APPIC Match Algorithm Explained

Posted by: Kenn on March 13, 2013 | No Comments »

Any student facing the APPIC Match process is at least somewhat concerned about how the Match process works. We’ve all heard throughout grad school that the Match occurs through some mystical algorithm. This is perhaps the most nerve-wracking part; the idea that a computer handles such an important decision is pretty scary. This article attempts to ease some of that fear by shedding light on how the Match algorithm works. As you read on, you will learn that students actually have the advantage as the algorithm runs, along with a simple strategy for choosing how to rank your sites.

After students have applied and successfully interviewed at various internship sites, they are asked to submit a Rank Order List of their preferred locations to APPIC. Actually, the list is not submitted to APPIC directly, instead students submit their list to the National Matching Service (NMS), a third-party organization which runs the matching algorithm. The internship sites are also asked to submit Rank Order List of their preferred candidates along with the number of positions their site has available.

The algorithm used by NMS is pretty simple and could actually be completed by hand, but a computer does it much faster and without human error. The algorithm starts with an attempt to place a student at the location they ranked highest on their list (their first choice).  If the student cannot be matched there, an attempt is made to match them to their second choice.  This continues until an applicant obtains a tentative match, or all of the student’s choices have been exhausted.

A student is tentatively matched to a program if the internship site has included the student on its Rank Order List and one of the below possibilities:

1. The site has an unfilled position available
2. The student is ranked higher by the program than another applicant who has already been tentatively matched to the program.

Matches are tentative because a student who matches to a program at one point in the algorithm’s process, may later be removed from the program to make room for an applicant who is ranked higher by the site.  If this happens, an attempt is made to re-match the student starting from the top of their Rank Order List.

This process continues until each student has either been tentatively matched to a program or all choices on the student’s Rank Order List have been exhausted.  When all applicants have been considered, the algorithm is complete and all tentative matches become final.

As an example, consider the below fictional Rank Order Lists from three Internship programs and three interested applicants:

1. The algorithm first attempts to place Jane at her first choice, which is the Child Center. The Child center ranked Jane on their list and has an available position, so Jane is tentatively matched to that site.
2. Next, the algorithm attempts to match Tom to his first choice, which is the Stress Clinic. As before, the Stress Clinic ranked Tom on their list and has an available position, so Tom is tentatively matched there.
3. The algorithm now attempts to match Kate at her first choice, which is the Child Center as well.  However, the Child Center does not have any available positions, because Jane currently fills the one position offered by the site.  This is when the internship’s rank order list comes into play.  The Child Center ranked Kate higher than Jane on their rank order list, so the algorithm takes Jane out and replaces her with Kate.  Therefore, Kate is now tentatively matched with the Child Center.
4. Since Jane was just removed form a tentative match with the Child Center, the algorithm attempts to find another match for her.  First, the algorithm attempts to match her with her first choice, the Child Center, but Kate who is a more preferred candidate fills that position.  So, the algorithm attempts to match her at her second choice, the Stress Clinic.  The Stress Clinic ranked Jane on their list and has an available position, so Jane is tentatively matched there.
5. The algorithm continues in this fashion. After each student’s list has been processed, the matching process is complete and all tentative matches become final.   Note that no applicant or program can be forced into a final match until all applicants’ rank order lists have been considered for the best possible tentative match.

To put it another way, an applicant’s rank order list is traversed “downwards”, from their most preferred site to their least preferred site until the student is tentatively matched or their list of choices is exhausted.  An internship site accepts applicants “upwards” on its rank order list, continually removing less preferred matches in favor of more preferred applicants, until all available positions are filled by the most preferred applicants who wished to be matched to the site.

In conclusion, with the above information in mind, there is a simple strategy to use when choosing which sites to rank.

1. Always rank the sites based on where you TRULY prefer to go.
2. Work hard to be ranked as high as possible on the internship’s list, which helps ensure other candidates cannot take your position.