Once I signed on to work with Paul Holes in late 2017, I was anxious to start putting together a profile of the man I was looking for. This is one of the first things that I do when I agree to work on a case. I knew there had been a Snapshot composite profile of the Golden State Killer produced by a DNA lab that had previously worked on the case, and I asked Paul if I could see it. (Snapshot is a forensic DNA-analysis tool that attempts to create a likeness of unidentified persons by pinpointing phenotypic characteristics such as hair and eye color.)
Paul explained that because I was a civilian and not a law enforcement officer, he was not permitted to provide me with any official documents. All he sent me was one page of the lab report describing the killer’s bioancestry information, which, I later discovered, was not even completely accurate. Paul did not send me any phenotypic data—for instance, hair color, eye color, freckled or unfreckled skin, or other information that might have been useful.
Up until then, I had been reporting to both Paul and FBI attorney Steve Kramer, but because Paul was in the process of wrapping up his cases ahead of his retirement, Steve Kramer became my sole point person on the investigation. Steve was assigned to the FBI’s Los Angeles field office, and I called him there to ask for a copy of the Snapshot profile. Steve was friendly, but he did not waste time. He, too, said he could not send me anything. I simply was not authorized to see any official material regarding the case.
“You are asking me to identify this guy and you won’t give me any information at all?” I asked, probably sounding more impatient than I intended.
“It’s not up to me,” Steve said. “It’s Orange County. They’re not letting any of it go.”
Orange County was just one of several California counties in which the Golden State Killer had struck. As Paul Holes also explained it to me, there were a lot of internal politics at play between counties, and a few grudges that long predated my entry into the case. Whatever the reason, it meant I would not have access to any official paperwork.
This was frustrating, but I had experienced such interdepartmental-type squabbling before. When I was a first-year graduate student at UC San Diego, I was caught up in a rivalry between two laboratories supervised by two professors, and I was even falsely accused of contaminating one of the labs with radioactive iodine by an unscrupulous postdoctoral fellow. I understood how competition between departments could get in the way of productivity.
I also knew that it was possible to begin a case having next to nothing to work with and still solve the case—as I had with the Lisa Jensen Project. In fact, I only really needed one thing to identify the Golden State Killer: his DNA.
But that, too, was a problem.
Apparently, we did not have his DNA.
In the course of his twelve-year crime spree, the Golden State Killer left behind a wealth of DNA evidence, primarily hair, skin cells, and semen found at his many crime scenes. Over the decades that followed, however, each new advance in DNA technology led to some of that evidence being used to conduct whatever the new tests were, in hopes of generating fresh leads.
For instance, in the late 1980s the FBI began developing a national DNA database that would come to be known as the Combined DNA Index System, or CODIS for short. Essentially, CODIS was a vast store of DNA profiles drawn from criminals and crime-scene DNA. Law enforcement can compare DNA from their crime scene against the CODIS database. As soon as CODIS was operative, officials working on what was then called the East Area Rapist case submitted DNA samples they had preserved from earlier crime scenes.
Unfortunately, those samples did not match any existing DNA profiles in CODIS. The new database was powerful but limited—if a perpetrator had no criminal record, the crime-scene DNA was not of much use.
Early DNA tests also required a relatively large amount of high-quality DNA material. This caused the supply of the Golden State Killer’s DNA to shrink quickly. Investigators had to balance the desire to keep up with new DNA technology against the risk of eventually running out of DNA material to test. (It is only in recent years that testing methods have progressed to where much less DNA is needed for an accurate test, which means that crime-scene DNA will last longer—but still not forever.)
That was the reality of the Golden State Killer case. In a way, it was a desperate race between rapidly improving technology and a rapidly dwindling supply of the killer’s DNA. What would happen first—the key breakthrough in DNA analysis that would lead to the identification of the killer, or the very last DNA sample being wasted on another dead end? Who would win the race—the good guys or the murderer?
