Why do people say 1L performance is unpredictable? Forum

[A. Nony Mouse]

you realize you're talking about the separation b/w the #1 and #last person in a class, not a median? you're speaking about 4 people total, and extracting from the 4 something about hundreds of people? It's very possible the #1 kid in that class had an amazing test, and the #100 kid drew a smily face (I'm assuming a smily face was not the correct answer).

Sure, but a 16-point spread is still really narrow (yes, unless out of 16 points) to distribute a 1L curve over. The point is probably that it's hard to write an exam that distributes grades over a broader range, because at a decent school everyone does know everything so there isn't much to distinguish the different exams. But a professor should be able to write such an exam.

[star fox]

You should work really hard and there are more stealthily lazy people than they appear. But in general, you don't know how much it will actually move the needle.

[Bilb0Baggins]

These #'s suggest that LSAT / UGPA is a somewhat reliable predictor of 1L performance, no?

http://www.lsac.org/docs/default-source ... -11-02.pdf

Which granted is not a guarantee -- but do suggest the chances of top performance are even higher if one attends a school where one is >75th GPA and >75th LSAT, yes?

[AReasonableMan]

These #'s suggest that LSAT / UGPA is a somewhat reliable predictor of 1L performance, no?

http://www.lsac.org/docs/default-source ... -11-02.pdf

Which granted is not a guarantee -- but do suggest the chances of top performance are even higher if one attends a school where one is >75th GPA and >75th LSAT, yes?

It's a bad link, but yes. There is SOME predictive value, but it's like using 40 times and benchpress reps as a measure of NFL performance. There's some value, but their actual rookie performance is always a 10x better predictor.

[chuckbass]

These #'s suggest that LSAT / UGPA is a somewhat reliable predictor of 1L performance, no?

http://www.lsac.org/docs/default-source ... -11-02.pdf

Which granted is not a guarantee -- but do suggest the chances of top performance are even higher if one attends a school where one is >75th GPA and >75th LSAT, yes?

It's a bad link, but yes. There is SOME predictive value, but it's like using 40 times and benchpress reps as a measure of NFL performance. There's some value, but their actual rookie performance is always a 10x better predictor.

There may be some correlation, but I wouldn't bet 3 years and X amount of $$ on that. Of the people I know, most either overpeformed or underperformed their numbers.

[AReasonableMan]

These #'s suggest that LSAT / UGPA is a somewhat reliable predictor of 1L performance, no?

http://www.lsac.org/docs/default-source ... -11-02.pdf

Which granted is not a guarantee -- but do suggest the chances of top performance are even higher if one attends a school where one is >75th GPA and >75th LSAT, yes?

It's a bad link, but yes. There is SOME predictive value, but it's like using 40 times and benchpress reps as a measure of NFL performance. There's some value, but their actual rookie performance is always a 10x better predictor.

There may be some correlation, but I wouldn't bet 3 years and X amount of $$ on that. Of the people I know, most either overpeformed or underperformed their numbers.

right but if Person A is above both then they are likely betting significantly less money. it certainly makes more sense than Person B who is below both #'s betting twice as much that they will outdo Person A. Person A has a better (albeit not much better) than 50% chance of > Person B.

[star fox]

These #'s suggest that LSAT / UGPA is a somewhat reliable predictor of 1L performance, no?

http://www.lsac.org/docs/default-source ... -11-02.pdf

Which granted is not a guarantee -- but do suggest the chances of top performance are even higher if one attends a school where one is >75th GPA and >75th LSAT, yes?

It's a bad link, but yes. There is SOME predictive value, but it's like using 40 times and benchpress reps as a measure of NFL performance. There's some value, but their actual rookie performance is always a 10x better predictor.

There may be some correlation, but I wouldn't bet 3 years and X amount of $$ on that. Of the people I know, most either overpeformed or underperformed their numbers.

Dual 75s = full ride usually

[Johann]

why dont we get the sealocust in here to brag about his 1Ls did? ( he had a class of 20 - i predict 7 or 8 are below median. )

[utahraptor]

why dont we get the sealocust in here to brag about his 1Ls did? ( he had a class of 20 - i predict 7 or 8 are below median. )

this seems misplaced and really petty

[BVest]

[responding to 16-point spread from best exam to worst exam]
That means your prof wrote a terrible exam for a curved class. That should never be the case, especially for a large section class of 60-80 students. The problem is that there's not enough range to allow the differences in scores to fall along the distribution that the prof is required to use.

My civ pro exam, OTOH, had a total of 200-something points. The high grade in the class (A) was 155 and the low grad (C) was a 67.

you're asusming both tests were outta 200 pts

No, I'm not. I assume that the previous exam was out of 100 actually. It's still not enough separation for a large-section curved class, and makes the results -- while not random -- closer to random than they reasonably should be.

you realize you're talking about the separation b/w the #1 and #last person in a class, not a median? you're speaking about 4 people total, and extracting from the 4 something about hundreds of people? It's very possible the #1 kid in that class had an amazing test, and the #100 kid drew a smily face (I'm assuming a smily face was not the correct answer).

What? How am I speaking about 4 people? And what in the world makes you think that I don't know whether I'm talking about #1 to median vs #1 to #last?

