Only three people know what happened on that Toronto street on the night of August 31, 2009. One is dead, and the other two aren’t talking publicly until the trial is over, if they ever will. The best version of events the rest of us can put together is this: At about 9:45 p.m., the stretch of Bloor Street through Yorkville—a chic downtown neighbourhood and shopping destination—is quiet. A black Saab convertible heads west, a man behind the wheel and a woman at his side. They approach a cyclist, also heading west, and there’s some kind of minor collision. An argument ensues, escalates, and the cyclist ends up hanging off the side of the car as the driver hits the gas. The convertible veers wildly around a crew of workmen in the centre lane and into the eastbound lanes, with the cyclist still clinging to it. The Saab smashes against roadside objects—first a tree, then a mailbox. The diversion may have been intentional or maybe the driver lost control, but eventually the cyclist falls into the path of the Saab’s rear wheels.

Paramedics rush him to hospital, where he soon dies. The driver, meanwhile, sits in the back of a police car a block from the accident site; he looks bewildered, dismayed, disbelieving. He knows the storm that’s coming.

Within hours, The Globe and Mail and the Toronto Star run brief online stories about the incident, describing it as a hit and run. Neither names the driver, though speculation about his identity is already making the rounds. Just after 5 a.m. the next morning, 680News takes a chance and goes with the circulating rumour—the man in police custody is Michael Bryant, former Ontario attorney general and then-CEO of Invest Toronto, an agency that promotes the city as an international business centre. He’s a political showman—a former “cabinet rock star,” according to the National Post—who’d been returning from a night on the town when he was involved in a confrontation that ended with the death of cyclist Darcy Allan Sheppard, a 33-year-old bike courier. It was a dramatic culmination of the increasingly angry, occasionally violent conflicts between drivers and cyclists that had dominated city streets and headlines all year.

As night turned to day in Toronto’s newsrooms, phones rang and inboxes pinged, a journalistic reveille rallying the incoming army of editors and writers. “It got pretty exciting pretty fast,” remembers Post reporter Matthew Coutts. National editor Scott Stinson roused him from bed at 7 a.m. and told him to get down to the police station where Bryant was being held. “Right off the bat it was all hands on deck,” says Coutts. “A lot of people get into journalism for that rush.”

Editors felt the same rush as reporters. “I shouldn’t call it a great day,” says Kelly Grant, then-Toronto editor for the Globe. “But when a story breaks that everyone wants to read and I have this stable of incredibly talented reporters I can throw at it, that’s not a bad day. That’s a great day.”

Unsubstantiated rumours quickly percolated through newsrooms—the dead man had been homeless, Bryant had been cavorting with a mistress, Bryant had been drunk. The pop culture allusion on everyone’s lips wasThe Bonfire of the Vanities, Tom Wolfe’s satirical novel about a Wall Street investment banker who runs down and seriously injures a young black man. “The dream was, let’s find the ritzy restaurant in Yorkville where Bryant ate and count how many bottles of expensive wine were on the table,” says Star city editor Graham Parley. “It would have been very sensational to find him drinking with some mystery woman before jetting off in his luxury convertible. You know, ‘What a story! Mmm, yeah!’”

The working theory was that Bryant, drunk on power and privilege (and booze, of course), possibly cavorting with a mistress, had killed a down-on-his-luck cyclist in a fit of road rage. “Newsrooms are famous for putting together a working theory,” says Larry Cornies, a former editor of the London Free Press. “We drive a story forward and make guesses as to what the next development will be. Ninety-five percent of the time we’re right. But not always.” Not this time. “In a crude way,” says Parley, “you could say we went looking for dirt and didn’t find anything.”

***

Over the next few days it became apparent that Bryant had actually been on an unexpectedly modest anniversary date with his wife, lawyer Susan Abramovitch, eating shawarma before grabbing some dessert and heading home. They consumed no alcohol. It wasn’t a hit and run, either. After Sheppard fell to the street, Bryant drove a block or so and pulled into the driveway of the nearby Park Hyatt hotel, where he called police.

And Sheppard, the Alberta-born Métis initially seen as a “merry warrior in the often intense subculture of bicycle couriers,” as Globe columnist Judith Timson described him, became more complicated as well. A friend told reporters he’d been drinking on the night in question and had ended up in the back of a police car after an altercation at his girlfriend’s apartment. He was also wanted on dozens of outstanding arrest warrants in Alberta.

