I'm not quite sure which makes me want to reach through the phone and strangle the cell phone customer service rep more: the "I understand your concern sir and am truly sorry for your inconvenience" scripted response to every issue they refuse to solve for you, or the fact that they are NEVER wrong when you tell them your phone is dropping service, or your bill is incorrect, or that you're pretty sure you've never downloaded a Shania Twain ringtone.
The vast majority of the time, any errors in the cell phone provider's data records are, at most, going to cost you the time spent arguing with them on the telephone (and I guess the appurtenant angina medicine required to keep you heart beating) and the money you're going to inevitably pay them once you realize you are fighting a losing battle. For some, however, these records might mean the difference between freedom or spending the rest of your life in prison. Such was the case with our client Charles Anthony Murphy, Jr. who was charged with conspiracy to commit murder, two counts of murder, and one count of attempted murder and who was recently tried in Orange County, California and defended by none other than The Law Talking Guys (criminal law defense attorneys Michael Molfetta and Paul Wasserman).
The primary pieces of evidence against our client were entirely derived from cell phone records. This required us to select smart jurors who were willing to not accept as Gospel cell phone company data that was replete with errors, inconsistencies and omissions. Fortunately, Mike and I picked such jurors and were able to convince five of the twelve that a man should not be convicted of murder using records the were demonstrably erroneous, and thus obtained a mistrial as the jury was hopelessly deadlocked 7-5.
So just how bad could records that the prosecution are relying upon to convict a man of murder be? Well..... pretty bad actually. In our case, the cell phone provider was T-Mobile. To demonstrate that our client was part of a criminal conspiracy to brutally slaughter an innocent family, the prosecution leaned heavily on two sets of records produced by T-Mobile in response to a search warrant: "billing records" for our client's cell phone (i.e. bills sent to him for payment), and supposedly corresponding back-end records that show what goes on "behind the scenes" when a call phone call is placed. These back-end records contained data that purported to show what "switch" a cell phone call utilized on the network (think of a router on your home computer network) and to what cell tower within that switch "family" handled the call. All well and good if the records prove to be accurate and complete.
This is where years spent as a computer and electronics geek (and being terrorized by jocks like Molfetta when I was a kid because of it) paid off. If there is one thing working with computers all these years has taught me, it's that data integrity can often be verified by cross-checking databases that contain the same information. Thus, if I were to look at an entry on the billing record, there should, of course, be a corresponding entry on the "back-end" records that show what tower and what switch connected to the cell phone dialing or receiving the call, what the time and date were for the call, and what the duration of the call was.
Unbelievably, there were a large number of call data entries where the information didn't match between the two records; one of the most striking examples being a call our client made where he was billed for 36 minutes of connection time when the back-end records showed the exact same call lasting for a little over 17 minutes.
More disturbingly, there were voluminous examples where calls billed to the client appeared on the billing records but NOT on the back-end records, and vice-versa. This missing data was particularly troubling as the prosecution was relying on existing data to prove our client was at a certain place at a certain time due to the cell tower location history. This, of course, begged the question of how reliable the data the prosecution actually had was, but also what call data we were missing that supported the fact that our client was not in the area the prosecution claimed he was at time when NO call data appeared in the records.
Time will tell if Mr. Murphy will be re-tried. However, the important lesson we learned in this trial AND admitted to by the T-Mobile expert on cross-examination: cell phone records are NOT designed and maintained with forensic evidence use in mind. They are first and foremost designed for one thing and one thing only: making a profit for their shareholders. Relying upon them as if they are the equivalent of DNA evidence is a dangerous practice indeed.