[MKC]

why dont we get the sealocust in here to brag about his 1Ls did? ( he had a class of 20 - i predict 7 or 8 are below median. )

this seems misplaced and really petty

I'm actually kind of curious if TSL figured out how to teach law schooling effectively.

[Johann]

why dont we get the sealocust in here to brag about his 1Ls did? ( he had a class of 20 - i predict 7 or 8 are below median. )

this seems misplaced and really petty

I'm actually kind of curious if TSL figured out how to teach law schooling effectively.

yeah not trying to be petty - im curious

[UnicornHunter]

why dont we get the sealocust in here to brag about his 1Ls did? ( he had a class of 20 - i predict 7 or 8 are below median. )

this seems misplaced and really petty

I'm actually kind of curious if TSL figured out how to teach law schooling effectively.

yeah not trying to be petty - im curious

+1 on the curiosity. Would be highly relevant to TT.

[star fox]

Its tough to say people did well because they paid money for a class on how to take law school exams or if they would have done well anyways and the kind of people to sign up for that kind of thing are the ones more likely to gun their way above median anyways. Its still the same "moving the needle" question.

[kcdc1]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed open-book issue-spotter?

[fats provolone]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed issue-spotter?

wut

[Johann]

Its tough to say people did well because they paid money for a class on how to take law school exams or if they would have done well anyways and the kind of people to sign up for that kind of thing are the ones more likely to gun their way above median anyways. Its still the same "moving the needle" question.

yeah we know its not science. people neurotic enough about grades to pay another 2k about them are clearly neurotic as hell. but still, its about the best data we can get that is easily accessible.

[utahraptor]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed open-book issue-spotter?

#teamgradesareprettyrandom checking in.

I've never done that and I've never scored below median.

but, uh, cool strategy, interested in hearing how that works

[chuckbass]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed open-book issue-spotter?

#teamgradesareprettyrandom checking in.

I've never done that and I've never scored below median.

but, uh, cool strategy, interested in hearing how that works

I think this just falls into different things work for different people, and it's then all dependent on your professor. I used some presets for torts which helped a little just on timing I'd think, and helped given the way my professor wanted the questions to be answered. If you can tell, there's still a lot of luck/randomness involved there.

[kcdc1]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed issue-spotter?

wut

(1) Assume every issue covered in the class will be in the exam
(2) Write out the rules governing issue #1
(3) Write out how Plaintiff argues issue #1 with blanks for relevant facts
(4) Write out how Defendant argues issue #1 with blanks for relevant facts
(5) Leave a space for retort and conclusion
(6) Repeat steps (2)-(5) until you've written an argument shell for every issue in the class
(7) Read exam fact pattern carefully, insert relevant facts while retyping argument shell from outline

[star fox]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed open-book issue-spotter?

I think it's better to just make flow charts and checklists for specific types of problems and get better at using them on practice exams. You'll probably spend a lot of time adjusting and revising your pre written answers to the specific fact pattern anyways.

[chuckbass]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed issue-spotter?

wut

(1) Assume every issue covered in the class will be in the exam
(2) Write out the rules governing issue #1
(3) Write out how Plaintiff argues issue #1 with blanks for relevant facts
(4) Write out how Defendant argues issue #1 with blanks for relevant facts
(5) Leave a space for retort and conclusion
(6) Repeat steps (2)-(5) until you've written an argument shell for every issue in the class
(7) Read exam fact pattern carefully, insert relevant facts

So were you number 1 in your class? Cuz if not I don't see how much of a point you can prove.

[jbagelboy]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed open-book issue-spotter?

#teamgradesareprettyrandom checking in.

I've never done that and I've never scored below median.

but, uh, cool strategy, interested in hearing how that works

I got my worst grade since starting law school (below median) in a class where I pre-wrote exam answers in the manner you described. Turned out the prof didn't read the essays and graded the course based on a multiple choice exam unlike anything we'd ever seen.

You never know.

[Johann]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed open-book issue-spotter?

this seems like a dumb strategy in my opinion. the goal is to be flexible and able to go where the question takes you. having a preconceived answer i think will kinda force you in the wrong way. plus you can do much better stuff with that time.

[UnicornHunter]

Honest question for the "Grades Are Random" crowd:

How often have you devoted a solid 40 hours purely to pre-writing an exam answer, complete with rule statements and pivots to sub-issues, and still scored below median on a timed issue-spotter?

wut

(1) Assume every issue covered in the class will be in the exam
(2) Write out the rules governing issue #1
(3) Write out how Plaintiff argues issue #1 with blanks for relevant facts
(4) Write out how Defendant argues issue #1 with blanks for relevant facts
(5) Leave a space for retort and conclusion
(6) Repeat steps (2)-(5) until you've written an argument shell for every issue in the class
(7) Read exam fact pattern carefully, insert relevant facts while retyping argument shell from outline

That's basically just an inefficient way to make an outline. Every outline should hit all the major issues and sub-issues anyway. I don't see the value-added from pre-writing stuff you're going to have to sub out anyway.

I will admit, some issues lend themselves to this approach. In contracts, you know you're going to have a battle of the forms question. Makes sense to queue up your arguments. Still, no reason you can't do that in the context of an outline.