beach_terror wrote:While LSAT/GPA are the best indicators for success in law school, there's absolutely no chance you can rely on those figures to predict performance like you are. For all the math you throw around, you seem to forget or maybe just don't know, how subjective the grading can be between the top 10% and the top 25%, then 50%, so on and so forth. Even if they have a decent probability of doing well, I'd never take the bet. It's just moronic. Everyone on this website will probably agree with me. You have to go to a school you'll be happy graduating at median at, because that's about the best you can accurately predict.

Actually, if I could bet against everyone who thinks they'll be top 10% of their class, I'd be pretty goddamn rich. I'd bet against every time, and I'd come out on top way more often than I'd lose.

Edit: While I have no formal math background, wouldn't any amount of subjectivity in grading severely damage a prediction curve like that, even if it was 100% accurate objectively?

Most people on this website will agree with you because they are just regugitating from other people: if the school in question != T14, then retake LSAT. if OP is thinking about transferring, then he/she must reconsider because its hard. While this is generally the case, however, this is without taking into consideration the specific factors in every case IF it is blindly apply in every scenario that is being thrown out.

Further, accuracy and precision is a matter of fantasy in the world of math and engineering. What the above mathematical function shows that for every collection of data / sampling points out there, be it LSAT or 1L grades, a guassian function closely models it in real world. That is, given a mean and a certain standard deviation, those who perform above said mean in one scenario (for ex. LSAT), is reasonably expected to perform just as well in another scenario (1L grade), if the guassian function is use to model both.

As unpredictable as law school grading is, there is a mandatory curve that dictates the median of the class: that is, every professor must grade the class on the basis of this median. Thus, to each individuals, it may very well be unpredictable. However, from a probabilistic perspective, mathematical function is still applicable. Here, the model is never meant to accurately predict anything, it is to model the most likely/reasonable outcome.

Suppose St. John's 1L median LSAT is X and 1L median is Y. OP's LSAT is X1, which is 2 standard deviations above X. While i agree that LSAT does not accurately predict OP's performance in 1L, however, it is still mathematically reasonable to expect OP to perform 2 standard deviations above Y or else the basis upon which Law school admissions admits applicants objectively will be annihlated. Is it possible that OP COULD underperform? sure, anything possible. Its not likely. Given the statistical probabilities, i am willing to bet, to gamble, to take the risk that OP will probably be very successful at St. John.

You have to go to a school you'll be happy graduating at median at

"You have to"? Who makes you the boss? No one has to do anything. However, everyone is "recommended" to go the best school he/she can get into or the school that would leave one with zero debt, given their

preferences. The main issue of this thread is "whether or not a full ride at St. John be turn down". I never suggested OP has to go to St. John. Rather, i suggested options that could optimize returns. Here, the fact that OP got a Full Ride indicates OP is someone with strong credentials and great probability for success. Thus, i drew a reasonable conclusion or prediction that OP will do very well in school. On the basis of this prediction, i offer him/her reasons why he/she should consider St. John as a possibility.