Tuesday, March 25, 2014

Residency Matching Algorithm (Also, I Matched!)

I am so so excited to report that I matched to my #1 program on my list, Georgetown (Go Hoyas)!!!!!!  I loved everything about the program, and I’m excited to try out DC!

For those who haven’t been lucky enough to hear my many explanations on how the match works, here’s an article my brother sent me that pretty succinctly explains how the process works: algorithm.

But what I found MORE interesting about that article is that the two people who developed this algorithm (Dr. Shapley who invented it and Dr. Roth who later modified it to fit the NRMP’s needs) won the 2012 Nobel Prize in Economics for their work!!

Dr. Shapley (with Dr. Gale) created an algorithm in the 1960’s to solve the “stable marriage” problem – which, in a very simplified way, is this: if there is a list of men and women, how do you match them up (note: this does not take into account same-sex marriages) so that there are no “unstable matches” in which all involved would have been happier in a different match? 

They solved this using "deferred acceptance," meaning that there would be multiple rounds of matches in which individuals would select a second option/mate (and then a third, etc.) if their most preferred option rejected them.

However, this was obviously stupid as that is not how marriage or people work.  So while the mathematics world was thrilled (maybe?), everyone else just went about living their lives as usual.

…I swear I’m going somewhere with this.

While it may not be how romance works, it turns out that it can be how residency works!  In fact, when the NRMP (National Resident Match Program) was established by med students in the 1950s (because getting into a residency was PuRe cHaOs before that), it turns out that’s exactly what they were using without even realizing it. 

In the 1980’s, Dr. Roth showed that the NRMP actually was using the Gale-Shapley algorithm, and he then went on to revamp the process in 1995 to favor the match based on applicants’ lists rather than programs’ list (after much controversy and student protest basically forced their hand).

So.  The moral of the story here is that if you are creating something with absolutely no real-world application, just hold on another 40-50 years, and maybe someone else will come up with a use for it and YOU TOO could win a Nobel Prize.

No, the real moral of the story is I’m going to be an internal medicine resident at Georgetown next year! Yay!

more sources not already linked: Nobel PrizeNRMPNobel explanation (this one really explains the general algorithm quite well)


  1. Your rank order list is confidential. Program rank order lists can be seen only by the program director, the institutional official, and NRMP staff unless you give your username and password to someone else.

  2. It requires time,money and energy to solve this problem. As the technology advances, it gonna have a solution with persistence.