Within a day, the Bonfire narrative looked shaky. Within two, it was pretty well discarded. Some observers attributed the story’s 180-degree turn to the machinations of Bryant’s communications firm, Navigator Ltd., hired just hours after the accident and almost invariably referred to in those first days as a “blue-chip” PR firm. (Senior partner Robin Sears was Brian Mulroney’s spokesman during the Oliphant inquiry.) Online, citizen journalists and cycling advocates worked to counter what they saw as a mainstream media smear campaign against Sheppard, boosted by Navigator. But reporters insist that the story’s turnaround was the result of old-fashioned shoe leather, good luck and nothing more.

Either way, the Bryant-Sheppard story proved that complicated, sensational stories don’t have to be journalistic debacles. That’s reassuring, considering that the mainstream media’s tendency to exploit flimsy leads, go-nowhere speculation and internet gossip (all rife in this case) has resulted in some appalling journalism. There are plenty of examples of what media pundits call “confirmation bias”: the readiness to report a story based on preconceived assumptions, accurate or not. In this case, it was new media outlets (YouTube and blogs, mostly) that twisted themselves into paranoid, judgmental contortions and did more to distort the story than to clarify it. Old media (print and broadcast) journalists made mistakes too, but by and large they kept up with the twists and turns. Most of the reporting unfolded in a credible and nuanced way. The biggest question in the aftermath is whether the success was thanks to good reporting, good luck or the guiding hand of a public relations firm.

The morning after the accident, 25-year-old Star reporter Robyn Doolittle received a wake-up call from her editor at around 5:15 a.m. She made it to the scene of the tragedy before 6 a.m. After poking around and talking briefly to some witnesses, Doolittle hailed a cab to the Traffic Services police station farther downtown, where Bryant was being held and would eventually make an appearance that afternoon. Doolittle remembers many of the reporters not-so-surreptitiously trying to dig up some dirt, probing for the name of the unidentified woman in Bryant’s car.

With only a small corps of reporters, the Post and the Toronto Sun were hampered during that first day, but the Star and the Globe were able to take a more aggressive approach. Parley deployed reporters to as many places as possible. They filed stories to the web while feeding extra information to Star reporter Cathal Kelly, who was writing a front-page feature.

By the afternoon, journalists were following three main threads: Bryant, Sheppard and the incident itself. Predictably, the papers with the most reporters on the story—the Star and the Globe—broke most of the new information. The Post, with fewer reporters, was unable to cover the developments as closely and adopted a more analytical approach. Navigator had shut down information from Bryant’s camp following a press conference the morning after the accident (for which the PR firm had Bryant change into a crisp new suit). Reporters then shifted their attention to Sheppard, who was turning out to be a far more complex character than anyone had anticipated.

“Not to boast at all,” says 25-year-old Star reporter Daniel Dale, “but I think I was the first to get a lot of stuff on him, and a lot of it was sheer luck. I messaged his entire Facebook list, and it was one of those days where people just get back to you. We also got stuff from his courier friends. They gave me his address.”

That address turned out to be a run-down apartment building on the east side of downtown, directly across from Seaton House, Toronto’s refuge of last resort for up to 434 destitute men. Rooming houses and smaller shelters cluster around it, while out on the street it’s a free-for-all—arguments, fights, drug deals. Few places in Toronto provide a starker contrast to Yorkville’s upscale environs.

Dale says Sheppard’s friends also volunteered that he was a heavy drinker. They thought it would make the courier seem more sympathetic, shifting blame to the police, who had come to his girlfriend’s apartment to investigate an argument, found he’d been drinking, and let him ride away. (Police said he’d been drinking but wasn’t drunk.)

The Globe’s chief librarian, Celia Donnelly, dug up a 2002 Edmonton Sun story mentioning dozens of outstanding arrest warrants for Sheppard. “Some days you get a name and it takes you nowhere,” saysGlobe reporter Kate Hammer. “Other days you get a name and it just keeps adding layers.” She  and veteran crime reporter Tim Appleby raced to verify that the Darcy Allan Sheppard in the Sun story was the same Darcy Allan Sheppard they were hunting. The crimes were minor. Sheppard was alleged to have cashed bogus cheques in small amounts to himself at local Money Marts.  The Star’s Parley says he wouldn’t have used that information even if he had it first. “The fraud charges have nothing to do with the Bryant incident. Why bring that up? To prove what?”

The Star scooped the Globe right back on September 3, with a story piecing together Bryant’s evening before the accident. This was yet another turning point. The story revealed that Bryant hadn’t been gorging on charcuterie and fine wine at a Yorkville boîte, but had been chowing on shawarma at a downscale take-out joint. His bill came to around ten bucks.

