Adnan Syed’s Post-Conviction Hearing, Day 2 (Part 2): Expert Gerald R. Grant Jr. Eviscerates The State’s Cell Site Evidence
I’m down in Baltimore for Adnan Syed’s post-conviction trial hearing and am posting daily recaps of proceedings. You can also follow me on Twitter at @xoamelia for short updates from the courthouse.
As I said earlier, Day 2 was LONG, so I broke my recap into two parts. Part 1 recapped Maryland Deputy Attorney General Thiruvendran Vignarajah’s cross examination of Asia McClain. Part 2 is below, and recaps the testimony of the defense’s cell site expert, who also recently testified in the Boston Bomber trial and has an eight-page resume!!! I’m gonna do my best to explain what he said about the issues that are important at this trial, rather than exhausting myself trying to understand and put into words all the technical specifics. I haven’t quite figured out how to connect my fucking iPhone to the Cloud, so….
Day 2, Witness #2: Gerald R. Grant Jr., expert in cellular phone forensics and historical cell site analysis
Direct Testimony: In advance of his testimony, Grant did the following:
- Reviewed the testimony of the State’s 1999 trial expert, Abraham Waranowitz
- Reviewed the State’s 1999 closing testimony
- Reviewed all of the cellphone records
- Reviewed the cell tower evidence used at the 1999 trial
- Spoke to Waranowitz about his testimony in 1999 (Waranowitz has since submitted an affidavit saying his testimony would have been different had he been provided with the instructions for how to interpret that cell site data)
Alright, so C. Justin Brown, Syed’s attorney, asked Grant to explain how the cellphone data was used in the 1999 trial. Grant said that the location of Syed’s phone on January 13, 1999, was determined by the coverage areas defined by the towers, not by charting specific individual calls. I believe what he means is that the prosecution looked at which tower was pinged during a specific call and surmised where in that area the phone likely was.
For example (and this is my example, not Grant’s): when Jay Wilds called Syed at 10:45 a.m. the cell tower that was pinged was L651A. Cell tower L651A. is near Woodlawn High School, which is where both Syed and the State says Syed was when he made that call. But hypothetically, let’s say that Syed had said he was at, I dunno, a McDonald’s near school when he made that call — so long as the McDonald’s was within the same coverage area of cell tower L651A, the data as it was used at trial could not have proven that he was lying. If he had said that he made the phone call from all the way across town, well outside of the cell tower reach, then this data could have been used to prove he was lying. Make sense? I believe this is what Grant meant when he said the State didn’t chart specific calls, because the cell tower data was a limited look at the phone’s signal making it impossible to pin down an EXACT location.
To make sure that the cell site data was accurate, Grant testified, Waranowitz used a test phone to make “controlled outbound calls.” All of the test calls were outgoing calls. He was able to compare the calls from the test phone and the cell sites they were pinging in order to confirm that the data AT&T had provided was indeed correct before testifying. Brown asked Grant if the State explicitly affirmed the reliability of the cell site data at trial, and Grant confirmed that they had.
Brown brought out Trial Exhibit 31, which is a five-page document made up of individual pages from three separate larger documents. Ex. 31 contains:
- Page 1: an authentication from AT&T
- Page 2: the last page of Syed’s REDACTED subscriber activity report, with the cell site columns blacked out (un-redacted records require a court order)
- Page 3-5: three pages from Syed’s UN-REDACTED subscriber activity report, where the cell site data is visible
Brown asked Grant whether, in his experience, cellphone companies generally send instructions on how to read the records (Answer: Yes.) Grant said yes, then testified that these instructions vary from company to company and from time period to time period, as there have been advancements in technology. While cellphone companies use the technology similarly, how they store/convey their records is different. Therefore, he testified, in order to read cell phone data correctly, the instructions must be from the same company/year as the phone records. Grant testified that, as an expert in this field, he relies on the instructions in order to do his analysis.
Next, Brown showed Grant Exhibit 15, which included a fax coversheet from AT&T followed by Syed’s un-redacted subscriber activity report, i.e. the source document from which pages 3-5 of Ex. 31 was pulled. Ex. 15 was not submitted into evidence in full during the 1999 trial. Grant testified that he is sure that Ex. 15 is a subscriber activity report because it says so on top of the last page. Now, the fax coversheet (page 1 of Ex. 15) has instructions which explicitly state, with emphasis:
Outgoing calls only are reliable for location status. Any incoming calls will NOT be considered reliable information for location.
