The data for Rank order list and residency was released few days ago. Not surprisingly the results indicate that those applicants who matched had longer ROLs. You can find the results here www.nrmp.org/wp-content/uploads/2014/06/Impact-of-Length-of-Rank-Order-List-on-Main-Residency-Match-Results-2002-2015.pdf
The average length of the rank order list was 10 for matched applicants as compared to approximately 4 for the unmatched applicants. This is not really that surprising since all it reflects is the fact that those applications that are viewed favorably get more invitations to interview for the match. In a hypothetical scenario where both applicant A and B are invited to interview at same number of places (N) then the probability of the applicants matching depends on how many programs the applicants place on their rank order list. If applicant A has interviewed at N places and puts N programs on his ROL contrast to applicant B who interviewed at N places but put <N programs on their ROL, then the probability of A matching is > than B since, A had a longer ROL.
Another interesting trend one can see if you look at the numbers closely is that in 2002 Approximately 80.5% matched and had a ROL length of 7.5, Every year the ROL length is increasing but % matched continues to drop. Fast forward to 2015, % of matched applicants is 76% and mean ROL length is 10. It is taking a lot more to match as the years go by and the match chances keep getting poorer.
So what does it mean for an international applicant? It means that in order to maximize the chances of a match you have to place ALL programs you interviewed at on your ROL. While it may be an option for US grads to not rank a program that they did not like, this may not be an option for IMGs. Thus unless the program had a highly unfavorable impression and there is no reason to considering training there (at the cost of perhaps even not matching) then it may be okay to consider not ranking a program.