Most of the early reporting also included reconstructions cobbled together from eyewitness testimony, police reports and security camera footage. This is where the mainstream media did engage in some irresponsibly speculative reporting—none more so than CTV affiliate CFTO, which led the 6 p.m. news on September 2 with an absurdly hyperbolic reconstruction of the incident.

After playing some fuzzy security camera footage showing Bryant driving the wrong way down Bloor Street, reporter Tom Hayes wonders aloud, “Why did the luxury convertible cross over into oncoming traffic? Was the driver trying to shake the man holding on to the car, or was the driver forced?” Next, the broadcast cuts to a wide shot of Hayes in a parking lot, crouching beside a white convertible of his own. The handheld camera rushes in on Hayes as he barks, “There were reports that the cyclist was hanging on, possibly to the steering wheel.” Hayes clasps the wheel of the car. “Could this have made it difficult for the driver to turn right…instead having to go left into the opposite lane?” Here, Hayes jerks the wheel back to the left, pointing accusingly. There’s a cut back to the security footage as Hayes announces conclusively, “And that’s where the cyclist struck the mailbox and later died.”

The way the segment strings together fact, speculation and unsourced reports is awkward enough, but Hayes hardly even bothers to specify which is which. And eyewitness testimony, the source of all the reconstructions, is no silver bullet anyway. “Witnesses are unreliable,” says Carleton University journalism professor emeritus Joe Scanlon, whose research focuses on rumour dissemination after high-profile disasters. “They reconstruct what they think must have happened. Every time you talk to someone they change their story a little bit.”

The public appetite for new information was voracious. The following week also saw stories about whether or not Bryant had received a “VIP stay in custody.” There were reports on the hazards of urban cycling. TheGlobe’s Christie Blatchford wrote a column titled, “In a city of drivers and cyclists at odds, the one on the bike is always right.” In it, she argued Bryant “may well have a solid legal defence, [but] it is trickier to see how he will be able to muster a moral one.”

And increasingly, there was a focus on Navigator. Rick Salutin, a freelance columnist for the Globe, is disdainful of PR. He fears that with spin doctors involved, the reporting becomes dubious. “It’s grossly unfair that if you have a lot of money you can hire people to spin your case,” he says, “and if you don’t, you’re largely at the mercy of those who do.” But reporters and editors insist they had no contact with the PR firm and respond to any suggestion otherwise with incredulous denials—of course we didn’t talk to Navigator, they say. Maybe someone else did, but not us.

On September 3, Star reporters Kenyon Wallace and Nick Kyonka put together a detailed chronology of Bryant’s evening. Someone, described only as “close to the family,” told Parley that Bryant and Abramovitch went for shawarma, then drove over to the east end for a walk, before finally getting dessert and heading to Yorkville. Kyonka and Wallace visited every destination, quoting one waitress who refused to be named, and another who gave only her first name. Both reporters were interns at the time, and neither knew who the original source was. (Parley won’t disclose the name.)

At first glance their story is indeed suspicious—unnamed sources, an angle sympathetic to Bryant, coming just as the weight of public sympathy was beginning to shift. “I just think it was a plant from the PR firm,” says Peter Kuitenbrouwer, Toronto columnist at the Post and acting Toronto editor at the time. “It was really frustrating because they obviously had access to Bryant. And with an unnamed source, how do you even know it’s true?”

But Parley refutes any suggestion of Navigator’s involvement. “If I were [Navigator chair Jaime] Watt I’d be delighted, because he’s getting credit for a lot of shit I don’t think he had anything to do with.” Parley says the original source didn’t approach the Star. Instead, editor-in-chief Michael Cooke came into the morning news meeting on September 2 and insisted on learning everything Bryant had been up to before the accident. So they started making calls.

Plan B was to deploy reporters around Yorkville to suss out Bryant’s whereabouts, an approach that obviously would have turned up nothing. “It would’ve been a better story for us if he had been drinking,” says Parley. “No question we were going out there with a bit of gotcha in mind, and the facts were the opposite of gotcha.”

On the morning of Tuesday September 1, Bryant was the villain and Sheppard was a near-innocent victim. By Wednesday evening, they’d almost switched roles. Sheppard’s image had been tainted by stories of a troubled past, police encounters and reports of alcoholism. Bryant’s image had been burnished—mainly in contrast to Sheppard. And though no one will admit to speaking with the PR company, Watt tells a different story. “Things were being reported that were not true,” he says. “That Bryant left the scene of the accident, that he was drunk, that he would be in court on October 19 when it was just an initial appearance by his lawyer. If we’d let them go they would’ve set like cement.”