Grant said that if he had been asked to testify in Syed’s 1999 trial and had received that fax cover sheet along with the subscriber activity report, he would NOT have ignored these instructions when making his analysis, and that he would not be able to plot the location of the incoming calls without further information from AT&T.
Grant was then shown Waranowitz’s affidavit, which states that he received Ex. 31 from the State “just prior” to testifying. Reminder: Ex. 31 included three pages of the un-redacted subscriber activity report (Ex. 15), but did not include the fax cover sheet and instructions that went along with it. His affidavit states that had he seen those instructions, he would not have given the testimony he did about the incoming calls to Syed’s cellphone at 7:09 and 7:16 on the evening of January 13, 1999. His testimony at the time supported the State’s claim that Syed was in Leakin Park burying Lee when those calls were received.
To reiterate: the State’s own expert witness from 1999 has submitted an affidavit essentially retracting his testimony about those two incoming calls, which the State used to place Syed in Leakin Park when Lee’s body was supposedly being buried. This was used to corroborate Jay Wilds’ testimony.
Cross Examination: I think Vignarajah was trying to do a couple different things with his cross examination, but it was hard to follow. Sometimes I thought he was trying to show that the instructions were really just boiler plate; sometimes I thought he was trying to cover former prosecutor Kevin Urick’s ass, since he’s potentially being accused of a Brady violation; but ultimately, what Vignarajah was trying to do was make a case for Grant being WRONG in his interpretation of the instructions.
Vignarajah kept asking Grant if he might be wrong in considering Syed’s phone records — i.e. Ex. 15 and pages 3-5 of Ex. 31 — a subscriber activity report, and Grant kept insisting that he was not wrong. Vignarajah tried to illustrate that Grant might be wrong by asking him if he has ever seen this particular issue, of the unreliability of incoming calls, used to “attack” cell tower testimony. Grant said he had not in his experience, but that he began doing historical cell site analysis AFTER 1999 and technology has improved, so issues with incoming calls not being reliable for determining location has not been an issue for him. Vignarajah kept trying to use this as if to paint Grant’s testimony as exceedingly rare, and that rareness equals strangeness equal POTENTIAL WRONGNESS. Grant found many of Vignarajah’s questions to be confusing, especially as he kept comparing analysis of different documents from different years, when Grant had been clear that you can do no such thing.
What Vignarajah literally could not seem to understand is that using these instructions about incoming calls to attack testimony about cell site locations would only happen in cases where that testimony ATTEMPTED TO USE INCOMING CALLS TO DETERMINE LOCATION. In other words, it would only occur in cases where the instructions were explicitly IGNORED. And perhaps Grant hasn’t experienced that because most cell site experts are provided with instructions on how to read the data they are analyzing. This issue is being used to attack the testimony and evidence in the 1999 trial because those instructions were ignored by the State and not provided to their original expert. And now that the original expert has seen the instructions, HE HAS RETRACTED HIS TESTIMONY ABOUT THE INCOMING CALLS.
Redirect: Brown asked Grant if he’s ever heard of an expert not being provided with instructions on how to read the data their analyzing. Grant said NO, that it was unusual.
State’s Rebuttal: Vignarajah asked Grant if he knew of a reason why incoming calls wouldn’t be reliable. Grant said he had heard of a case where the incoming calls were showing cell tower pings from the CALLER’s location, not the call recipient’s.
Then Vignarajah went back to the issue of “subscriber activity reports” and whether the “locations” referred to in the instructions was in fact referring to cell site locations, not an entirely different column on the subscriber activity reports. Grant insisted that Vignarajah is wrong. Vignarajah then said he would be calling his own expert — a new one, you know, since their expert from 1999 no longer stands by his testimony — who will testify about Grant and Waranowitz interpreting the instructions incorrectly.
My Assessment: The cellphone records came with instructions on how to read them that were not provided to the State’s expert and the jury in 1999. The State’s own expert, Waranowitz, has since essentially retracted key portions of his testimony, and both he and Grant are in agreement about how those instructions should be read and would have been read back in 1999. THIS IS WHAT MATTERS. That sound you hear is half the State’s case being put through the shredder.