By the time Paul Holes called me in March 2017, he had run out of all the Golden State Killer’s crime-scene DNA to which he had been given access. He was going to have to find another county that still had DNA and was willing to share it with him. The obvious starting place was Orange County, the third-most populous county in the state, and also the scene of four of the Golden State Killer’s murders.
In the fall of 2017, around the time I was able to work on the case in earnest, Paul reached out to the Orange County Sheriff’s Office for help.
They turned him down.
Once again, I was astonished. How could officials from nearby counties not find a way to work together to solve such a high-profile case? With resignation in his voice, Paul again cited a network of internecine spats and grudges that seemed to govern intercounty cooperation in California.
“That’s Orange County,” he explained. “That’s just the way they are.”
But, as it turned out, the race was not quite over—and failing to get a sample from Orange County proved a blessing in disguise.
While Paul Holes went down the list of California counties in search of a DNA sample, I went to work building an initial profile of the serial killer with what little information I had, and with whatever I could deduce from newspaper accounts and public records.
What did I know about the suspect? Well, I thought, let’s start with his age. I knew he had begun his rape spree in the mid-1970s, and I surmised that when he had escalated from burglary to rape in 1976, he had to be at least twenty years old. Why did I think that? Because he did not rape just one time, he raped multiple times. If we were looking for someone who had committed only one assault, I would have to allow for the possibility that the perpetrator could have been as young as, say, fifteen, and that the attack had been an isolated impulse offense. But given that he had struck so often—and that he had been able to elude police by scaling tall fences and outrunning them in the woods—I guessed that he had to be at least twenty when he started. He had to be capable of a certain level of physicality. This was not a boy committing these heinous, often strenuous acts.
These were a man’s crimes.
So if he were twenty or older in 1976, he had to have been born in 1956 or earlier. He likely lived in California—possibly Northern California—and had either a military or a law enforcement background. Forty years previously he had been described as athletic, about five foot ten and having blond or light brown hair. Beyond this, there was not much material that I could use to build a profile, although I would be able to get some additional information, such as eye color, from his DNA.
Paul Holes had not only built his own extensive physical and psychological profile of the killer, he had also described his profile in detail in a four-part TV series. The series aired right around the time I started work on the case, and I watched all four segments and took notes.
One of Paul’s theories was that, since many of the rapes had taken place in new housing developments, the Golden State Killer likely worked in an occupation that had something to do with real estate, such as a Realtor or developer. Possibly, Paul guessed, he might have worked in the construction business.
Paul’s profile was based primarily on supposition, whereas my profile included only factual information or educated biographical estimates—a possible year of birth, likely residential locations, or observed physical characteristics. To these, I would add verifiable data that would, hopefully, be revealed by the killer’s DNA.
For two months—October and November 2017—Paul came up empty in his quest for DNA for our project. Many other county investigators had run out of DNA material, too, or did not have enough to share, or had samples that were too mixed with a victim’s DNA to be of much use. To achieve the results we wanted, we needed a preferably single-source, uncontaminated sample of the Golden State Killer’s crime-scene DNA. It was beginning to look as if one might no longer exist outside of Orange County.
I did not know it at the time (although I have certainly learned it since), but solving really challenging DNA cases or crimes often seems to require a random twist of fate—a moment when something unexpected happens that defies probability and ends up being just the thing you need to keep going. When that happens, it feels as if the universe is nudging the case along. A document that was believed destroyed is found; a person who refused to do a DNA test has a change of heart; information that once appeared completely inaccessible no longer is. I have referred to these lucky happenings as my guardian angel.
In this case, it was one of the serial killer’s tragic victims—one of the two he slaughtered most violently—who reached out from the past, through decades of silence, and gave us precisely the thing we needed most.
Everything seemed off.
A side door that was normally unlocked to let in the landscapers was locked. The front door was open, although it usually was not. Three days’ worth of newspapers were stacked up on the front porch. Inside the home, the living room looked a mess, with sofa cushions out of place, although the house was always kept tidy. A wake-up alarm rang from behind the bedroom door, even though it was nearly noon.
