Top Law Schools

I know the median is around 3.2, does anyone know where 3.4 is on the curve?
And any bidding advise for NYC?

My guess would be top 35%.

i know for a fact that 1L median is ~3.2

G. T. L. Rev. wrote:

Anonymous User wrote:I know the median is around 3.2, does anyone know where 3.4 is on the curve?
And any bidding advise for NYC?

Michigan clerkship applicants usually include a letter from MLS which provides average cutoffs from the past five years. According to the current version of this letter, top 50% at graduation is, on average, 3.369. Median after 1L year is almost certainly lower, such that a 3.4 after 1L year probably puts you closer to top 30-40% or so.

Anonymous User wrote:i know for a fact that 1L median is ~3.2

G. T. L. Rev. wrote:

Anonymous User wrote:I know the median is around 3.2, does anyone know where 3.4 is on the curve?
And any bidding advise for NYC?

Michigan clerkship applicants usually include a letter from MLS which provides average cutoffs from the past five years. According to the current version of this letter, top 50% at graduation is, on average, 3.369. Median after 1L year is almost certainly lower, such that a 3.4 after 1L year probably puts you closer to top 30-40% or so.

How do you know this? (Question to Anonymous). Isn't the target mean 3.19 (but the actual number floats closer to 3.25)? The upperclass mean is usually around 3.45, so yeah, having a 3.XX after 1L would certainly be a better percentile rank than if you had that same GPA at graduation.

career service said so.

Helmholtz wrote:

Anonymous User wrote:i know for a fact that 1L median is ~3.2

G. T. L. Rev. wrote:

Anonymous User wrote:I know the median is around 3.2, does anyone know where 3.4 is on the curve?
And any bidding advise for NYC?

Michigan clerkship applicants usually include a letter from MLS which provides average cutoffs from the past five years. According to the current version of this letter, top 50% at graduation is, on average, 3.369. Median after 1L year is almost certainly lower, such that a 3.4 after 1L year probably puts you closer to top 30-40% or so.

How do you know this? (Question to Anonymous). Isn't the target mean 3.19 (but the actual number floats closer to 3.25)? The upperclass mean is usually around 3.45, so yeah, having a 3.XX after 1L would certainly be a better percentile rank than if you had that same GPA at graduation.

I know that we broke the curve last semester (3.30 median/mean (can't remember) resulted from a 3.19 target and a 3.25 maximum) so that might affect things. Still, very unlikely employers would know about this unless they're regular readers of Res Gestae

goodolgil wrote:I know that we broke the curve last semester (3.30 median/mean (can't remember) resulted from a 3.19 target and a 3.25 maximum) so that might affect things. Still, very unlikely employers would know about this unless they're regular readers of Res Gestae

I'm about 95% sure that they later admitted that info was wrong due to some miscalculation, but I could be wrong.

goodolgil wrote:I know that we broke the curve last semester (3.30 median/mean (can't remember) resulted from a 3.19 target and a 3.25 maximum) so that might affect things. Still, very unlikely employers would know about this unless they're regular readers of Res Gestae

Where did you hear this? I was struggling to figure out how the average 1L GPA could have possibly been 3.30 when not a single class had an average GPA that high. The grading curve on the registrar's website still says 3.30 for fall 2010, though.

Anonymous User wrote:

goodolgil wrote:I know that we broke the curve last semester (3.30 median/mean (can't remember) resulted from a 3.19 target and a 3.25 maximum) so that might affect things. Still, very unlikely employers would know about this unless they're regular readers of Res Gestae

Where did you hear this? I was struggling to figure out how the average 1L GPA could have possibly been 3.30 when not a single class had an average GPA that high. The grading curve on the registrar's website still says 3.30 for fall 2010, though.

As I alluded to above, I think the registrar reported the wrong number (although I don't think they've gotten around to changing the official documents), which caused Res Gestae to report the wrong number. I've heard secondhand that the registrar confirmed the mistake, and I believe something was printed in the hard copy of Res Gestae mentioning the error.

Helmholtz wrote:

Anonymous User wrote:

goodolgil wrote:I know that we broke the curve last semester (3.30 median/mean (can't remember) resulted from a 3.19 target and a 3.25 maximum) so that might affect things. Still, very unlikely employers would know about this unless they're regular readers of Res Gestae

Where did you hear this? I was struggling to figure out how the average 1L GPA could have possibly been 3.30 when not a single class had an average GPA that high. The grading curve on the registrar's website still says 3.30 for fall 2010, though.

As I alluded to above, I think the registrar reported the wrong number (although I don't think they've gotten around to changing the official documents), which caused Res Gestae to report the wrong number. I've heard secondhand that the registrar confirmed the mistake, and I believe something was printed in the hard copy of Res Gestae mentioning the error.

I remember hearing the same thing. I am almost positive the 3.30 was a mistake.

I think that 3.4 is likely to be around top third at this point.

3.4 is much higher than a third. You can roughly convert gpa to percentiles if you have some basic stats knowledge.

