How Accurate is Law School Predictor? Forum

[mpasi]

I played around with it the other day and was pleasantly surprised with my results, albeit somewhat skeptical. What are the results based on?

[anmo]

A lot of people here think it is useful, though it is very hard to predict URM cycles. Also, soft factors play into decisions, even if minimally, and that can perhaps sway a "consider" to "strong consider"... Again though, LSP for URMs is anything but conclusive.

[TTTennis]

I played around with it the other day and was pleasantly surprised with my results, albeit somewhat skeptical. What are the results based on?

It is the most accurate predictor in the world. If it says it, that's what will happen. Want proof? It said I was "admit" at San Diego. Check. "Consider" at Minnesota. Check. They considered and decided to waitlist me. "Consider" at Illinois-UC. Check. They considered alright. Considered me not quite as good as some applicants and placed me on the waitlist. Correction: I placed myself on the waitlist

In all seriousness, I think it is pretty accurate. No idea what the results are based on. I'm guessing they took data from applicants with certain GPA/LSAT scores, and based on whether these applicants were accepted, rejected, or waitlisted at the different schools, created a formula that spit out a probability of acceptance, rejection or waitlist.

[twert]

i could be wrong but i think a lot of it is based on law school numbers. i think law school numbers is more helpful since you can look at individual cases. also, i'm a splitter and LSP was awfully optimistic

[Stringer Bell]

It's decent but not super accurate for splitters. I honestly don't remember the origination of the formula it uses for the LSAT-GPA weights for each school.

[Geist13]

Honestly, I don't think it's any more useful than looking at the number profiles of each school. Let's face it, when it comes to your "deny" and "admit" schools, LSN doesn't tell you anything you don't already know. It's potential use, as I see it, is for those schools that would fall within the "weak consider" to "strong consider" range. But in those cases, it's much more useful to know the admissions emphasis of each particular school. I have a 167 and only a 3.09. Despite LSP's generally negative outlook, I knew I would get in to WUSTL and Illinois and probably Minnesota just from seeing those monster LSAT walls on LSN. One could argue that it is inherently more difficult for LSP to predict splitters like myself. But if you're not a splitter, does LSN tell you anything you don't already know? If you're median at both %'s LSP will probably give you either a "consider" or a "strong consider," which is exactly what you were probably thinking, i.e. a good but not overwhelming chance. Just looking at the LSN graphs is, I think, the most useful way to see the general tendencies of certain schools which you can then apply to your particular application.

I do think LSN has a cool set up and is really fun to play around with using hypotheticals (I did this a ton before I had my LSAT score). I don't think it contributes anything new and its inability to handle splitters properly (both standard and reverse) makes it less useful than the LSN graphs.

[09042014]

I played around with it the other day and was pleasantly surprised with my results, albeit somewhat skeptical. What are the results based on?

Aren't you a URM splitter? If so it doesn't work well for you at all.

[TheSharklord]

I think it draws all of its data from LSN. It's the same as looking at the numbers yourself, with the provision that to do it as well as LSP you'd have to be able to look at all the data simultaneously, process decision trends, generate theoretical LSAT/GPA weights, etc. Which is hard to do with any unbiased accuracy.

Anyhow, like any statistical model, it's only as good as your data. And for URMs and heavy splitters, and goodness knows the intersection of the two (which I am), there's not as much data to support the conclusions because you're on the edge of the numbers.

Anyhow, while LSN may not be perfect for individual data points (w/ flames, fudged numbers, etc.) on the whole, there are a ton of data by now so the trends are fairly strong if you're a "standard" applicant...

My 2 cents.

[lawduder]

Based on my cycle thus far, applying early January, I've been accepted to four schools. Two of these schools were "admit" at >95%, and one was "admit" at 93%. The fourth school was "consider" at 58% where I've also been offered a scholarship that covers just a little over half of my total tuition.

[mpasi]

i could be wrong but i think a lot of it is based on law school numbers. i think law school numbers is more helpful since you can look at individual cases. also, i'm a splitter and LSP was awfully optimistic

I figured there wasn't much difference between the two, except that LSP may be using admissions stats from years past, whereas people on LSN can lie their asses off. I think LSP was overly optimistic with me too, but still pretty much accurate.

[pearl_earrings]

.

[mpasi]

Law School Predictor is a great resource and it's definitely helpful when you're deciding where to apply... but don't take its predictions as a sure thing. It wasn't completely accurate for my cycle (Penn waitlisted me, even though LSP said I was an "admit"). But, it provides predictions, not promises... and I think this has been a relatively unpredictable cycle.

Yeah, I'm using it to set up a list of schools to apply to. Like someone said a few posts up, it doesn't take URM status into account, nor does it consider soft factors. I think it's given me an excellent starting point, though.

[CyLaw]

I think it draws all of its data from LSN. It's the same as looking at the numbers yourself, with the provision that to do it as well as LSP you'd have to be able to look at all the data simultaneously, process decision trends, generate theoretical LSAT/GPA weights, etc. Which is hard to do with any unbiased accuracy.

Anyhow, like any statistical model, it's only as good as your data. And for URMs and heavy splitters, and goodness knows the intersection of the two (which I am), there's not as much data to support the conclusions because you're on the edge of the numbers.

Anyhow, while LSN may not be perfect for individual data points (w/ flames, fudged numbers, etc.) on the whole, there are a ton of data by now so the trends are fairly strong if you're a "standard" applicant...

My 2 cents.

It does not draw its data from LSN. The LSN data was only used for accuracy testing and refinement, AFTER the models were created.

http://www.lawschoolpredictor.com/?page_id=173

[ATOIsp07]

Law School Predictor is a crock. It said I had a less than 5% chance of being admitted to a T1 (WCL). Of course, I was admitted.

[TTTennis]

Law School Predictor is a crock. It said I had a less than 5% chance of being admitted to a T1 (WCL). Of course, I was admitted.

Geez, if you use logic and evidence like this in all your cases as a lawyer, I don't see you losing any case. Ever.

I plugged that prediction into LSP and it is giving me back "strong consider." Your future looks bright my friend

[TheSharklord]

I think it draws all of its data from LSN. It's the same as looking at the numbers yourself, with the provision that to do it as well as LSP you'd have to be able to look at all the data simultaneously, process decision trends, generate theoretical LSAT/GPA weights, etc. Which is hard to do with any unbiased accuracy.

Anyhow, like any statistical model, it's only as good as your data. And for URMs and heavy splitters, and goodness knows the intersection of the two (which I am), there's not as much data to support the conclusions because you're on the edge of the numbers.

Anyhow, while LSN may not be perfect for individual data points (w/ flames, fudged numbers, etc.) on the whole, there are a ton of data by now so the trends are fairly strong if you're a "standard" applicant...

My 2 cents.

It does not draw its data from LSN. The LSN data was only used for accuracy testing and refinement, AFTER the models were created.

http://www.lawschoolpredictor.com/?page_id=173

Law School Predictor is of far greater use to applicants when applicants can gauge the accuracy of LSP predictions. These results are based off of more than 33,500 Law School Numbers-listed law school admission decisions from the 2008-09 cycle for all ABA schools listed on both LSN and LSP. The version of LSP tested was Version 2.4. A big thanks goes out to TLS forum member CyLaw for compiling the data from LSN; I couldn’t have done it without CyLaw.

Not sure I follow. That pretty much reads as the LSN being the data source to me. Insofar as data and statistical modeling are inextricably tied together (with no data a model is as good as a theory with no proof, made without any natural observation--ie nothing at all), I think I disagree, unless I have a fundamental misunderstanding of how LSP works (which is always possible--say, "Weak Consider" =P)

[ATOIsp07]

Law School Predictor is a crock. It said I had a less than 5% chance of being admitted to a T1 (WCL). Of course, I was admitted.

Of course I was half-joking. But my personal example does go to show that, at least in my case, LSP is inaccurate. thereby, my ultimate point was that it is hardly an infallible predictor.

[mcdad]

LSP is as accurate as the statistics will allow. If it says Consider then you have a definite 50% chance, so it really hasn't read your future. It just gives you your chances.
Another source of error is yield protection. My daughter was admitted to one "admit" but waitlisted at a lower ranked "admit". A LOCI might deal with that.

[Notor]

Just look at the medians... if you are above both medians, admit. Below both, reject. Above one, below the other, you have a 50-50 shot depending on what they are favoring and how far off the median you are. LSP is fun, but looking at the medians is more accurate.

[ATOIsp07]

People must remember that LSP lacks accuracy for its failure to gauge chances with consideration to softs, particularly, URM status, instances of overcoming adversity, work experience, etc.

I certainly have never lent any of my faith to LSP.

[BioEBear2010]

People must remember that LSP lacks accuracy for its failure to gauge chances with consideration to softs, particularly, URM status, instances of overcoming adversity, work experience, etc.

I certainly have never lent any of my faith to LSP.

+1. Softs and essays matter a lot more than people on TLS realize.

[Reinhardt]

Pro: It's faster than LSN graphs.

Con: Speed is not really a concern. Use LSN graphs.

I predicted my own results based on LSN, and so far that's been a lot more accurate.

[TheSharklord]

People must remember that LSP lacks accuracy for its failure to gauge chances with consideration to softs, particularly, URM status, instances of overcoming adversity, work experience, etc.

I certainly have never lent any of my faith to LSP.

+1. Softs and essays matter a lot more than people on TLS realize.

+2 Here's to jimmying the odds in your favor! (*loves him some holistic review*)

[nyyankees]

iirc, when he was making the predictor he said he was using unique formulas from each school, not sure where he got the formulas from, but i was under the impression it wasnt entirely LSN based.

[lovethelaw88]

I think you have to take what it says with a grain of salt. So far, I've gotten into 9 schools and got WL'd at two. No rejects yet. Out of the schools I heard from, it told me I'd get into a few, but mostly it told me I'd be denied or "considered".