Gary Smith, then twelve years old, had bicycled to his father’s home in Ventura County to mow the lawn—his weekly chore—and he had no idea why things seemed so strange that day, March 16, 1980. He walked into his father’s bedroom and saw the bedcovers pulled up over two people and, inexplicably, a thick fireplace log placed between them on the bed. Slowly, Gary pulled back the bedcover on one side, and saw his father, Lyman Smith, lying there.
The back of his father’s head was caved in.
Gary called the police and waited in a daze on the front porch for them to arrive. The house filled with investigators. Both Gary’s father and Gary’s stepmother, Charlene Smith, were dead, bludgeoned by a blunt object, presumably the bloodied log, which the intruder had apparently taken from a pile outside the house. Lyman Smith was found naked and facedown, his hands and ankles bound with a length of cord from the drapes. Charlene had on a T-shirt and was faceup, her wrists also tied. Both were bound, and the cord was tied using a distinctive knot called a sailor’s knife lanyard knot, or diamond knot. In the days after the brutal double murder, newspapers referred to the unknown assailant as the Diamond Knot Killer.
Shortly after the bodies were found, the Ventura County assistant medical examiner at the time, Dr. Peter Speth, arrived at the house. Not all medical examiners attended crime scenes to collect samples; most preferred to take their samples when the bodies arrived at the crime lab. But Dr. Speth almost always made it to the scene of a rape. He believed that moving and transporting bodies caused seminal fluids in a body cavity to shift and mix, compromising the purity of a sample later taken in a lab.
In the Smiths’ bedroom, Dr. Speth went to work.
Because DNA testing was not yet possible in 1980, Dr. Speth concentrated on tests for blood groups and enzymes, which could be used to identify possible suspects, provided the samples were as close to pristine as possible. Employing a standard sexual assault kit that had been created just a few years earlier, Dr. Speth used a swab to take a vaginal sample from Charlene and put the swab, cotton-end up, in a rack of vials. He air-dried the vial, sealed it with a capsular desiccant, and put it on dry ice he had brought to the scene with him. Once he was back in the Ventura County forensics lab, the sealed vial went into a freezer.
The process went smoothly and without incident, and Dr. Speth had the pristine sample he needed. His work at the crime scene, in theory, should have been finished. But it was not.
In the bedroom of the Smiths’ tragically violated home, Dr. Speth did something that was not part of any rule book, not required of any medical examiner, and not a common practice among any medical examiners anywhere—not in 1980, and not today.
If Peter Speth was not, as he believes, the only medical examiner in the world to routinely do what he did next, he was certainly among an infinitesimally small number of forensic practitioners who have done it.
Over the years, Paul Holes had been judicious with the DNA material from the three Golden State Killer crimes that had occurred in Contra Costa County. Still, he had no choice but to dip back into the samples every time there was a breakthrough in DNA technology. Advances in DNA analysis using Y-STRs, or short tandem repeats on the Y, or male, chromosome, looked promising enough that Paul spent six years—and the last of his DNA material—working the new technology to try to get a surname for the suspect. Unfortunately, he came up empty again. There were just no useful matches, suggesting that the men in the suspect’s paternal line may have been recent immigrants.
When I joined the case, Paul was running out of counties to ask for DNA samples. Finally, he and Steve Kramer briefed Steve Rhodes, a Ventura County prosecutor and a member of that county’s cold-case task force, on the new technology that I had used to identify Lisa Jensen. Ventura County was the scene of two of the GSK’s (Golden State Killer’s) murders: Charlene and Lyman Smith. Paul asked Rhodes if he had any GSK crime-scene DNA available to share.
“Let me dig around,” Rhodes said.
At least it was not a no.
Rhodes called the Ventura County Medical Examiner’s Office and asked a technician there to look through the sexual assault kits stored in the lab’s freezer.
It turned out that the county’s chief medical examiner, Dr. Ann Bucholtz, recently had begun the process of cleaning out the lab’s cluttered and disorganized freezer, which was packed with samples and other evidence that no one seemed to need. Many of the samples had already been transferred from an older freezer, and if that transfer had been handled improperly, it was likely that the samples had thawed out and been ruined. Dr. Bucholtz assigned a technician the task of cleaning up the freezer and told her, “You can probably discard everything.”
