How long does the average law school applicant wait to hear back? Are those with high LSATs and low GPAs—or with low LSATs and high GPAs—more or less likely to be admitted? Are early candidates more qualified, strictly by the numbers? Does it really matter if you apply in November instead of in February? Drawing on self-reported data from LawSchoolData.org and utilizing Plotly, this project aims to provide users with the analysis tools necessary to answer these and related questions.

While some general trends are discernible—it tends to help if applicants apply by the end of November, for instance—the takeaway broadly is that the admissions context varies greatly from school to school. What holds for Yale may not for Northwestern, and vice versa.

Current as of 7/28/2022 and updated approximately weekly. The first plot is live for the 2021/2022 cycle; all other plots will be updated for the current cycle once more data become available, likely near the end of February, 2022.

---

Usage Note:

The plotting package Plotly makes possible an interactive experience, among other things allowing users to: pan and zoom, using the toolbar at the top of each plot; adjust the x and y axes, by hovering over an axis at the middle or at either end; and highlight and hide the data sets represented in the legend, by single- or double-clicking on the legend markers.

Note: Historical percentages and likelihood are calculated based on the past four cycles (17/18, 18/19, 19/20, and 20/21). The Notified trace includes only those who have received an acceptance, rejection, or waitlist—thus does not include withdrawls or holds. This trace is calculated by reference to the total number of applicants who received an acceptance, rejection, or waitlist notification. The A, R, and WL traces are calculated by taking the number of acceptances, rejections, or waitlists up to a given date and dividing by the total number. Acceptance Likelihood is calcuated by dividing the number of acceptances remaining at any given point by the number of applicants who at that point had not yet received an acceptance, rejection, or waitlist and who had not withdrawn. The data filtering requirements discussed in the technical details, above, for plots 1, 3, and 4, mean that the Acceptance Likelihood trace is likely to be of limited reliability. The faint diagonal lines against the background represent months of waiting: y=x (0 months), 1 month, 2 months, ..., 6 months. Marker outlines in blue represent splitters (>75th percentile LSAT and <25th percentile GPA) while marker outlines in black represent reverse splitters (<25th percentile LSAT and >75th percentile GPA). Until 2022 percentile data become available, percentiles for 2022 are assumed to be percentiles from 2021. The dashed blue line marks the date of the last data update.

Note: The top-left plot shows the acceptance rate by date of application submissions, calculated by dividing the number of acceptances among applicants who submitted their application in a given month by the total number of applicants who submitted applications that month, whatever their end result, including Pending. This method of calculation assumes, therefore, that the overwhelming majority of those applicants who were admitted updated their LawSchoolData.org status to reflect their admission and that there are no significant correlations between application date and the liklihood of updating a LawSchoolData.org profile. Because the plot above reports the total number of acceptances, rejections, and waitlists only, the n= totals may differ between this plot and the plot above. The remaining subplots in the present figure draw on the same data set as does the top-left plot. The top-right and bottom-right plots show the mean of the LSAT and GPA, respectively, of all applicants who applied in a given month. The bottom-left plot shows, for each month, the percentage of the total volume of submitted applications submitted that month. For example, if December of the 18/19 cycle were to show 23% it would indicate that 23% of the total number of applications submitted that cycle were submitted in December. Error bars represent one standard deviation.

Note: Wait times are calculated for each group (A/R/WL, Accepted, Rejected, Waitlisted) by averaging the number of days from the time an applicant sent their application to the time the applicant receieved a decision. Holds are not included. See plot below for a further breakdown of wait time data.

Note: Error bars represent one standard deviation. Use this plot in conjunction with the plot above.

Note: Acceptance rates are calculated by dividing the number of regular, splitter, or reverse splitter applicants admitted by the total number of applicants from the relevant category who applied, whatever their end result, including Pending. The number of splitters and reverse splitters who applied and were admitted is for several schools so small as to make meaningful inferences impossible. For each school, Index values are calculated by dividing the acceptance percentage of splitters or reverse splitters by the acceptance percentage of applicants who were neither. The greater above 1.0 a value is the easier it was for splitters or reverse splitters to gain admission, compared to regular applicants; the lower below 1.0 a value is the harder it was. Splitters are applicants with an LSAT score greather than the 75th percentile and a GPA less than the 25th percentile; reverse splitters have a low LSAT and high GPA. Actual acceptance rates come from the ABA—but until 2021 admissions data become available, actual percentages for 2021 are assumed to be actual percentages from 2020. Table columns may be rearranged by dragging.

Note: Percentages are calculated by dividing the total number of applicants and admits from LawSchoolData.org by the total number of applicants and admits as reported by the ABA. An applicant counts as having used LawSchoolData.org to report their applying to or hearing from a given school if they entered at minimum their LSAT score and GPA. Until 2021 volume data becomes available, volumes for 2021 are assumed to be volumes from 2020.

Technical Details:

Unless otherwise noted: 1) all admissions data is drawn from LawSchoolData.org; 2) holds and withdrawls are not included in calculations of wait time, acceptance rate, and other derivative values; and 3) and waitlisted, waitlist -> rejected, waitlist -> accepted, and waitlist -> withdrawn reports are treated identically, as waitlists. Note that, because there appears to be some inconsistency in the self-reporting of the decision date for waitlist -> rejected, waitlist -> accepted, and waitlist -> withdrawn reports—some users update their decision date upon second notification while others leave it unchanged since receiving the original waitlist notice—there is likely to be some error in waitlist timing data.

If cycles are not reported in the format 15/16 then the cycle number listed represents the year in which applicants for that cycle would begin law school; e.g., cycle 17/18 is equivalent to cycle 18.

The plots below are generated from the pool of applicants who have entered on LawSchoolData.org at minimum their LSAT score and their GPA. The data pool for plots 1, 2, 3, and 4 has the additional restriction, because these plots analyze time data, that applicants are included only if they entered on LawSchoolData.org the date at which they sent their application; the pool for plots 1, 3, and 4 is further restricted by requiring applicants to have entered the date at which they received their admissions decisions. Accordingly, counts for plots 1, 2, 3, and 4 may differ from those available on LawSchoolData.org itself, which allows users to update their decision without providing any other information.

Those interested in examining the source code may wish to visit the GitHub repository for this project. Feedback, suggestions, and other notes may be directed to u/IneffablePhilosopher.