I am confused about match

[TJMAXX]

---

Members do not see this ad.

If you look at the matching graph from last year, 7 out of 9 people who ranked more than 16 programs didn't match. While people who ranked less than this number had more chances to match? How would you explain that?😕

 

[podtsan]

Maybe the applicants ranking that many programs were only those well below average (hence ranking so many to make up for poor application), with the result that many of them didn't match because they were so weak.

 

[yaah]

exactly.

 

[scurred]

sometimes people rank multiple specialties besides being a poor applicant...that also hurts them...nobody wants someone wavering on what they want to do.

 

[TJMAXX]

If you look at the graph carefully, the 16 programs they ranked were all pathology 😛

sometimes people rank multiple specialties besides being a poor applicant...that also hurts them...nobody wants someone wavering on what they want to do.

 

[jrs1234]

If you look at the graph carefully, the 16 programs they ranked were all pathology 😛

I don't know what your status is, but they were also all IMG. It could have been some difficulty with getting their visa or application in general.

 

[bioguy]

Using my model, there are two factors that would lead to a low match percentage:

P(match) = 1 - (1-p)^n

low n: although the candidates ranked 16 programs, it does not imply that 16 programs ranked the candidate. A program may decide not to rank a candidate even after interviewing them. Thus, the candidate may have ranked 16 programs but was only ranked by 10​

low p: - the candidate has a low probability of matching at any individual program (say, say p= 0.02).​

Assuming 9 identical candidates (a simple but unrealistic scenario), there are a variety of ways you could get a find a combination of (n, p) that would produce a 7/9 failure rate. For example, a candidate who ranked and is ranked by 16 programs but is ranked very low (p=0.03) would, on average, match successfully about 28% of the time which is close to 2/9. Or, assuming nonidentical candidates (more likely) there could have been two strong candidates who matched and seven very weak candidates who had no prayer of matching.

As you can see, it is easy to come up with scenarios under which 7 of 9 candidates who rank 16 programs fail to match. Thus, there is nothing really surprising about the match data.

dude!!! whats wrong with you? you shouldn't be in Pathology, you should be an Astrophysicist trying to decipher the eleven dimensions

 

[TJMAXX]

Thank you! I understand your theroy.
But How do you explain why the high failure ratio only happens to people who ranked more than 16 programs not to other people who rank much less number of programs? 😳
So I Believe in more about the week cadidates reason.😉

Using my model, there are two factors that would lead to a low match percentage:

P(match) = 1 - (1-p)^n

low n: although the candidates ranked 16 programs, it does not imply that 16 programs ranked the candidate. A program may decide not to rank a candidate even after interviewing them. Thus, the candidate may have ranked 16 programs but was only ranked by 10
​

low p: - the candidate has a low probability of matching at any individual program (say, say p= 0.02).
​

Assuming 9 identical candidates (a simple but unrealistic scenario), there are a variety of ways you could get a find a combination of (n, p) that would produce a 7/9 failure rate. For example, a candidate who ranked and is ranked by 16 programs but is ranked very low (p=0.03) would, on average, match successfully about 28% of the time which is close to 2/9. Or, assuming nonidentical candidates (more likely) there could have been two strong candidates who matched and seven very weak candidates who had no prayer of matching.

As you can see, it is easy to come up with scenarios under which 7 of 9 candidates who rank 16 programs fail to match. Thus, there is nothing really surprising about the match data.