The technician, Janelle Payne, methodically went through the samples. At the very bottom of the freezer she found a tub, and in the tub was an old but still sealed rape kit and other sealed evidence bags. She was surprised to find an unopened rape kit, since rape kits were almost always opened and the samples analyzed early in an investigation. Yet here was a completely pristine kit. That was uncommon enough for Payne to bring the kit to Dr. Bucholtz’s attention.
“Why do we have this unopened old kit?” she asked Dr. Bucholtz.
“I don’t know. Let me call the prosecutor’s office.”
The prosecutor Dr. Bucholtz talked to had no idea as to whom the kit belonged, either. Eventually, though, the existence of the kit was made known to Ventura County investigator Steve Rhodes—the man Paul Holes and Steve Kramer had asked for Golden State Killer DNA.
Rhodes asked for the name and date on the rape kit.
The name was Charlene Smith, and the date was in 1980.
Peter Speth had long wanted to be a pathologist—the medical professional who conducts lab tests, studies body tissues, and helps reach diagnoses. “A pathologist is the doctor’s doctor,” says Dr. Speth. “You have to know every disease in every part of the body, and you have to be able to recognize it by how it feels, and also simply by eyeballing it. Why did this person bleed from the bowel? What kind of brain tumor is this? You need many years of training to become a pathologist.”
Dr. Speth underwent that training and earned a fellowship under Dr. Boyd Stephens, San Francisco’s first-ever chief medical examiner and a world-renowned expert in forensic pathology. “He was a unique guy, in that he went to every crime scene and established himself as the boss there,” says Dr. Speth. “Nowhere else was the medical examiner in charge of investigators and the other criminalists at the scene of a crime.”
Dr. Speth, who drove everywhere in a small red Renault (his CB handle was “Red Rascal”), tagged along with Dr. Stephens to every crime scene and learned how to be aggressive and persistent in retrieving and processing good samples. But he developed his most distinctive practice all on his own. “It occurred to me that we, as doctors, should somehow be able to preserve evidence so that it will be available for future research using new technology,” he says. “It just seemed logical to me to save evidence from every single case I worked.”
Dr. Speth saved evidence by taking not just one sample at a crime scene, but two samples—something no other medical examiner did.
One of the samples was for the investigating officers working the case. The second one was safely stored away, its seal unbroken, should there ever be a need to reexamine the evidence.
At the home of Charlene and Lyman Smith, Dr. Speth took four vaginal swabs and placed two of them in one red-topped vial, and two in a different vial. He eventually sealed both vials and labeled them precisely to begin and preserve an impeccable chain of custody. Whoever it was that broke the seals would be photographed doing it and become part of the custody chain.
The first vial went to the investigators assigned to the Smith double murder. The second vial was stored away in a freezer.
That second vial, miraculously, survived unopened and forgotten for the next thirty-seven years—until technician Janelle Payne found it in a tub at the bottom of a crime lab freezer that was deemed so disastrously disorganized its contents hardly seemed worth saving.
Dr. Speth, for one, had no idea the second kit had survived, nor did anyone tell him when it was found. “The Charlene and Lyman murder scene had haunted me since 1980,” he says. “I lived year after year waiting and hoping that the true monster who carried out those murders would be caught before he died, and before I died.”
So it was that Ventura County prosecutor Steve Rhodes called Paul Holes and Steve Kramer and delivered the most remarkable news: he was in possession of a pristine unopened rape kit from the scene of the Smith double murder. The odds of such a thing happening were incalculably small. And yet it had happened—the second kit existed. Because of Dr. Speth’s unique practice of assembling two kits, Paul Holes now had the precious DNA sample he had hoped beyond hope to find. He now had, he told people, “a gold mine of Golden State Killer DNA.”
What lay ahead was unprecedented and historic, and would, in time, introduce to the world the use of investigative genetic genealogy to identify a suspect in cases involving violent crimes.
