Law School Predictor: The Thread" Forum

[lawlover829]

thank you

[realworldescapee]

Great Program: I plugged in my stats (currently prospective) and five friends who are now in law school and it was 22 for 23.

Some points to improve on:

I've had my browser crash on me when I try and change the GPA before the 160/0.00 completely loads. A warning to let the page fully load should be added to help prevent this from occurring. The issue has happened multiple times. I use the latest version of Firefox and and the latest version of Mac OS X 10.5 Leopard (should it matter).

I'd also reorganize the columns with the more applicant desirable stats to the left. Maybe I'm wrong, but I personally don't care (in the grand scheme of things) what a school's LSAT/GPA ratio is. I do care what my index score is, what the 25th, 50th and 75th percentile scores and how far my index score is from the aforementioned percentiles. I'd personally arrange the columns (after city) as follows: Score, From Mid, From 25, 25th score, median score, 75th score, %+/-, You %, Mid %, From 75, From Perfect, LSAT/GPA ratio, raw data for LSAT and GPA. IMO it organizes the info with categories.

A little nitpicky, but awesome program.

[YCrevolution]

..

[YCrevolution]

..

[YCrevolution]

..

[caoyun]

Update complete. Law School Predictor: Version 2.1 is now live at LawSchoolPredictor.com.

If only it were accurate. Admit at NYU? Strong consider at Columbia? If only. Nice program, though. It was right from everything Chicago and down.

[coolkatz321]

UIUC places a lot of emphasis on the GPA...most people with my stats were either admitted or WL. Maybe tweak that a bit?

[YCrevolution]

..

[FjornF]

Excellent program, I am sure this will save many people alot of time.

A couple of thoughts:
1. Create a normalized score column. Maybe something like 0 if score is less than 25th, 1 if above 75th (corresponding to deny/admit) and then linear in between. This would be alot more useful than "strong/weak" considers.

2. Release the XLSX file. Yes someone could repost it to their site, or whatever, but now that you are first to post and have the backing of TLS I think it is pretty safe. Being protective of the source only obscures its accuracy and usefulness.

[MlhopeTC]

This thing is great, thanks for your work!

[YCrevolution]

..

[bakemono]

besides being a bit pessimistic about splitters with low GPA and high LSAT, it's the most functional and accurate predictor thus far. I haven't even applied yet, but I've looked at the stats religiously, and this seems to have great application. Your nerdiness is MUCHO appreciated.

[YCrevolution]

..

[Top-Notch]

Not very accurate from what I can tell.

[YCrevolution]

..

[Helmholtz]

I'm a weak splitter, but I definitely noticed a difference in predictions (Michigan and Duke moving up to "admit" status and UVA moving up to "strongly consider" status). I'm really hoping the newest formula is better than the old one (purely out of self-interest).

[YCrevolution]

..

[realworldescapee]

YCrev,

Someone pointed this out to me after they toyed around with the median numbers and index formulas, and noticed that if you entered the median LSAT/GPA numbers they don't match up with the 50th percentile index score. For instance they had a 163/3.65 stats and were given an index score above the 50th percentile or UIUC. But from my general impression, they would be a weak consider at this school. True the GPA is on the higher end of the school, but the LSAT is close to the 25th percentile and 3 points below the median. Moreover, if you input Illinois median GPA/LSAT numbers, into the index formula, that score is higher than your stated median index score. I didn't have an answer to the question why the UIUC median socres resulted in a higher 50th percentile index score than you reported. So i'm wondering what the explanation is for this result. Personally, I wouldn't have given them a weak chance at UIUC given the LSAT score.

UIUC fyi was just one school that they gave me that this occurred with.

Thanks

[YCrevolution]

..

[Mr. Matlock]

Hey ycrev!!

Hopefully you'll see a bump in site activity. We've been pushing it in the LSAT score thread. It really is, imho, the cream of the crop for admitting data and how it's presented.

Thanks again for all of your hard work.

[YCrevolution]

..

[Ken]

Thanks Evan, it is looking great!!

[bluejayk]

wow, this is cool, it's picked up my spirits a bit after receiving a somewhat disappointing LSAT score. Thanks for all the work!

[James Bond]

Thanks Evan, it is looking great!!

Agreed, I've recently become addicted in predicting my future admissions cycle. I've formed a spreadsheet comprised of schools I like as far as their location and environment, coupled with possible LSAT/GPA combinations, and charted my likelihood of admission depending on how I perform my junior year and on the LSAT. Call me a dork, but I like to know what to aim for and what my options are:

--ImageRemoved--

[FjornF]

Excellent program, I am sure this will save many people alot of time.

A couple of thoughts:
1. Create a normalized score column. Maybe something like 0 if score is less than 25th, 1 if above 75th (corresponding to deny/admit) and then linear in between. This would be alot more useful than "strong/weak" considers.

2. Release the XLSX file. Yes someone could repost it to their site, or whatever, but now that you are first to post and have the backing of TLS I think it is pretty safe. Being protective of the source only obscures its accuracy and usefulness.

1. I'll look into it. I know of a couple others who expressed a similar wish although there are others that also didn't want a prediction that involved numbers. If I implement it, I'll probably put it alongside the textual predictions.

Users can also see how far away they are from 25/75 right now. I hesistate slightly to quantify the prediction which would suggest a precision that probably isn't there. I'd also need to figure out a way to present it so that people don't misinterpret the predictive number.

2. I don't maintain a master XLS file anymore (unlike previous versions, instead I have multiple files, which aren't nearly as structured as the HTML files), but I understand what you're saying. The method(s) used to calculate the predictions is already there basically (in the About page) and the published admission index formulas can be found on the LSAC website. I can publish the formulas separately on the website sometime if people want to see them. I don't include them right now in the spreadsheet just because it slows down loading times, and most people don't seem to care what the formula is.

In the meantime, I'm (slowly) considering the logisitics of a beta-testing program where qualified TLS forum members will get advance access to new versions of Law School Predictor (in XLS and HTML format), and then can provide feedback and point out any errors. I think I'd be looking for people who have been on TLS "a while" and who have a large number of posts or have already contributed some significant time/energy to helping me develop Law School Predictor.

Awesome, it is exciting to see the rate at which you are improving this thing. It really is a wonderful resource. Hope that facebook advertising works out.