting the percent admitted in 10-11 from the percent admitted in 11-12; if it's positive, then it's green (the more positive the more green) and if it's negative then it's red (the more negative the more red).

Interesting, because it helps to show the strategies that schools are employing to try to maintain their numbers (mostly, it seems, becoming more numbers-focused in admissions: it seems to me that you can tell this when you have a green area right around and slightly below the median, with a huge red area right below that).

[florida1949]

Between recently stumbling upon this site, --LinkRemoved--, and http://www.lstscorereports.com/, I've become so inundated with and engrossed in law school admittance and statistics lately...

+infinity

[sinfiery]

Hey everybody, I have a request.

If anyone here has used mylsn.info (LinkRemoved) and it has proven to tremendously aid you in the law school admissons process, please consider voting for LSATSCORES2012 in this poll. He is the creator of this tool and because of the degree to which it has aided me in my application process, I feel he should without a doubt be the winner of the tittle, "Most helpful poster" for 2012. If you agree with me, please vote for him too.


Thanks!

[Ti Malice]

Many thanks for the terrific tool you've constructed.

I have one suggestion for improvement: It would be very helpful if the impact of submission date could be viewed on a much more granular level instead of in the three broad categories of "early," "in between," and "late." For one thing, the endpoints of these categories are too far apart to really deserve being placed in the same category. For many schools, there is a significant difference in odds for an identical app being submitted on September 1 versus November 30, or on December 1 versus January 31. (As an issue purely of nomenclature, I think it would be good to dispense with the category names as well; schools' admissions timelines are substantially different from one another, such that "in between" -- the equivalent of "on time" -- at one school is much more like "late" at another.)

An application submitted to Berkeley between mid- and late November, for example, is actually late, given that a substantial majority of admitted applicants in a given cycle at Berkeley are admitted from late November through the third week in December. An app submitted to Michigan on December 1, on the other hand, can be considered on-time, since a large proportion of applicants admitted to Michigan are admitted from Christmas through the middle of March. An app submitted to Chicago on November 20 is at the late end of "timely," but one submitted on December 15 (meaning it will likely not go complete until January) is likely very late, given that the vast majority of applicants admitted to Chicago are traditionally admitted in the middle of January.

Not being able to view submission times with greater precision makes it impossible to discern more precise patterns at individual schools. It would be helpful if the default setting were "any time," as it is now, but with the user also being able to select individual boxes for different months, as with the application years. Even more useful would be breaking the months of November, December, and January -- the months in which small differences in submission date (like first week versus last week in November) can make a substantial impact on admissions chances at many schools -- into sections of 10 to 11 days each.

Thanks again!

[LSATSCORES2012]

Many thanks for the terrific tool you've constructed.

I have one suggestion for improvement: It would be very helpful if the impact of submission date could be viewed on a much more granular level instead of in the three broad categories of "early," "in between," and "late." For one thing, the endpoints of these categories are too far apart to really deserve being placed in the same category. For many schools, there is a significant difference in odds for an identical app being submitted on September 1 versus November 30, or on December 1 versus January 31. (As an issue purely of nomenclature, I think it would be good to dispense with the category names as well; schools' admissions timelines are substantially different from one another, such that "in between" -- the equivalent of "on time" -- at one school is much more like "late" at another.)

An application submitted to Berkeley between mid- and late November, for example, is actually late, given that a substantial majority of admitted applicants in a given cycle at Berkeley are admitted from late November through the third week in December. An app submitted to Michigan on December 1, on the other hand, can be considered on-time, since a large proportion of applicants admitted to Michigan are admitted from Christmas through the middle of March. An app submitted to Chicago on November 20 is at the late end of "timely," but one submitted on December 15 (meaning it will likely not go complete until January) is likely very late, given that the vast majority of applicants admitted to Chicago are traditionally admitted in the middle of January.

Not being able to view submission times with greater precision makes it impossible to discern more precise patterns at individual schools. It would be helpful if the default setting were "any time," as it is now, but with the user also being able to select individual boxes for different months, as with the application years. Even more useful would be breaking the months of November, December, and January -- the months in which small differences in submission date (like first week versus last week in November) can make a substantial impact on admissions chances at many schools -- into sections of 10 to 11 days each.

Thanks again!

Good idea, thanks for the input! Will need to take some time thinking about the best way to implement this.

[LSATSCORES2012]

In response to something someone posted in another thread, I made http://mylsn.info/graph_money.php

Basically it's a scholarship graph. The more solid the point the greater the scholarship (where a fully solid point is a $150,000 scholarship or greater). You can also see the exact average at a point (and in parentheses is the number of applications taken into consideration) by hovering.

