Friday, July 11, 2008
‘110 percent certain’
Arlo Olson: ‘Half shot’ yet positive in his testimony about what he saw
CONTENT WARNING
This series contains references to vulgar language and violent acts that may be objectionable to some readers and that parents may find inappropriate for their children.
Editor’s note: This series is the product of a six-year investigation by former Daily News-Miner reporter Brian O’Donoghue and his journalism students at the University of ֱֻ Fairbanks, with support from the News-Miner.

Between 1995 and 2000, Arlo Olson was convicted seven times for assaulting his girlfriend and twice for assaulting people who tried to protect her. In 2000, he served time for hitting and choking his mother. Olson, who provided key testimony for the state, said in 2003 in a jailhouse interview that he was not sure what he saw down the street from Eagles Hall.
He’d been drinking for hours by the time he eyed four men fleeing a robbery.
Those standing alongside later swore they didn’t notice the crime unfolding more than a block away. A perception expert argued that no one can recognize faces at such distance.
Yet jurors believed the man from Kaltag. Arlo Olson sounded that good on the witness stand.
Then-prosecutor Jeff O’Bryant underscored Olson’s significance in the final trial of John Hartman’s four accused slayers.
“Simply put,” he told jurors weighing murder charges against Kevin Pease and Marvin Roberts, “if Arlo didn’t see what he saw, and you throw out some of the state’s evidence, the state doesn’t have a case. No doubt about it.”
Others, including a nationally recognized authority on criminal identification, remain troubled by the weight attached to identifications made under remarkably unfavorable conditions. “The idea an identification of this sort could still result in a conviction is rather astonishing,” said Gary L. Wells, an Iowa State psychology professor and former chair of several federal panels on eyewitness evidence
Yet jurors not only found the hard-drinking 20-year-old credible, in doing so they discounted a sober, expectant mother’s time-referenced alibi for Roberts, the driver of the car that Olson identified as fleeing that distant robbery.
Though he, too, shared the Athabascan heritage, Olson’s starring role in all three Hartman trials fed doubts — and, for many Interior Natives, perceptions of racism — that a decade of legal wrangling hasn’t overcome.
Linked robbery with murder
Fifteen-year-old Hartman lingered in a coma until his death Sunday evening, Oct. 12, 1997. By then, police had four suspects, ages 17 to 21, including two — George Frese and Eugene Vent — who confessed.
Earlier that day, Detective Aaron Ring stopped Olson on the street, mistaking him for someone else. Hearing that he, like the suspects, had attended Friday’s big wedding reception, the detective urged Olson to come by the station to give a statement.
But Olson was reluctant; he had assault charges pending from two alcohol-fueled brawls back
in his village. When police later sought him at his grandfather’s apartment, he hid.

Five days after the discovery of Hartman sprawled unconscious on a downtown street, Olson bowed to his grandfather’s requests and visited the police station. He volunteered that he had seen the murder suspects jump another guy minutes earlier. He said he recognized the group from the mug shots published in the newspaper.
Police had already characterized the robbery of Franklin Dayton as the start of a violent spree culminating in Hartman’s murder. Now they had a witness tying the murder suspects to that earlier assault.
Adding to Olson’s significance, the witness also said he had run into Frese outside a nearby bar about the time Hartman was rushed to the hospital. Frese was limping and acting “jittery,” Olson told police. He said the suspect also bragged that he had “kicked a guy’s ass.”
Three years after delivering guilty verdicts against Roberts and Pease in the third and final Hartman trial, a pair of jurors discussed Olson’s impact with University of ֱֻ Fairbanks journalism students.
“The key thing was the eye-witness,” juror Gary Montini said in 2002. “I forget the guy’s name, young guy that was out on the steps of the hall, he identified them while they were beating the crap out of some guy.”
He called Olson “courageous” for speaking out against fellow ֱֻ Natives.
Jamie Smykalski found Olson particularly credible compared to the suspects’ witnesses. “Every single person that came up to speak for the defense reeked of disbelief,” the juror said. “This Arlo guy is either the world’s best liar, in my opinion, or he saw what he saw. He was very convincing.”

