The Nobel Prize that Keeps Affecting My Life

Doctor's graduate from medical school every year - but not all of them will practice medicine.  Thousands of them will never get the chance.
CNN wrote a great article entitled: Why your waiter has an MD that explains the problem.

Here's the gist of it:
All med-school graduates who want to practice in the USA have to complete a "Residency Training Program."  If you want to specialize in anything such as Anesthesia, Psychiatry, OB/GYN, Family Practice, Surgery etc… you need to complete a 3-5 year residency and then pass your specific specialty board exams.  After that you can do EVEN MORE training called a “fellowship” to be a sub-specialist; such as a vascular surgeon, a child psychiatrist, or a pediatric oncologist. 

There are a few problems with this system.

In 2012 :
38,377 doctors applied to residency training programs in the United States:  there are only
24,034 slots available.

1^st – 14,000 doctors who graduated from medical school and passed their general medical boards – will not be able to practice in the USA because they didn’t get into a residency.

2^nd – How do you decide who goes to which residency?  Let’s say I wanted to be a dermatologist.  I applied to 45 different dermatology residencies.  I interview at 8 of them. 
Those 8 programs have interviewed anywhere from 15 to 100 applicants.  Every residency wants the best applicants, and every applicant wants the best program (or sometimes any program).

What is the best way to decide who goes to which residency?  Everyone knows you (the applicant) are interviewing at 7 other places - and you know they are interviewing another 50 people for the spot you want.  Who picks?  If you aren't picked by your top choice - do you then apply late to your second choice?

How do you make it fair and most advantageous for all parties involved?

SWITCH GEARS!

A while ago I read the book “A Beautiful Mind” – the biography of John Forbes Nash.   I blogged about it because it was interesting and insightful - and I got to see the life of a genius with schizophrenia.  There was an entire chapter devoted to his mentor – Lloyd Shapely.

Shapely was a genius of game theory. 
(He and John Nash and two others even invented their own game called “So Long Sucker.”)

Shapely did a lot of theoretical mathematics in game theory – like figuring out how to give the best result for all players when everyone is competing against each other.

 – Which brings us back to the problem of 38,000 doctors applying for 24,000 residency slots.

Let’s say one residency can accept 2 applicants per year.  They interview 20 candidates each year and then rank those 20 doctors from first choice to last choice.

Those applicants have interviewed at 15 different residency programs, and have ranked them from first choice to last choice.

How do you match them up?  How do you make sure that after you’ve weeded through 38,000 applicants applying to 4,427 residency programs offering 24,000 slots – that every had the best result?

How do you make sure that there is not a residency and an applicant that would both prefer each other over what they ended up with?

Let’s say Stanford Dermatology has 3 slots.

I ranked them as my #1 choice, and I am their #20 choice.

I ranked UCLA as my #2 choice and I am their # 5 choice.

I ranked Texas as my #3 choice and I am their #2 choice.

Where do I go?  Do I go to Texas because it's the best match?  Just becuase the combined ranking score at Texas (3+2) is less than UCLA (2+5) or Stanford(1+20)?  What if Stanford’s top 19 picks ranked some place else as their #1? Do I become Stanford's #1 pick because I'm the only guy left who really wants them?

Whose list takes precedence?  Does Stanford send out 3 letters of acceptance, wait to see if they get rejections, and then try to fill the slots with the next 3 people on their list?

There is a system called the National Resident Matching Program (NRMP).  It’s designed to take every applicants rank list, and every residency programs rank list – and match them up. The applicants choice is considered first, then the Residency's.

Many applicants think - Can I beat the system?  How can I get into a better program than I’m really qualified for?  Could I rank programs at the top of my list that I have no chance of getting into – and then hope that the system will feel sorry for me and give me the first program I have a shot at (which is my REAL 1^st choice but I ranked it low to game the system?)

You may think I’m kidding.  I’m not.  I had friends in medical school trying to play this game with their training sites and med-school rotations.  They tried to beat the system, to play the game.

Luckily – the Matching system (NRMP) is based on Shapely’s theories about game theory.

A man named Alvin E. Roth proved that the NRMP was both stable and strategy-proof.  It gave everyone the best possible outcome, and it was impossible to “win the game” or “beat the system” by some strategy.

It is in everyone’s best interest to be completely honest in their rankings.  You have the best chance of getting your 1^st choice, and so on down the list.  It is true for the applicant, it is true for the residency.

This seems to make perfect sense – but how hard was it to make that system?  How do you allow for so many variables.

Some residencies only have 1 slot, some have 20.  Some applicants only rank 1 program, some rank 30?

That is why Shapely and Roth are winning the Nobel Prize this year.  Their combined efforts have developed a perfect system.  It can’t be beaten, tricked, swindled, twisted, or played.  Everyone uses the exact same strategy.

3 years ago I applied to 10 programs.  I interviewed at 5 programs and made my "rank list."  On “Match Day” every graduating medical student found out where they were working for the next few years.  I found out I was moving to Reno, my first choice.

2 months ago I applied for a “sub-specialty fellowship.”  I interviewed last month.  I submitted my “rank list” last week.

January 9^th 2013 is “Match Day.” It’s the day I find where I’m training for the next 2 years (if I matched at all).

December 8^th 2012 is the Nobel Prize Ceremony.  That’s the day that Roth and Shapely will give their Nobel Lectures.  I’ll be listening, and silently thanking them for making my life, education, and career - much easier.

To Dr. Roth and Dr. Shapely – thank you.


P.S. (Random Fact) - The Nobel Prize in Economics is not actually a Nobel Prize.   There are Nobel Prizes in Physics, Chemistry, Physiology/Medicine, Literature, and Peace.
It's actually awarded by the Central Bank of Sweden and is called "The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel"