Yet another statistic from this Animals 24-7 group that doesn't appear to be peer reviewed or available online. Is this one group the source of all these claims?

Claims about pit bull attacks inherently rely on local officials deciding how to classify the breed that attacked. That likely isn't going to be particularly accurate, considering shelter staff (which work with dogs day in and day out) frequently can't correctly categorize dogs as pit bulls or not. 48% of the dogs labeled pit bulls in the study had no pit bull ancestry. If shelter workers are this bad at identifying breeds by sight, I don't have high expectations for cops.