FYI, at Mich 3.15 to 3.25 is the mean target not median. There is a substantial difference between the two measures. For instance, if a professor gives out 10 As and 5 A+s, the class will have a higher mean than if the professor gave out 15 As but the median hasn't moved. This is very important understanding individual class grading guidelines:

--LinkRemoved--

Note how for instance a professor can give out A+s to anywhere from 0-3% of the class and A-s to anywhere from 13-17% of the class. This can have the effect of significantly pulling up the mean GPA for a class without affecting medians, particularly if the top half is substantially better than in previous years and the bottom half of the curve looks like previous years. A 3.3 mean is therefore entirely plausible without affecting medians.

These percentages allow you to roughly map out a continuous bell curve and calculate your percentile using a markov model. I did this using excel. It's not perfect and assumes that across your 6 classes, you average out to the target percentages in that pdf. But to give you an idea, 3.4 will put you at anywhere from 75% to 85%. This should make intuitive sense to you because as with any bell curve, you have a lot of crowding around the median, which is roughly 3.2.

G. T. L. Rev. wrote:

Anonymous User wrote:3.4 is much higher than a third. You can roughly convert gpa to percentiles if you have some basic stats knowledge.

FYI, at Mich 3.15 to 3.25 is the mean target not median. There is a substantial difference between the two measures. For instance, if a professor gives out 10 As and 5 A+s, the class will have a higher mean than if the professor gave out 15 As but the median hasn't moved. This is very important understanding individual class grading guidelines:

--LinkRemoved--

Note how for instance a professor can give out A+s to anywhere from 0-3% of the class and A-s to anywhere from 13-17% of the class. This can have the effect of significantly pulling up the mean GPA for a class without affecting medians, particularly if the top half is substantially better than in previous years and the bottom half of the curve looks like previous years. A 3.3 mean is therefore entirely plausible without affecting medians.

These percentages allow you to roughly map out a continuous bell curve and calculate your percentile using a markov model. I did this using excel. It's not perfect and assumes that across your 6 classes, you average out to the target percentages in that pdf. But to give you an idea, 3.4 will put you at anywhere from 75% to 85%. This should make intuitive sense to you because as with any bell curve, you have a lot of crowding around the median, which is roughly 3.2.

Are you seriously suggesting that a 3.4 goes from top 15% after 1L year to just a hair above median at graduation? I do not doubt your methodology; however, I am skeptical that upper level classes could change the grade distribution so significantly.

Me too. I think top third or so is correct.

i can't believe there's no concensus on this.

I had roughly a 3.4 at Michigan after my 1L year. I pegged this at roughly top third at Michigan. I bid very conservatively, but ended up with callbacks in the V5 and offers in the V10. That was 2009. I can only imagine it'll be better this time around.

Based on the information listed above (the clerkship letter that lists the 5 year average of *graduating* GPA percentile cutoffs), I have extrapolated the following percentile cutoffs for 1Ls. Really, this should be a worst case scenario and your perecntile might be marginally better.

3.25 -- top 50%
3.39 -- top 33%
3.46 -- top 25%

After that, the extrapolation is most likely too inaccurate. These numbers should be approximately correct, though.

Disclosure: I am a umich 1L.

Anonymous User wrote:I had roughly a 3.4 at Michigan after my 1L year. I pegged this at roughly top third at Michigan. I bid very conservatively, but ended up with callbacks in the V5 and offers in the V10. That was 2009. I can only imagine it'll be better this time around.

Don't mean to thread-hijack, but I have a 3.46 heading into OCI. You got any suggestions for firms to bid to (I'm solely interested in NYC)? You mentioned V5/V10, but I looked at all those firms and they were mostly in the higher (3.5+) GPA bands... any suggestions would be welcome. Thanks.

Anonymous User wrote:3.4 is much higher than a third. You can roughly convert gpa to percentiles if you have some basic stats knowledge.

FYI, at Mich 3.15 to 3.25 is the mean target not median. There is a substantial difference between the two measures. For instance, if a professor gives out 10 As and 5 A+s, the class will have a higher mean than if the professor gave out 15 As but the median hasn't moved. This is very important understanding individual class grading guidelines:

--LinkRemoved--

Note how for instance a professor can give out A+s to anywhere from 0-3% of the class and A-s to anywhere from 13-17% of the class. This can have the effect of significantly pulling up the mean GPA for a class without affecting medians, particularly if the top half is substantially better than in previous years and the bottom half of the curve looks like previous years. A 3.3 mean is therefore entirely plausible without affecting medians.

These percentages allow you to roughly map out a continuous bell curve and calculate your percentile using a markov model. I did this using excel. It's not perfect and assumes that across your 6 classes, you average out to the target percentages in that pdf. But to give you an idea, 3.4 will put you at anywhere from 75% to 85%. This should make intuitive sense to you because as with any bell curve, you have a lot of crowding around the median, which is roughly 3.2.

