ROL_2010, you're right. The extra interviews may help some people only marginally and may not be worth their extra time and effort if they match at their number one and their number one would not have changed because of that last interview or two. But who are those people? No matter how many rank lists you may think you are at the top of, the programs decided that not you.
As mentioned before, ofcourse stronger applicants will get more interviews. However, I think it's pretty clear that if you don't go to an interview your chances of being ranked at that program are 0%. If you do go you likely will be ranked somewhere on the list. Being somewhere on 14 programs' lists gives you a better chance than being somewhere on 13 programs' lists. Sure, maybe all 14 programs will match their top 4-10 and you are number 15 on every list so it didn't help you. The point is, why take that risk? Especially when, as mentioned by others, you may end up ranking that 14th program very highly if you just go check it out. After investing so much in this process to determine your future, why not just finish the job?
I realize this means nothing to those who've submitted rank lists already, now that I realize what the date is. However, those who read this stuff and will apply next year should strongly consider trying to make as many interviews as possible. Whatever the "odds you will match at your last choices", just going to the interviews cannot hurt your chances of matching at a program you will really like. Isn't that why you're doing this anyway? Is it worth the marginal gain? The statistical gain may be marginal in your mind if you think you're at the top of everybody's list. However, the reality of matching vs. not matching and possibly passing up a program that you might have loved makes the statistics a little less important. Statistically, very few of your future patients will get a post-op PE, but I dare you to skip the DVT prophylaxis based on statistics. The clinical significance sometimes outweighs the statistical significance. That's all I'm saying.