Note that this also includes data from people who entered their scholarship values as the value per year (ex, 22.5k instead of $67.5k). Also, it discards all values greater than $200,000, and for anything less than four digits long it adds three zeroes to the end (ex: 200 -> 200000, 5 -> 5000.

Really a beta feature, let me know if you have any trouble/ideas to make it more helpful.

[manofjustice]

very cool site.

[Yukos]

Don't think there's any way to fix this, but I realized YP is almost certainly underestimated on your site. If YP waitlists don't come out until April but HYS are accepting Dec-March, I'm sure significant numbers of people are withdrawing from UVA, Penn or whoever else while their app is still "pending." Thus they're never officially marked as waitlist on LSN, throwing off mylsn's percentages. Just a thought.

[Yukos]

Don't think there's any way to fix this, but I realized YP is almost certainly underestimated on your site. If YP waitlists don't come out until April but HYS are accepting Dec-March, I'm sure significant numbers of people are withdrawing from UVA, Penn or whoever else while their app is still "pending." Thus they're never officially marked as waitlist on LSN, throwing off mylsn's percentages. Just a thought.

I think that someone withdrawing before they get YP WL'd and thus it not showing up as part of the WL stats is the best case scenario. If I'm trying to figure out my chances at a reach school, I'd like to not have to parse through the idea that some of the WLs at that institution are YPs.

Good point, but the data are already confounded by YP (Applicant has twice as good shot to get into Harvard as UVA) so I think it's worth pointing out that if you are in the YP range, your chances of being YP'd are probably higher than mylsn implies.

ETA: Just realized that the problem you bring up is probably minimal. If a school is a reach than WLs from that school for people with your numbers are, by definition, not YP.

[Yukos]

Don't think there's any way to fix this, but I realized YP is almost certainly underestimated on your site. If YP waitlists don't come out until April but HYS are accepting Dec-March, I'm sure significant numbers of people are withdrawing from UVA, Penn or whoever else while their app is still "pending." Thus they're never officially marked as waitlist on LSN, throwing off mylsn's percentages. Just a thought.

I think that someone withdrawing before they get YP WL'd and thus it not showing up as part of the WL stats is the best case scenario. If I'm trying to figure out my chances at a reach school, I'd like to not have to parse through the idea that some of the WLs at that institution are YPs.

Good point, but the data are already confounded by YP (Applicant has twice as good shot to get into Harvard as UVA) so I think it's worth pointing out that if you are in the YP range, your chances of being YP'd are probably higher than mylsn implies.

ETA: Just realized that the problem you bring up is probably minimal. If a school is a reach than WLs from that school for people with your numbers are, by definition, not YP.

Yes but lets say that I have Mich numbers and I'm trying to figure out whether or not its worth throwing an app at NYU. Every YP WL that gets calculated into my chances discourages me a little more. Everyone who's numbers are below 3.9/174 has reach schools so I dont think this is minimal.

NYU doesn't YP so let's say you have GULC numbers (170/3.6-3.7 I guess) and are thinking about applying to UVA. When you search your own numbers on mylsn any UVA WLs will NOT be due to YP because UVA doesn't YP 170/3.65 people. If something is a reach for you then you're beneath at least one if not both medians so when you search mylsn for those numbers you won't see any YPs, only "legit" WLs.

[LSATSCORES2012]

First I want to say that I apologize to everyone who has been using Internet Explorer - I just installed Windows on my computer and thought I'd see how well it worked... and, well, lots of stuff didn't. But it should work now, so if you experience any issues please let me know. (Also, I'd advise switching your browser to something more modern.

To Yukos and Wormfather, I think you both raise good points. I think there are lots of problems in general with the data on LSN, and this is one of them, but while half or even more of the people who would be YP'ed at a certain school will withdraw and, thus, not have the would-be waitlist recorded, there are still a few people who will (as we have seen) and, thus, there should still be a record of the YP occuring. But what you say is particularly interesting because the YP could be significantly worse than one would expect based upon LSN.

In general, I think the graphs are a much better way to examine one's chances at a school than the search feature, because with the graphs you can see if the point you're searching is just some sort of statistical anomaly, and you can see how much of a buffer you have. By that, I mean that if you are a green point, and you have a bunch of green below (and to the left) of you, then you're safe. On the other hand, if you're a green point, but you're just above a sea of red, your chances might be heavily dependent upon the cycle you're applying in and maybe even your softs (crazy, right? )

In any case, that's a bit besides the point. You have a great point, and I think it really serves to show one of the many ways in which the meaning of the information on LSN, and thus MyLSN, is obscured. It's important to bring things like this up, though, because it does create the possibility of someone coming up with some sort of way to avoid the problem (though, unfortunately, like you suggested, I, at the very least, can't think of a way to solve this one).