Nah. Nice try though. First, nearly every 1L professor grades towards the higher end of the allowable distribution, so assume that the mean of all the grades in a given class in 3.25. Since the professor will probably give out 25% A+'s, A's, and A-'s, and 30% B+'s, the median in any given class will almost always be a 3.30. But this is irrelevant because every student has six grades from 1L year, so the median GPA for individuals in the class should approach the mean class mean, which is about 3.25(assuming the 3.3 number the registrar provided is wrong, and I think it has to be). So, assume that the median after 1L is 3.25. According to Dean Baum's letter, the average median at graduation between 2006 and 2010 was about 3.37. So, just to do a rough calculation, assume that the average student raises his GPA about .1 points between the end of 1L year and graduation (although people with higher GPA's probably raise them less, so maybe the cutoffs I provide for top 10% - 1%. are a little low). This is a rough calculation, but even so, here is what you get using the stats from Dean Baum's letter, and I suspect its pretty accurate:

3.27 = top 50%
3.39 = top 33%
3.46= top 25%
3.63= top 10%
3.70= top 5%
3.90= top 1%

Anonymous User wrote:

Anonymous User wrote:I had roughly a 3.4 at Michigan after my 1L year. I pegged this at roughly top third at Michigan. I bid very conservatively, but ended up with callbacks in the V5 and offers in the V10. That was 2009. I can only imagine it'll be better this time around.

Don't mean to thread-hijack, but I have a 3.46 heading into OCI. You got any suggestions for firms to bid to (I'm solely interested in NYC)? You mentioned V5/V10, but I looked at all those firms and they were mostly in the higher (3.5+) GPA bands... any suggestions would be welcome. Thanks.

Same anonymous as above here.

I think certain firms are definitely out of range: S&C and Cravath. But otherwise, I think it's definitely worth looking at places like Simpson, Paul Weiss, K&E, and Skadden. Davis Polk, Debevoise and Cleary are going to be a tougher, but if you think you have very strong interviewing skills, they're worth a shot. Really, though, you're in the range for some of those firms and pretty much every firm below them. I'd structure the bidlist to put the more reach-y of the above firms at the bottom. Those firms tend to be undersubscribed in terms of bidding since so many like to play it safe. You'll still get interviews with them, I think. For the more target-ish firms, I'd place them at the top of the bid list, just to ensure that you're putting yourself in front of good firms that you have a legitimate shot at. And then I'd put every other firm (i.e., the other target-safeties, and safeties) in the big juicy middle.

Another Michigan rising 2L here. I agree with 3.27 being roughly median (using Dean Baum's letter is the best bet). I think that poster's 10%-1% is off-mark, but he/she admitted as such, and it's really hard to tell where the top of the class is.

You really should look at the GPAs posted on Career Services website to get a gauge at firms you are competitive at, keeping two caveats: (1) these GPAs do not represent hard cutoffs (or at least not all of them do -- they are means/medians, not mins); (2) they may be slightly lower than they should be since they are from 2005-2010 (so they include pre-crash numbers).

Associate Dean West provided an explanation of why the article was wrong in a letter to the editor. I could not find it online and do not have a hard copy anymore.

Can this be right? Looking at the grade curves for Winter 2010, the 1st year mean was 3.39. I realize the summer starters take more electives but still, could the registrar have made this error 2 semesters in a row? Is it possible the mean is really that high?

Sounds like top third is credited. It's nice to see so many Michigan rising 2Ls posting (although maybe we're overusing the anon feature just a tad). I think it would be good to have a Michigan OCI 2012 thread. Maybe this should be it.

Any thoughts on whether OCS's (or whatever the hell they're called now) staunch belief that one should apply to 1-2 markets maximum is valid? (I know this was the subject of some controversy on TLS last year.) What about Chicago and New York with no ties to either plus a few applications to a tertiary market with strong ties?

You have to consider that while you receive 30 bids, you don't receive nearly as many interviews. The average tends to be around 11-12. You don't want to split 11-12 interviews between two markets. And, as it stands, I think 11-12 are too few even for one market.

Anonymous User wrote:You have to consider that while you receive 30 bids, you don't receive nearly as many interviews. The average tends to be around 11-12. You don't want to split 11-12 interviews between two markets. And, as it stands, I think 11-12 are too few even for one market.

Oh wow I didn't even know you only get 11-12 interviews. (I should probably read those emails more carefully.) Yeah that makes one market make a lot more sense. Damn. So do the vast majority of people go outside OCI then, even people with top grades?

Can't you just take all the grades for first year classes on the website and take the average of all of those and get the mean? If you do a small example: say you know that there are 3 people in an entire hypothetical law school class of 2013, they took 3 classes, so there are 9 total grades. Say there is 1 A, 6 Bs, and 2 Cs. You don't know who got those grades or in what combination, but no matter what the combination is for those 3 people individually, the average will always be the same.

Or I could be completely wrong...

But I put this in excel today and it came out to 3.24.