Arlo Olson was standing in front of Eagles Hall (top) with a group of people when he says he saw Franklin Dayton being robbed down the street. Dayton told police he was mugged just past the Poldine Carlo building (bottom). The distance between the two locations is 550 feet. Four days later, Olson identified the men he says he saw robbing Dayton as George Frese, Kevin Pease, Marvin Roberts and Eugene Vent — after he saw their mugshots in the newspaper. Police contend the group continued on in Marvin Roberts’ car and fatally assaulted John Hartman roughly a half-mile away.
Vague about vantage
Olson couldn’t actually see the muggers’ faces from where he stood under the canopy fronting the Eagles Hall. That would have required binoculars, he said under cross-examination. But Olson said he had seen them all riding in the car earlier that night and was “110 percent certain” he recognized Frese, Vent and the others committing the robbery based on their haircuts, profiles and posture. “That’s like you see Michael Jordan a half a mile away,” Olson testified, “you’re going to recognize that’s Michael Jordan.”
The witness remained vague about the distance separating him from the mugging. He gave grand jurors the impression Dayton’s assault took place in a nearby parking lot. Testifying at the first Hartman trial, 16 months later, Olson said: “I don’t know, maybe a 100 yards, 150 yards,” and characterized it as a “football field.”
It didn’t come out until the third trial, 22 months after the crime, that Fairbanks police had gone back to measure how far Olson stood from the sidewalk section pointed out by the mugging victim. Using a roller device, then-Sgt. James Geier measured the distance at 183 yards, or 550 feet.
Geier, who prepared step-by-step reports on other aspects of his role in the Hartman investigation, left this field survey out of his case file. “They asked me to go do something,” he explained under cross-examination. “I did it, and let them know what it was, what the results were.”
He couldn’t recall who wanted the measurement or who he called with the result. “It probably was somebody with the police department.”
Roberts’ defense attorney, Dick Madson, took aim at that omission. “You didn’t want us to know about it,” he said. “Did you?”
“The place is still there, sir,” countered Geier. He added that Olson’s account seemed plausible when he gazed back toward Eagles Hall from Dayton’s mugging location. “I thought well, yeah, I could see the place fine.”
How far is too far?
Geoffrey Loftus, a University of Washington psychology professor and visual perception expert, contends identifications at such great distance are not reliable. “You simply wouldn’t be able to distinguish anything about what that person looks like,” said the professor, who testified as a paid expert for Pease.
Even under the best lighting and with no time pressure, the professor assured jurors, an alert observer’s chance of recognizing someone’s face farther than 200 feet away is “essentially nil.”
Olson wasn’t talking about faces, however. He said he based his identifications on the haircuts, posture — the suspects’ general appearance.
There’s a real difference, Loftus acknowledged last fall.
“Indeed when you see an entire body, there’s more information than in a face,” he noted by e-mail. “Body shape and motion come into play.”
But the professor contends body recognition lacks the certainty attached to facial identification. “It’s pretty easy to tell one human’s face from another; one person’s face is unique or almost unique in that respect.” Body builds and postures are more interchangeable, he pointed out. “Determining someone is 6-foot-2 and 200 pounds doesn’t really distinguish any one individual from many other individuals.”
Stung by the trial verdict, Loftus and a University of California at Los Angeles colleague later experimented with subjects attempting to distinguish facial types and celebrity photos at a distance of 200 feet. Their findings were published in a 2005 article in Psychonomic Bulletin and Review, a journal of experimental psychology.
The article features images of actress Julia Roberts’ face, altered to demonstrate the effects of distance, taking into account light frequency, contrast degradation and other visual considerations. At 172 feet, the movie star’s image blurs past recognition.
The decade since Hartman’s murder has given justice officials and researchers new cause for scrutinizing eyewitness testimony, particularly in murder and rape trials. Misidentification by claimed witnesses is the leading common factor among DNA-driven exonerations, offered as evidence in 75 percent of convictions later overturned through science.

Geoffrey Loftus, a University of Washington psychology professor and visual perception expert, maintains that identifications at distances such as 550 feet are not reliable. “You simply wouldn’t be able to distinguish anything about what that person looks like,” said the professor, who testified as a paid expert for Kevin Pease.
All of those proven errant IDs, according to Wells, came from observers who stood closer, enjoyed better lighting and other circumstances more favorable than Olson possessed that night outside the Eagles Hall. “Those witnesses had not been drinking and so on,” the professor noted by e-mail after reviewing this story. “And yet they mistakenly identified innocent people rather than the perpetrators.”
Hunters know better
Out in the villages, no one needed light frequency calculations to assess the witness’s credibility. “How could he see 500 feet? During the evening, you know, in the dark?” demanded Gerry Roberts, the father of the alleged getaway car driver.
Hunters who have attempted to distinguish bull from cow moose in twilight, even using a rifle scope, will grasp the absurdity, he said in a 2004 interview.
“Five hundred feet is a long way to see moose, too, you know? Somebody you could recognize?”
Jurors from a village would have known better, he insisted.
Four years ago, ֱֻ journalism students discovered that members of one Hartman jury wanted to see for themselves. The entire panel left the Anchorage courthouse during deliberations for a field test. Several jurors paced off what they took to be the distance at issue. Others eyed them, weighing Olson’s claims.
Neither Judge Ben Esch nor any of the attorneys knew about the excursion that sunny afternoon. Disclosure of the unauthorized experiment held clear legal implications, a pair of former ֱֻ attorneys general said in interviews in February 2003.
“It’s totally improper,” said Charlie Cole, a Fairbanks attorney who served as the state’s chief lawman under Gov. Walter Hickel. “I suspect it’s grounds for a new trial.”
Cole’s successor, Bruce Botelho, agreed. “You’ve got jurors that have, on their own, tried to reconstruct a situation that clearly could not approximate the setting.”

Top to Bottom: George Frese, Kevin Pease, Marvin Roberts, Eugene Vent
In August 2004, Judge Esch overturned Pease’s conviction due to the jury’s experiment. Though the same facts apply, Roberts’ conviction stood; ֱֻ law only allows one opportunity at overturning a verdict through new evidence.
Last July, two of ֱֻ’s three appellate judges concluded that jurors simply tested the vision expert’s general assertion that, even under optimum conditions, faces are unrecognizable 200 feet away. Since this was the purpose, Judge David Mannheimer wrote for the majority in its ruling, “there is no reason to believe that the experiment yielded a false or misleading result.”
Chief Judge Robert Coats seconded that view. “Jurors engaged in a sensible effort to resolve the credibility of Dr. Loftus’ testimony,” he wrote. “They understood both the significance and the limitations of their experiment.”
The ruling blocked the new trial ordered for Pease.
In a dissenting opinion, Judge David Stewart pointed out jurors were weighing Olson’s entire account, rather than solely examining the vision expert’s remarks. And their experiment fell well short of duplicating the witness’ vantage on Dayton’s robbery, the judge noted. “The jurors were looking at each other during the daytime at a distance more than 100 feet less than Olson faced,” he said, referencing accounts provided by the participants.
“Because such an experiment is a constitutional violation,” he argued, “a conviction cannot stand unless the court can conclude, beyond a reasonable doubt, that the misconduct did not contribute to the verdict.”
Wells, who has been testing witness reliability since the 1970s, suggests the jurors made a “classic error” in their street experiment.
“If you expect to see a particular person at a distance, you can have the impression that you in fact can recognize them,” he explained by e-mail. “But you know it was going to be them, you knew what they were wearing and so on. What if, unbeknownst to them, someone different had been substituted for their fellow juror?”
The significance of the jury’s experiment is now before the ֱֻ Supreme Court, along with Roberts’ constitutional arguments that he, too, deserves a new trial if Pease prevails.
Only ‘half shot’
Walking to the reception that Friday, Olson joined several friends swigging from a fifth of Bacardi. His pregnant 17-year-old girlfriend wasn’t drinking.
The couple reached Eagles Hall about midnight. She itched to dance. He wasn’t ready. “I don’t like to dance unless I’m high,” Olson explained in court.
Guys kept buying Olson’s younger sister mixed drinks. He downed “two or three” of these. He also sipped Mountain Dew spiked with the remaining rum.
By about 1 a.m., he recalled, “I wasn’t drunk. My speech wasn’t slurred and I wasn’t staggering. I still didn’t want to dance.” Or, as he put it during another court appearance, “I was high. Yeah. Was half shot.”
In this state, he headed outside for a cigarette, joining a crowd of about 30 gathered by the front canopy.
“Look, fight,” Olson recalled a friend shouting. Glancing up, he told police, he saw Dayton knocked to the ground and kicked. “Give us your money, bitch,” he recalled hearing someone yell. The victim handed up his wallet, the witness swore.
And he said he recognized Vent and Frese dashing toward a little blue car.
“How long did that take?” asked Pease’s attorney, Lori Bodwell. “When you saw them running to the car?
“Probably three.”
“A few seconds?”
“Three or four seconds, yeah.”
Police never found anyone else who acknowledged seeing the robbery.
The friend Olson said directed his attention to the fight, Kaltag basketball coach Robert Nicholas, recalled standing alongside Olson as people got in and out of cars. But he didn’t remember watching or discussing a fight.
Local housing officer George Yaska, another person Olson said stood nearby, testified that he was inside dancing when word about Dayton’s mugging spread inside the hall. Yaska, who didn’t recall bumping into Olson at all, declared that he would have rendered assistance. “I usually end up jumping in anyways,” he said. “You have to.”
The victim, meanwhile, saw his own robbery differently.
Dayton, who like the others had had plenty to drink, didn’t remember handing up his wallet, as Olson described. He recalled being belly down on the sidewalk and feeling his pocket picked. Looking up as the car sped away, he glimpsed a good-sized, white or tan vehicle. He sketched it for police as a four-door model.
Olson returned inside before Dayton made his way back to the hall. His girlfriend caught him dancing with another woman sometime before 3 a.m. and stomped out. He and another friend moved on to the Elbow Room. From there, the pair continued to the ֱֻn Motor Inn, where Olson started a food fight that carried into the parking lot.
Afterward, the witness sipped 100-proof Wild Turkey at another friend’s apartment before landing on his grandfather’s couch about 5 a.m.
‘Party poopers’ table’
Critics familiar with the people involved have trouble understanding why jurors, aware of Olson’s state of intoxication, apparently discounted the observations of a woman who avoided the alcohol flowing so freely that night.
That Friday, Tanana resident Mary Ann Wiehl had kids to watch. It was 11 p.m. before she escaped to the reception. She found a table with Gareth and Carry Edwin, a Nenana couple, and Eileen Newman, a 28-year-old Alyeska Pipeline Service Co. clerical employee then engaged to Gary Moore, a contributing student reporter on this series.
Being pregnant, all three women shunned alcohol. Welcome to the “party poopers’ table,” they joked together.
Marvin Roberts, whom Wiehl babysat as a child, settled on the table’s opposite bench, the women recall. Wiehl and the others weren’t consciously watching the 19-year-old. They didn’t notice, for example, that he exited the hall at least four times and gave friends quick rides. But his dance partners — the sister of one, then the niece of another — periodically drew their attention.
Approaching 1:30 a.m., Newman asked a younger woman to accompany her to the ladies room, “I can’t,” she reportedly said. “Marvin just asked me to dance.”
Newman went with Carry Edwin instead. On their way to the restroom, the pair passed Dayton’s relatives reporting his robbery using the hall’s payphone. They saw the injured man at the foot of the stairs, surrounded by concerned friends.
That 911 call, logged by the dispatch center at 1:35 a.m., coincides with the approximate time frame police established for Hartman’s fatal beating a half-mile away.
When she returned to the table, Edwin shared what the women heard from the victim’s sister. Roberts was present for that news, her husband Gareth later recalled.
Wiehl said she phoned police soon after Roberts’ arrest. “There’s no way he could have left, beaten somebody to death, and then come back like nothing happened,” she recalled telling police.
She got the impression that police were receiving similar calls from other reception guests. “I was asked if I was related to him,” said Wiehl, whose account can’t be verified in available police records. She said she sensed the interviewer’s attitude change as she outlined a family connection through her grandfather in another village. “Basically, the officer told me we all stick together.”
Weeks passed. Edwin mulled over that 911 call. She and Newman compared recollections about Roberts’ presence relative to that ladies room stroll.
Newman didn’t know Roberts that well. She grew up in Rampart, a tiny Yukon village. Roberts was her younger brother’s contemporary and was said to be a quiet, studious boy. She couldn’t vouch for that, but she could pick Marvin out in any crowd.
You don’t forget faces from the closest village downstream.
In August 1999, Newman explained to jurors how she knew Roberts wasn’t behind the wheel of that getaway car.
Roberts’ conviction came as a slap.
“If they think I was lying, they don’t know me,” Newman told UAF students in a videotaped interview. “And I would have to be lying.”
Jailhouse retractions
An assortment of offenses landed Olson behind bars for weeks, then months and, more recently, years, following the Hartman trials in 1999.
While serving time in 2000, Olson disavowed portions of his trial testimony in conversations with Frese’s defense attorney, Robert Downes, who now serves as a judge, and with an investigator working for Roberts and Pease. The inmate declined their requests to tape those conversations.
Ring soon visited. “I was talking about recanting,” the detective said Olson assured him, “because I wanted out of here.”
In May 2001, the witness backed that up under oath. At a hearing largely prompted by his wavering, Olson characterized the jailhouse retractions as lies motivated by his desire to get out of jail. Judge Esch found the witness’ conflicting statements disturbing but concluded that Olson’s sworn trial testimony deserved more weight.
Olson again sounded off in a series of jailhouse interviews stretching from December 2002 through April 2003. These interviews were conducted both face-to-face at Fairbanks Correctional Center with a reporter and through calls initiated by the inmate. A half dozen of the telephone sessions were recorded for this story without Olson’s knowledge, a practice lawful in ֱֻ by a person taking part in a conversation.
“I was threatened with perjury,” he said in an interview recorded in January 2003, “and then I was threatened with going to jail and them sending the troopers out to get me.
“I didn’t want to testify,” he added. “I told them I wasn’t sure. And they kept showing me bits and pieces (of the interrogation statements). I guess to make me, you know, feel sure of what I was doing. And it did.”

Curtis Sommer discusses the Hartman case in August 2004 with other members of the Marvin Roberts Project, a tribal justice group, which spent weeks reviewing trial transcripts in the basement of Tanana’s village health clinic.
Getaway car perspectives
Olson also discussed on tape how he came to identify Roberts’ blue hatchback.
“I said a beige car,” he recalled. “Then they brought me up and showed me Marvin Roberts’ car. Then it was, ‘No one else drove that car but Marvin Roberts.’ You know?’ He (Ring) kept saying that. We went through that three or four times before he decided to put it on the record and tape it.”
Police confirmed during trial that Olson identified Roberts’ blue car through viewing a single vehicle inside the station.
That amounts to a “tainted lineup” contends William Satterberg, a local defense attorney with no connection to the Hartman case. “You’re showing the guy one car in the police garage. He knows why it’s there.”
While it may not have been practical to present the witness with a group of similar vehicles, the attorney said, Olson’s identification of the car would be more credible if it resulted from reviewing photos offering a comparable selection of cars.
Wells, the eyewitness evidence researcher, concurs. “All tests of identification should be conducted using the lineup-type method in which there are fillers,” the professor commented by e-mail. “This is just as true of vehicles as it is with people.”
He regards the garage showing “highly suggestive” and hardly a fair test of Olson’s memory. “It is perhaps merely a test of the witness’s willingness to cooperate with and corroborate the theory of the police.”
As recently as August 2006, Ring denied the witness ever wavered on the getaway car’s color. “Arlo Olson never stated anything about a beige car,” the detective stated in an affidavit filed opposing Vent’s ongoing bid for a new trial.
The detective had no trouble buying Olson’s account of the distant robbery, he said in a 2002 interview, because earlier that same evening the witness had encountered the same group, riding in the same car. “So he saw them. Basically face to face.”

Kathy Mayo Roberts, far right, refers to a map of downtown Fairbanks posted on the clinic wall. With Roberts is Julie Roberts Hyslop.
In 2004, residents of Tanana launched their own investigation. After several weeks sifting trial transcripts in the basement of the local health clinic, members of the tribally sanctioned Marvin Roberts Project zeroed in on the same exchange outside the Eagles Hall.
The witness said that a little two-door blue car rolled up alongside him during an earlier cigarette break. “George leaned out the passenger side window and asked if I wanted to get high,” Olson testified at one trial. “He was kind of leaning up through the window,” he said during another trial appearance.
Here’s the clue that sent Tanana’s amateur detectives in hot pursuit: Roberts, they realized, owned a hatchback model ‘92 Dodge Shadow.
Olson placed Frese riding in the rear of the hatchback, sitting directly behind wide-shouldered, 200-pound Pease.
“That rear window doesn’t open,” former tribal judge William “Pongee” Carlo pointed out to a visiting reporter. “So how is he leaning out of it?”
This February, police allowed the News-Miner to inspect Roberts’ car, which remains in city storage as evidence. The hatchback’s side rear windows are indeed immobile. However, the 54-inch-long front passenger door extends far enough back to allow a rear passenger to poke his head into its window opening. If Olson’s account is reliable, big Pease would likely have to have cooperated inside the car by tilting his front passenger seat forward.
‘I am Spartacus’
Setting aside the merits of the verdicts, the conduct of the Hartman trials left many Athabascans seething about perceived racist stereotypes.
“Assumptions were made about Native people, in general, sticking together. That was very hurtful,” declared Shirley Lee, then the executive director of Fairbanks Native Association, speaking outside the courthouse at a rally staged on the 10th anniversary of Hartman’s death and the suspects’ arrests.
The tribal detectives’ effort extended to documenting perceived anti-Native bias in the last trial.
Asked for an example, Carla Klooster, a non-Native council employee and 25-year-Tanana resident, reached for a sticky note. “Page 3,145, line 17,” she said. At least five people testified they saw Roberts around 1:30 a.m. He was out on the dance floor with this girl or that girl. He gave a couple a lift or fetched a woman’s orange juice.
Prosecutor Jeff O’Bryant, who recently retired from the Fairbanks District Attorney’s office, made the most of those overlapping sightings. “Mr. Roberts has got some problems with his alibi,” he pointed out to the jury. “He can’t be in all those places at once.” The prosecutor implied it wasn’t simply confusion.

Prosecutor Jeff O’Bryant, seen here in an unrelated 2007 case, caused a ripple effect through the Native community during Marvin Robert’s trial by comparing Native loyalty to the slave scene in “Spartacus.” O’Bryant’s comments came as a response to witnesses saying Roberts was seen dancing at a wedding reception at Eagles Hall the night of John Hartman’s fatal assault.
“It reminded me,” he told jurors, “of the movie where the Romans have a bunch of prisoners, slaves, and there’s an uprising amongst the slaves because of the conditions. And the leader of the uprising, apparently, was Spartacus.”
When the Romans came looking for Spartacus, O’Bryant observed, “much like the witnesses here” slaves stepped forward declaring “I am Spartacus,” one after another.
In the village clinic basement, heads nodded as Klooster reread the passage.
The dramatic scene from director Stanley Kubrick’s 1960 Oscar-winning movie still inspires frequent reference in popular culture; the slaves’ self-sacrifice is commonly held up as embodying true loyalty.
But the prosecutor’s argument wasn’t taken that way in Tanana, recalled Klooster in an interview this past December. “What’s the basis of the loyalty?” she said, pointing out that while people standing up for Roberts knew him, most weren’t close friends. “It’s basically saying the loyalty stems from cultural identity; that one Native will back up another Native.
“That’s such a ridiculous statement,” added the former tribal administrator, citing decades of experience observing life in a small Athabascan community, “because it’s so absolutely not reflective of Native culture. Natives don’t back up other Natives just because they’re Native.”
In an interview in May 2007, O’Bryant declined to discuss the reactions to his Spartacus comparison.
As a young man, he double-majored at ֱֻ in Yupik Eskimo and linguistics. His professional silence toward accusations of bias leveled in letters and tribal resolutions shouldn’t be taken as indifference. “It rankles me more than you can possibly know,” O’Bryant said, leaning forward, eyes wide. “It rankles me.”
Downward spiral
Olson, between 1995 and 2000, was convicted seven times for assaulting the same girlfriend and twice for assaulting individuals who sought to protect her. In 2000, he served time for hitting and choking his mother.
A year later, he landed in jail for punching a Unalakleet woman.
In August 2002, police arrested him on a charge of drunken driving, his second such charge in a month. In one incident, both of the couple’s kids were in the car when he rear-ended another vehicle.
That November, Olson kicked in his girlfriend’s apartment door, held a knife to her throat threatening her life and, according to the indictment against him, kidnapped her for several hours.
Troopers rescued the woman.
At the time of his arrest, Olson’s blood-alcohol content registered 0.281, more than triple ֱֻ’s intoxication standard. He had downed two kinds of prescription pills.
His life was ruined, Olson sobbed to troopers in a taped interview. He blamed stress from the Hartman case.
“A few years ago I testified as the state’s key witness in the John Hartman murder case,” Olson told Judge Richard Savell at his April 2003 sentencing on a charge reduced through a plea bargain. “I did what I thought was my civic and patriotic duty. Most of all, I did what I thought was right.”
In a voice cracking with emotion, Olson described life as an outcast. “No longer am I welcome at Native gatherings, functions, places of work and homes.”
He urged the judge to consider his special burden. “Jail time imposed on me will be hard and unbearable,” he said. “I’ll have to do it in solitary confinement. I am known by everyone as a rat and as a snitch.”
“What has changed,” Savell countered, “since the first assault against the female in January 1996? I can’t point to anything,” the judge said. “No improvement, only escalation, increased frequency and more of it.”
He sentenced Olson to four and a half years in prison, with an additional five years suspended as incentive to reform.
Former Yukon Kuskokwim Schools superintendent Glenn Olson approached as a bailiff slipped handcuffs on his son.
“God love you,” he called softly.
Mugging linked to murder by distant witness
1) Mugging victim Franklin Dayton told police he was knocked to the ground on the sidewalk just past the Poldine Carlo Building, near a parking lot entrance. Police measured the distance separating the victim from witness Arlo Olson — 550 feet — but left it out of case reports.
4) Police theorized that the group responsible for Dayton’s robbery fled in a car down 1st Avenue, fatally assaulting John Hartman minutes later at 9th Avenue and Barnette Street.

A road map showing a portion of downtown Fairbanks
2) Arlo Olson was standing in a crowd on the front landing of Eagles Hall when he observed Dayton’s robbery. Four days after the crime, Olson identified the murder suspects from newspaper mug shots.
3) 911 alibi? One of the men Olson identified, Marvin Roberts, was observed inside the hall near the dance floor when Dayton’s robbery was reported via 911. Others, however, said they saw Roberts elsewhere at the time, a point that the prosecution said undermines the alibi claim.
Tomorrow:
Brian O’Donoghue is a UAF assistant professor of journalism. Former students Tom Delaune, Laurel Ford, Gary Moore, Robinson Duffy, Kelsa Shilanski and Sharice Walker contributed to this story.
