Fifty-Seven Halloweens: Pam Hobley, Patty Spencer, and the Man Who Ran Out the Clock

Pam Hobley and Patty Spencer walked away from Oscoda High School on Halloween afternoon in 1969 and were never seen again. Police questioned a teenage musician that year. He lived free for 56 more years, and when DNA finally named him in a different girl’s murder, he died before anyone could ask him about Oscoda again.

Two Girls on River Road

I am publishing this on Halloween on purpose. Fifty-seven years ago today, two girls in Oscoda skipped out on the afternoon of the biggest night of their fall, and two families have been setting extra places ever since. Every year the date comes back around, the local stations run a short segment, and the case goes back in the drawer. This year there is a name attached to it, and the name comes with a deadline that already passed.

Oscoda is a small town on the Lake Huron shore in Iosco County. On Friday, October 31, 1969, Oscoda High School had its homecoming football game against Ogemaw Heights, followed that night by a dance and Halloween parties. Pamela Sue Hobley, 15, and Patricia Ann Spencer, 16, were schoolmates who ran in the same crowd but were not considered close friends. That afternoon, they left school together.

According to the Western Michigan University Cold Case Program’s case summary, they were last seen around 2 p.m. walking east on East River Road, away from the school. Pam wore a white fur coat with brown trim over a brown and white plaid skirt, white knee socks, and chunky shoes. Patty wore a brown sweater, a tweed or plaid skirt, a gray and green plaid jacket, and thick-heeled brown shoes. Her missing-person profile notes a jacket with a peace sign painted on the back and a peace-sign pendant. Neither girl had a purse or ID.

Accounts of the next stretch differ in their details. The Charley Project, the national missing persons database, reports that a passing driver picked them up on River Road and took them to a gas station; other accounts say he dropped them downtown. Police have said other information they have not released also places the girls downtown that day. The driver has been cleared.

Pam’s family went trick-or-treating that night. When they got home and Pam had not, her mother started calling around and learned Patty was missing too. Pam’s sister Mary Buehrle was 8. She has spent most of her life since then asking people to remember.

The First Week Went to the Runaway Theory

The early investigation followed a script that shows up in missing-girl cases across Michigan in that era. For roughly the first week, the working assumption was that Pam and Patty had run away, with Flint as the likely destination. Police issued a statewide alert. Residents pooled a $600 reward.

Three weeks in, Oscoda Police Chief William Tarnosky told The Flint Journal, “There just isn’t a trace.” In May 1970, the girls’ mothers, Arlene Spencer and Lois Hobley, published an open letter in The Bay City Times. They acknowledged that teenagers run away, said their daughters would not have gone that long without making contact, and described themselves as worn down to the point of losing hope.

Their reasoning was sound on the facts. Two girls who leave without money, identification, or a change of clothes, on a night they had plans to be seen at, are not behaving like runaways. The families knew that in the first month. Pam’s sister Bettye Myers has said that by Christmas, when the girls did not come home for the holidays, the concern finally deepened.

Fifty-Seven Years of Tips, Barns, and Persons of Interest

The case never fully closed, but it moved slowly. In 1985, police received information that the girls had been killed by two local men and buried near a barn known as a teenage party spot. According to the Oscoda Press, the local police chief directed a cadaver-dog search of the area in 2011, 26 years after the tip. It found no evidence of human remains. Other places teenagers frequented were searched with the same result.

A new detective took the case in 2010 and began re-interviewing witnesses. In 2013, police said they had a person of interest but needed more information to move forward. Over the years, Pam’s family heard claims that the girls had made it to Hale, Flint, or Saginaw. Hobley’s dental charts and both girls’ DNA profiles were entered for comparison against unidentified remains nationwide. In December 2015, the FBI issued a ViCAP alert to homicide and missing persons units across the country, stating that foul play was suspected.

In December 2023, Michigan State Police detectives from the West Branch Post and the Oscoda Township Police Department announced a joint cold case team. Det. Sgt. Erik McNichol told ABC12, “This has been a cloud over the community for a long time.” The same report said investigators believed potential suspects in the case could still be alive. Students from Western Michigan University’s Cold Case Program, a partnership with the state police, digitized and organized the entire case file and produced investigative support documents for detectives.

Flint, 1983: The Case That Named Him

To understand where the Oscoda case stands, you have to start with a different girl. Sheri Jo Elliott was 16 and had recently moved to Flint to live with her mother after living with her father in Charlotte. She attended Carman High School and babysat part time. WMU’s case summary describes her as shy, smart, and careful around strangers.

On November 16, 1983, she left home to catch her school bus and did not come back. Four days later, her body was found along a rural road in Blumfield Township, in Saginaw County. According to the Michigan State Police release, she had been shot multiple times and sexually assaulted. The case went cold. In the years that followed, members of her own family came under suspicion, a burden families in unsolved cases carry far too often.

The state police release says the Third District Cold Case Unit reopened the investigation in 2023. WMU’s program page says the case had been reopened multiple times since 1983. Working with the MSP Forensic Science Division and Othram, a private Texas lab, detectives used forensic-grade genome sequencing and genetic genealogy to build a profile of the unknown man. It led them to Roni Collins of Grand Blanc, a bassist who went by the nickname Hendrix and had been a fixture in the Flint music scene for decades.

Before detectives could get a voluntary DNA sample, Collins died by suicide in January 2026. DNA collected during his autopsy matched the evidence preserved from Sheri Jo’s body in 1983. The state police announced the case solved on April 13, 2026, and credited the WMU students with reorganizing and digitizing decades of investigative material.