I think this is my longest post on TLS...

Hope you all are having a great weekend!

[suralin]

First I want to say that I apologize to everyone who has been using Internet Explorer - I just installed Windows on my computer and thought I'd see how well it worked... and, well, lots of stuff didn't. But it should work now, so if you experience any issues please let me know. (Also, I'd advise switching your browser to something more modern.

To Yukos and Wormfather, I think you both raise good points. I think there are lots of problems in general with the data on LSN, and this is one of them, but while half or even more of the people who would be YP'ed at a certain school will withdraw and, thus, not have the would-be waitlist recorded, there are still a few people who will (as we have seen) and, thus, there should still be a record of the YP occuring. But what you say is particularly interesting because the YP could be significantly worse than one would expect based upon LSN.

In general, I think the graphs are a much better way to examine one's chances at a school than the search feature, because with the graphs you can see if the point you're searching is just some sort of statistical anomaly, and you can see how much of a buffer you have. By that, I mean that if you are a green point, and you have a bunch of green below (and to the left) of you, then you're safe. On the other hand, if you're a green point, but you're just above a sea of red, your chances might be heavily dependent upon the cycle you're applying in and maybe even your softs (crazy, right? )

In any case, that's a bit besides the point. You have a great point, and I think it really serves to show one of the many ways in which the meaning of the information on LSN, and thus MyLSN, is obscured. It's important to bring things like this up, though, because it does create the possibility of someone coming up with some sort of way to avoid the problem (though, unfortunately, like you suggested, I, at the very least, can't think of a way to solve this one).

I think this is my longest post on TLS...

Hope you all are having a great weekend!

Damn, impressive work. I really really like the graphs , it's such a good way of seeing where different combinations of numbers end up.

One question though: after completing a search, where did the stat go that showed how applications with higher/lower indexes do? I thought that was super useful. If it's still there somewhere, I apologize for being a dumbass.

Thanks for all your work man!

[North]

Another fantastic update with the 'more info' feature. Great idea. It used to be such a pain to download the Excel sheet to use for profile hunting on LSN (but still far less frustrating than using LSN's awful search function, of course).

And put some ads on there or something. It's time you made FAT STACKS off your 20,000 unique visitors. LSP certainly wasn't bashful about that.

[facile princeps]

Another fantastic update with the 'more info' feature.

+1

That "more info" feature was a golden touch.

[LSATSCORES2012]

One question though: after completing a search, where did the stat go that showed how applications with higher/lower indexes do? I thought that was super useful. If it's still there somewhere, I apologize for being a dumbass.!

You're right, I did take it down, but I can put it back up... I'll try to get it done this weekend.

Another fantastic update with the 'more info' feature. Great idea. It used to be such a pain to download the Excel sheet to use for profile hunting on LSN (but still far less frustrating than using LSN's awful search function, of course).

And put some ads on there or something. It's time you made FAT STACKS off your 20,000 unique visitors. LSP certainly wasn't bashful about that.

Another fantastic update with the 'more info' feature.

+1

That "more info" feature was a golden touch.

Credit where credit is due - this was Mr. Elshal's idea. Glad you like it

[LSATSCORES2012]

One question though: after completing a search, where did the stat go that showed how applications with higher/lower indexes do? I thought that was super useful. If it's still there somewhere, I apologize for being a dumbass.

Alright, this is back up now.

And put some ads on there or something. It's time you made FAT STACKS off your 20,000 unique visitors. LSP certainly wasn't bashful about that.

Your wish is my command I really am thankful for all the support TLS has given to the website

[suralin]

One question though: after completing a search, where did the stat go that showed how applications with higher/lower indexes do? I thought that was super useful. If it's still there somewhere, I apologize for being a dumbass.

Alright, this is back up now.

And put some ads on there or something. It's time you made FAT STACKS off your 20,000 unique visitors. LSP certainly wasn't bashful about that.

Your wish is my command I really am thankful for all the support TLS has given to the website

Awesome! Thanks for all your work!

And I agree, it's about time you made some $$$ off this.

[LSATSCORES2012]

Also, just added a page that compares complete date with decision date for all the schools on which there is data. It uses the google charts API, which wouldn't let me switch the axes into a more intuitive way, but I still think they're readable/helpful.

See: http://myLSN.info/datevdate.php

[LSATSCORES2012]

To whoever sent me the feedback, that was a glitch and should be fixed now. Please let me know if you have any other troubles

[LSATSCORES2012]

Latest updated data is labeled 2012-2013 (testing) - if you notice any problems, please let me know! It's using the latest data from this cycle, but I don't want to completely replace the 2012-2013 until I'm sure the new parser is working.

[violettre]

Wow this is super useful. Thanks for putting this together.