Let the music play: A night on the dance floor comes into question
Marvin Roberts’ claimed alibi rests upon observers who recall him dancing at the Eagles Hall at the approximate time John Hartman suffered his fatal assault.
But was anyone dancing during those fateful minutes?
According to the prosecution’s final pair of witnesses, the band was on break when a 911 caller reported an unrelated robbery near the Eagles Hall.
Nick Monroe, a member of the band playing that night in October 1997, testified that he was on break when he first heard that someone associated with the reception had been kicked or hurt. Reception guest Claude Andrew Koyukuk also remembered the band taking a breather as he saw relatives bandaging Franklin Dayton, who’d staggered into the hall following a mugging.
Others described Roberts dancing or heading for the dance floor just before Dayton’s return caused a stir.
Gareth Edwin, of Nenana, described Roberts among those listening as word spread of the crime.
Eileen Newman’s account encompassed both aspects. Her niece begged off a trip to the ladies room to dance with Roberts. On her return, Newman testified, she bumped into Dayton’s relatives using the pay phone for that 911 call.
Citing the band’s break, prosecutor Jeff O’Bryant challenged Roberts alibi. “He can’t be dancing when there’s no music.”
Defense attorney Dick Madson reminded jurors another witness mentioned hearing about the mugging as he came off the dance floor.
“What does that mean?” Madson asked. “He wasn’t out there dancing when there was no music going on. It means the band was playing.”
By all accounts, few in the rocking crowd kept close watch on the clock. Only one event taking place within the hall that night is time certain: Police at 1:35 a.m. logged the 911 call reporting Dayton’s assault. That roughly coincides with Hartman’s fatal beating a half mile away.
Hours of recorded police and emergency traffic survive in the Hartman case files. Audio of that particular 911 call might shed light on the band’s activity. However, that call wasn’t found on case tapes reviewed for this story.
— Brian O’Donoghue
ABOUT ‘DECADE OF DOUBT’
This seven-part series offers no proof of guilt or innocence. It does document gaps in the police investigation that raise questions about the victim’s last conscious hours. It points out that the group convicted of John Hartman’s murder may have been prosecuted with forms of evidence identified later in national studies as contributing to some wrongful prosecutions elsewhere. And it shows how rulings from this state’s courts have undermined ֱֻ Native confidence in the justice system by keeping juries from weighing all that’s known about the crime.
Among the series’ observations:
- The police investigation remained focused on suspects flagged through a pair of confessions, subsequently retracted, despite lab tests that yielded no supporting evidence.
- Jurors remained unaware that state crime lab experts couldn’t match Frese’s boots with photos of Hartman’s bruises. Though it bore the lab’s logo, the suggestive exhibit presented at trial was a non-scientific photo overlay assembled by police and the district attorney. Recent studies have shown that evidence lacking forensic merit often figures in convictions that are later overturned.
- Detectives referred to fictitious evidence throughout the interrogations that yielded confessions from Vent and Frese. Employing such trickery on suspects who profess no memory of a crime, while standard practice in 1997, today draws specific cautions in the nation’s standard-setting criminal interrogation manual. The revisions reflect lessons learned from re-examining tactics used obtaining confessions later proven false in cases that sent innocent people to jail.
- The state’s case strongly relied upon identifications made by an eyewitness standing 550 feet from a robbery. The distance raises the possibility of witness misidentification, which has emerged as the leading common denominator among hundreds of errant murder and rape convictions.
- Police paid scant attention to the last person known to have been with Hartman. Chris Stone, a 14-year-old self-described methamphetamine addict, had been hospitalized following a similar assault only weeks prior. And jurors never heard about Stone’s attention-getting entrance into Carrs-Foodland about the time Hartman lay dying in the street. Also, no one involved in the Hartman case had access to Stone’s sworn statement, sealed in an unrelated juvenile proceeding, suggesting, under one interpretation, awareness of his friend’s plight.
All of this has contributed, in the eyes of many, to a decade of doubt.