One detail in the coverage matters more than the headlines gave it. WMU program director Ashlyn Keursten told A&E that detectives, with students’ help, discovered untested biological evidence in the Elliott file.

Unless Collins was already in a DNA database, which nothing in the public record shows, a test in the 1990s or 2000s would not have produced a name by itself. Genetic genealogy is what produced the lead, and that tool has only been in wide use since 2018. But an untested sample is not a neutral fact. A profile developed years earlier could have been compared against every man who had been questioned in a similar case, including a musician whom police had interviewed about two missing girls in 1969. The public record does not show whether anyone ever lined those two files up before 2026.

The Musician in the Van

On April 22, 2026, ABC12 reported that state police confirmed Collins had been questioned in the Hobley and Spencer disappearance. In 1969 he was about 19, living on Harry Street on Flint’s north side, and playing in bands in the Hale and Oscoda area. Police say the girls were seen talking to someone in a van before they disappeared. Collins was questioned by police in Flint and said he knew nothing about it.

Mary Buehrle remembers him from her childhood: he spent time near the beach by the little park, she told ABC12, and he drove a van. Her own words about where she has landed after 57 years: “My emotions are up and down but I truly believe he took the girls.”

Keursten pointed to the same overlap. Collins played in a band that performed for high schools across the region where Sheri Jo, Pam, and Patty lived, and all three girls were about the same age when they were taken. After his death, investigators across the Midwest began checking their own cold cases against his DNA.

That is the full extent of what has been made public about Collins and Oscoda. He was there, he was questioned, he drove a van, and he denied it. There is no forensic link, because there are no remains and, as far as the public record shows, no biological evidence from the girls to test. State police have renewed their call for tips, and that is the honest status of the case: a strong person of interest, now dead, and a file that needs a witness.

Dead Before the Handcuffs

Collins is not the only Michigan cold case suspect to die at the finish line in the past year.

In August 1986, a woman on a bicycle trip to Sault Ste. Marie stopped at a public restroom at the Big Fishdam River access site on U.S. 2 in Garden Township, Delta County. A man forced her into his vehicle at gunpoint, drove her to multiple locations, sexually assaulted her repeatedly, and let her go. The case went cold. In the fall of 2023, state police partnered with Northern Michigan University’s Cold Case Program, determined the case was a candidate for forensic genetic genealogy, and sent evidence to a private lab with the MSP Forensic Science Division.

The lead pointed to a 78-year-old man in Argonne, Wisconsin, who had lived in Harris, Michigan, about 45 minutes from the crime scene, in 1986. Working with the Wisconsin Department of Justice and the Forest County Sheriff’s Office, detectives got a court-authorized DNA sample. It matched. The Delta County Prosecutor authorized kidnapping and first-degree criminal sexual conduct charges, and the arrest was set for July 29, 2025. He was found dead at his home two days earlier.

Police withheld his name because he died before arraignment. The case was not announced publicly until June 2026, nearly 11 months later. State police say it is believed to be the oldest sexual assault case in Michigan solved through genetic genealogy, and the survivor has been told who attacked her.

The detectives who closed these cases did the work, and a survivor and a family got answers they had been told for decades they would never get. The harder lesson is about time. When an investigation is built to end in a charge, and the suspect is an old man who has had 40 or 57 years to think about it, the gap between “we know” and “we arrest” is a window, and in two recent Michigan cases that window closed on the wrong side.

What a Dead Suspect Does to a Missing-Person Case

In a homicide where the body was found, a suspect’s death closes the courtroom but not the evidence. Sheri Jo’s case could still be solved after Collins died, because her killer’s DNA was on file and could be matched at autopsy.

Oscoda is the opposite kind of case. There are no remains. Investigators said in 2023 that the case was solvable because people who know something were likely still alive, and that tips from the community were the way in. When the only person of interest the state has publicly confirmed dies, the case loses its most important potential witness, and it loses the one conversation that might have told two families where to look.

Let me tell you what I keep coming back to. Collins was questioned in 1969. He was a free man in mid-Michigan for 56 years after that. He was still playing bars around Flint in his seventies. And for most of that time, the evidence that would eventually name him in Sheri Jo’s death was sitting in a file. If that evidence had been tested and run against a short list years earlier, investigators would have had something they never had in the Oscoda case: leverage, sitting across a table from a living man.

Why This Matters Beyond Oscoda

The university programs deserve real credit here. WMU students digitized both the Elliott and the Oscoda files. NMU students helped crack Delta County. In both programs, the breakthrough came from someone finally doing a full inventory of what was in the box. That should not take a student practicum. It should be the first step of every reopening, at every agency, and it should happen on a schedule rather than when a grant or a class makes it possible.

There is also a policy gap in how Michigan handles a suspect who dies before charges. One man was named; another was not. Families and survivors deserve a consistent standard, and so does the public, which in Oscoda is still being asked to supply the tip that cracks the case. If investigators believe another case may be connected to a dead suspect, the public needs enough information to know what to call in about.

And there is the question every Michigan cold case unit should be asking right now. Which other files hold untested biological evidence and a person of interest who is in his seventies or eighties? Every one of those cases is on a clock that no amount of later lab work can reset.

Pam Hobley would be 72. Patty Spencer would be 73. Pam’s parents died without knowing where she is. Her sisters are still here, still asking. If you were in Oscoda in the fall of 1969, if you ran with the crowd that went to those parties, if you knew a band that played Hale and Oscoda that year, or if someone ever told you something about that Halloween that you have carried since, this is the year to say it.