These are great data, but I'm concerned that the methodology is fatally flawed. Do we have any real reason to believe that whatsthefare requests track closely to actual ride requests?