Watt is unabashed. Of course the agency spoke to journalists. And it did so because the working theory, as Cornies might call it, was leading reporters down a distinctly one-sided, anti-Bryant path. “We did what we needed to do to change the dimensions of the story, and then we stopped talking,” Watt says. “We know that journalists and newsrooms have been the subject of huge cutbacks. We have more general reporters working on more topics, and we have a huge pressure to file quickly. So there are lots of examples of pack mentality setting in.”

Navigator was most interested in what would make its client look good. But that may not have been too far from the truth. “It wouldn’t surprise me at all to learn that the story became more accurate after the PR people got involved,” says Ira Basen, producer of CBC Radio’s Spin Cycles, a documentary series about the intersection between PR and the press. “That would only surprise you if you thought they were there to be untruthful. And sometimes they are. Not outright liars, but they speak through what I call a sort of managed truth. But sometimes that’s very close to the absolute truth. Hopefully journalists don’t need PR people to turn them in the right direction,” he says. “But we all have our blind spots.”

Within 24 hours of the incident, members of the cycling community rallied hard to correct what they saw as a smear campaign against Sheppard. Donald Wiedman, a PR professional and cycling advocate who took up Sheppard’s case pro bono, says the public was “duped with Navigator’s help.” He even set up a Twitter feed called Bryant Truths, a counter to Bryant Facts, Navigator’s feed that operated alongside a blog of the same name. (Navigator rarely updated those accounts, though for the rest of the year there was predictable hyperbole about Twitter’s importance—the December 2009 issue of Toronto Life even included “political damage control by tweeting” as one of the “25 Ideas That Are Changing the World.”)

Mess Media, an activist website that claims it “corrects media reporting errors” in stories on bike couriers, set up the Bryant Watch blog, featuring posts about Watt’s own 1985 fraud conviction. The site strongly alleged that Navigator was feeding the press with anti-Sheppard information. Mess Media complained about biased coverage, but much of what the site reported was impossible to prove and patently libelous. “Bryant lost it,” reads a post from December 7. “He was overtaken by complete rage. If he had a gun he may have shot Al in the back. If he had a bat he may have bashed Al’s head in from behind.” Another blog, The Bike Joint, dubbed Bryant the “butcher of Bloor Street.”

Wiedman was also in touch with another, more mysterious activist, who goes by the moniker “honestedits” and who will only describe himself as working in the “information technology sector.” “Who else would risk speaking against the powerful than the anonymous?” he asked via e-mail. He alleges that television reports chopped up the security camera footage and presented it out of order. His “improved footage” YouTube videos, complete with annotation explaining what’s happening onscreen, were posted days after the accident and received over 40,000 hits.

Some bloggers were more scrupulous. In Torontoist’s year-end round up, Hamutal Dotan wrote, “It would behoove us all to sit down, shut up, and let that justice system do its job. None of us knows what precisely occurred that night: the specifics…ought not to be a matter on which we speculate, theorize, emote, or onto which we graft explanations as they suit our world views.”

***

The Star’s Doolittle has another theory as to why the Bryant-Sheppard coverage, after a shaky start, turned out relatively well. She thinks the story of Victoria Stafford, an eight-year-old girl kidnapped in Woodstock, Ontario, made reporters and editors a lot more cautious about jumping the gun on breaking news. “I think Tori Stafford was a teachable moment in newsrooms across this country.”

On April 8, 2009, an unidentified woman kidnapped Tori as the girl walked home from school. A security camera at a nearby high school captured her leaving with the abductor. It was the last time Tori was seen alive—three-and-a-half months later, police found her remains. The abduction created a speculative bubble that imploded in an especially embarrassing and appalling way. “We always do our best to avoid preconceptions,” says Doolittle, “and we know the dangers of them, and we sometimes get these reminders of those dangers.”

Though estranged, Tori’s parents Tara McDonald and Rodney Stafford came together to appeal to the media and the community for their daughter’s safe return. As the days and weeks went by, there was little news about Tori’s whereabouts or the perpetrators, but there was still plenty of public interest. So attention shifted to Tori’s parents—their strained relationship, McDonald’s OxyContin addiction and a trust account she set up, which some speculated was an attempt to profit from the tragedy.

