HOW THE LOTTERY WORKS
Students rank studio preference on the ballot distributed by the Program Council Representatives.
“1” indicates first choice, “2” indicates second choice, etc.
All studios must be ranked for the ballot to be accepted
The lottery process is anonymized by student CUID number which is not known by the representatives
Only numbered rankings are considered – comments are disregarded
Program Council runs the lottery.
The final studio placements are determined by representatives of the Program Council and checked by the Program Administration.
The studio ballots are collected and, as of summer 2026, processed by three different sorting algorithms that balance maximizing first-choice placements with the best average placement:
- Option 1 – Water Filling: This algorithm protects students from bad outcomes by resolving overloaded studios. Every student starts in the first-choice studio, then, for each overfull studio, the student with the best available alternative gets moved to their next best open studio—repeating until no studio is over capacity. This output has the highest first-choice rate and best average placement with no notable tradeoffs.
- Option 2 – Stable Matching (Deferred Acceptance): This algorithm maximizes the number of first-choice studio placements. Each unmatched student “proposes” to their top remaining studio. Studios accept up to capacity but bump their lowest-ranked student if a better proposal comes in. Bumped students re-propose to their next best choice until everyone is placed. This output has no standout strengths but has the highest average placement and highest worst-case placement.
- Option 3 – Minimum Regret: This algorithm minimizes the overall dissatisfaction across the whole group. The student with the fewest good options gets placed first into their best available studio, repeating until everyone is assigned, then pairs of students are swapped wherever a trade would reduce total disunification. This output has the best worst-case placement but the lowest first-choice placement rate.
For each algorithm, the best of ten iterations is presented in a spreadsheet. The representatives compare the three lists and make adjustments to refine the best of these three options. The algorithmically produced placements are comparable to the manual placement system used previously but significantly streamline the process. The final placement is still determined by the representatives.
All lottery moves are recorded in an Excel spreadsheet that allows representatives and administration to track the average ranking (how many students end up in first, second, third, etc. studios divided by total students). The overarching goal is to have the lowest average number possible without sacrificing students to their lower choice studios. Both representatives and administration track this average number against previous years.
In the event that students must be moved to a studio that they ranked unfavorably (typically 6+ for Core 1 – Advanced 4, and 8+ for Advanced 5 and 6) in order to fill studio quotas, they will be compensated with the option to select first choice in the following semester.
The Administration checks the lottery results.
- Once the lottery is complete, all ballots and Excel spreadsheets are verified by administration and names are matched with CUID numbers.