Advice to job candidates

This is almost certainly too late, repeats what has already been said elsewhere, contradicts what has been said elsewhere, and reflects only my own peculiar views. But for what it's worth, here are some reflections on what can make a job application better or worse, and what strikes me as a search committee member reading applications. I have read close to 40 applications for our job so far and the general quality is amazingly high.

1. Unfortunately, our ad in Jobs for Philosophers is an abbreviated version of the ad on the VMI website. The full version asks for evidence of teaching excellence. Not everyone includes such evidence, and some provide much more than others. More is generally better than less (see below for exceptions to this rule).
2. Including a statement of your teaching philosophy seems to be obligatory, but most don't really provide much useful information. A really good one might help, but a generic one won't hurt you. 
3. If your student evaluations are really good you should include them. If they used to be poor but they have got much better, maybe only include the recent ones. Make sure you say what scale is used for numerical evaluations (especially if your school has a maximum score of 4 when most others have a maximum of 5). I wish every school used a 5-point scale, but they don't. 
4. Beware of including negative comments from students. Ideally comments will be provided and will be unedited, but they will also be uniformly glowing. If yours aren't, then think about not including them at all. Of course, a couple of bad ones among hundreds of great ones is OK.
5. You should have at least one teaching letter, and this should go into detail on what you do well. It should be more than half a page long. 
6. Other material can be impressive too, for instance details of efforts you have made (seminars attended, etc.) to improve your teaching.
7. Someone should check your letters to make sure they are detailed enough to be useful and don't contain bad typos (e.g. misstating the likely date of your dissertation defense, thereby suggesting that you don't meet the criteria stated in the ad).
8. A very minor point thrown out for what it might be worth: I find that I like to know who someone is when I read their letters, so cover letters that include the applicant's name on the first page (perhaps because they are no more than one page long) seem best to me.
9. It looks as though you pretty much need to have some publications to be competitive, but, against my better judgment, I find myself worrying that an applicant with lots of publications in more selective journals than I usually publish in will not want to stay here. Caveat: "pretty much need" does not mean need in our case, if other aspects of the application are strong enough. But we do want reason to believe that you will have published enough for that not to be a problem by the time you come up for tenure (and we expect our job to become tenure-track, even though it isn't advertised that way).
10. Although it shouldn't, I think it helps if you have something search committee members can connect with, such as letters from people they know, similar research interests, etc. This can probably be faked finessed to some extent by looking up the school and presenting yourself as working on, or wanting to teach, the stuff that people there work on or teach. 

OK, ten points is probably enough. I hope this helps someone.