I don't actually know what my print runs have been. I have some guesses, but nothing from the publisher. But then I got to thinking...
We know the median first novel advance for a SF/F author is probably around $5000 or so. That's the boilerplate first offer I got from Baen (which then fell through, but that's another story). Average is a little higher than the median, but I'm going to stick with $5000 for ease of math.
We also know not all novels earn out their advance, especially first novels. $5000 is a best-guess on the part of the publisher as to how much they should invest in your new book.
Sticking purely with mass market paperbacks for the moment, let's say you get royalties at 8% (fairly standard but not universal for an original mass market, I believe) and a cover price of $7.99 (also standard U.S. cover price for mass markets). So you're earning $.64 per book. Juggle the numbers, and a $5000 advance means you're going to need to sell roughly 8,000 books (7,812.5) in order to earn out. In my case, I'd guess the publisher probably did a print run between 10,000 and 15,000 books, but that's a total guess, and hopefully more experienced publishing folks can speak to that piece. (ETA: ramblin_phyl points out that there's also a break-even point in the cost-efficiency of first print runs, which might mean the numbers on that run were a little higher.)
Hardcovers and e-books add more variables, as the royalties are different, but I'm trying to keep things as simple as possible for this example.
The numbers are important, but so are the expectations. Goblin Quest got a $4000 advance from DAW (lower than average, in part because it was a reprint). In 2.5 years, it's sold somewhere around 15,000 copies or so. I don't have exact numbers, so that's a best estimate based on royalty statements and other sources. 15,000 is well above and beyond the number needed to earn out the advance, which means the book is doing pretty well, and my next advances were higher. Go me!
Compare that to a book that gets a $40,000 advance. Here the expectations are very different, and 15,000 sales would be abysmal.
There's a lot I don't know, and every book is different. All else being equal, a debut YA urban fantasy will probably get a higher print run and better expectations than a debut comedic fantasy, because the former tends to sell better. Total number sold also depends on how long the book stays in print (DAW is good about keeping books in print, and Goblin Quest has probably sold an extra 1500-2500 copies in the past six months alone, two years after it first came out).
So, helpful data or just me overobsessing and playing with numbers again? Any questions or things I didn't cover? I'm hopeful that others will also fill in some of the gaping holes left by my own lack of knowledge.
Edited to fix my bad math skills.