“It’s seldom that you have a parent like Tara McDonald making herself so available,” says Cornies, now a columnist for the London Free Press, one of the major newspapers closest to Woodstock. “Most people in that situation are shell-shocked and grief-stricken into hiding, but Tara was inexplicably almost brave and in front of the cameras and microphones daily. And because that was such a departure from standard behaviour amongst parents of missing youth, all the alarm bells start to go off among reporters: ‘Why is this parent so different?’ And then the theories begin to spin out.”

In the Stafford case, the working theory was that Tori’s mother was losing her grip on sanity and was maybe even the kidnapper. In the information vacuum that enveloped the case, there was little to substantiate that, but little to disprove it either.

The nadir of the Stafford story came on April 21, when police released a composite sketch of the woman in the school video. Some comments on a Facebook page speculated it resembled McDonald. The flimsily sourced rumours made it into newspapers across the country, which reported the allegations and McDonald’s denials.

This whole sordid soap opera continued until police arrested Michael Thomas Rafferty and Terri-Lynne McClintic on May 20. Cornies wrote an editorial for the Free Press after the discovery of Tori’s body that concluded with the news of a funeral. “We’re not invited,” he wrote, referring to the media. “Nor should we expect to be. But…it seems right that we should encircle McDonald and Stafford now—in solidarity, in sorrow and in abject apology.”

Montreal-based media critic Craig Silverman thinks it was a textbook case of pack reporting, exacerbated by the expectation that public interest must always be met with more coverage, whether there’s anything new to report or not. “It is so intense when something happens and the details are sketchy, but you’ve got to write about it,” says the author of Regret the Error: How Media Mistakes Pollute the Press and Imperil Free Speech. “The idea that we could stay silent when the competition is out there reporting things takes an incredible amount of self-control because you want to beat your competitors.”

After all, says Kelly Grant, the goal when reporting a breaking story is, “To report the hell out of it, to talk to as many people as humanly possible, to take any suggestion that could be controversial and just report the shit out of it.”

Silverman thinks the Stafford case is an example of “confirmation bias,” as were the early stages of the Bryant-Sheppard story. “Once you’ve decided what’s happened, you’re going to write to that,” he says, “and if something is reported without due diligence, it’s going to spread because people will not have the self-control to stay away from it, and because the news cycle is so quick right now.”

In 2006, for example, the Post reported that Jews, Christians and Zoroastrians living in Iran would have to wear ID badges reminiscent of those worn by Jews in Nazi Germany. The paper ran the piece on the front page even after a rabbi at New York’s Simon Wiesenthal Center said he couldn’t confirm it, though he believed it to be true. The New York Post picked up the story, but it turned out to be a complete fabrication. “It seemed like it could be true. It fit certain perceptions of what Iran might do, so it was believed easily,” says Silverman. “But it was entirely made up. They went with it because they wanted the scoop.”

***

Reporters chased down a lot of scoops during the first week of the Bryant-Sheppard story, and in the beginning they were thrown off by a lot of glib assumptions and false innuendo. In the end, though, a more complicated and probably more truthful story emerged. Maybe it really was just diligent reporting. Maybe it was simply that reporters kept finding new information and didn’t have the time to report idle speculation.

More troubling is the suggestion that the PR spin may have inadvertently kept them honest. Even if Navigator wasn’t behind the story about Bryant’s night on the town, it’s possible that it was behind some other reported details, or that the move to shut down communication from Bryant’s camp forced reporters to look into the other side to keep feeding the public appetite. “That made a lot of people turn toward Sheppard,” says Matthew Coutts. “It didn’t take much to find out about his past and family in Edmonton, and you had Bryant’s story coming out in a far more controlled manner.”

The coverage wasn’t an unqualified success. There were those initial reports of a hit and run. There were too many irrelevant, salacious stories about Sheppard’s past, too many dubious reconstructions of the accident and there are still too many questions. There are also hints of PR meddling on the day Bryant emerged from Traffic Services. The Globe’s Tim Appleby says that an unidentified man insisted photographers turn their cameras so the police station would be out of view and Bryant would appear against a bright blue morning sky instead. (“I don’t know anything about that,” says Watt.)

But the mainstream media’s missteps pale next to the blogosphere’s. “When I started doing this, the media controlled information,” says Appleby. “People got their information from the mainstream, and that’s absolutely not the case anymore.”

He believes that civilian journalism is both good and bad. “Probably more good than bad,” he says. “But the bad thing is that everyone has an opinion. And it’s harder and harder to separate the real from the imagined.”