Just a note about bite statistics... unless they are expressed as a function of the population of the breed in question (which they almost never are), they are pretty much useless. Even the CDC decided at one point to stop tracking bites by breed because saying "there were 35 bites by breed X and 65 bites by breed Y in 2010" is pretty much useless if you don't know how many X's and Y's there are in the country right now.
Here is why:
Assuming that breed identification is 100% correct, let's say in 2010 there were 100 serious or fatal bites by Breed X and 10 serious or fatal bites by breed Y in Anytown in 2010. Breed X is looking pretty bad right now.
But, let's say there were 10,000 Breed X's and 1000 Breed Y's living in Anytown in 2010. So the rate of bites per population for both breeds is exactly the same, 0.01 or 1%. That means whether you mean an X or a Y, your chances of meeting a member of that breed who will inflict a serious or fatal bite on you are pretty much equal, even though looking at the raw numbers Breed X looks far more dangerous.
What makes it even more wooly is that there is really no way to count the population of unregistered dogs of any breed. So a breed or type or whatever you want to call it like APBT or bullies or pit bulls or whatever you want to call them probably has a far lower bite rate than it appears by looking at raw numbers, because I suspect there are a relatively high number of unregistered members of those breeds/types.
Once I actually sifted through a crapload of CDC data and calculated the rates of bites as a function of population for a crapload of breeds, and for the most part they all came out pretty much the same. I sincerely wish I had kept that in a spreadsheet somewhere because I am far too lazy and unmotivated to do it all again.
I don't particularly disagree with any of the other points made about concerns with these breeds/types, dog bite statistics just happen to be a pet peeve of mine.
