Game Theory

When Channel 4 covered cycling in London recently, the content was predictably disappointing, but it does at least serve as an example of why coverage of this type is so problematic.

It came in three parts, which are best covered in reverse order.

The interview

Still, it could have been worse.

The third piece was a studio interview, with Nicola Branch of Stop Killing Cyclists and Steve McNamara of the Licensed Taxi Drivers’ Association interviewed by Jon Snow. It seems barely conceivable in this country that any current affairs interview, on any subject, could be anything other than binary and polarised, and this was no different, with Snow (himself not only a regular cyclist but president of the CTC) first bringing up McNamara’s notorious—but almost universally misunderstood—”Isis of London” comment and then asking Branch “You don’t shoot red lights, then?” with a slight undertone that belied his previously-voiced opinion: “cyclists behave extremely badly and I don’t know if there are any cyclists who haven’t gone through a red light.” Still, it could have been worse, and the interview hit the schedule buffers before anything incendiary—or, for that matter, useful and enlightening—could be said.

But hey, there we go: People in cars, and people on bikes. A polarised view, and the majority of the audience will have more in common with the former.

The VT

At this point, I noticed I was halfway through eating my own shoe.

The piece that preceded it was somewhat worse. It didn’t really help that Channel 4 had tossed the assignment onto the sports desk rather than finding their transport reporter, so we were treated to Keme Nzerem—who was sufficiently wet behind the ears to seem surprised that helmets were a contentious issue—kicking off the piece with some well-worn YouTube clips of spectacular action, voiced over with a delivery indistinguishable from caricatured sensationalist Ted Maul. “Shocked? You shouldn’t be! Just another day in the saddle… or ON A CAR BONNET!”—I fully expected a caption to flash up to tell us the report was coming from Cowsick.

And who better to help shore up this sensationalism than “self-styled vigilante” Dave Sherry? It seems, in this case, that “style” involves dressing up as a mobile CCTV van, sticking your chin as far forward as possible, and providing soundbites that start out cringeworthy and end with “No retreat, no surrender.” (At this point, I noticed I was halfway through eating my own shoe with the foot still inside it, and had to rewind a bit while I sicked it back up again.)

But hey, there we go: People who’ve not been involved in spectacularly acrobatic collisions, arguments or physical assaults; and people like Dave Sherry. A polarised view, and the majority of the audience (and this time I mean absolutely everyone apart from Traffic Droid) will have more in common with the former.

But, for all the tediously familiar binary world view, the faux tribalism, the sensationalism, and the suggestion that riding a bike has a noteworthy connection to reinventing oneself as a one-man surveillance platform, the most interesting piece was the online teaser article that preceded the actual broadcast.

The teaser

Retrospective blame may be unhelpful, but it’s worth seeing where responsibility lies.

“The tragic consequences of one cyclist running a red light,” read the plaintive headline to Nzerem’s article, neatly obscuring the myriad other contributory factors in Brian Dorling’s death.

As Nzerem himself writes of Brian’s widow, Debbie:

Debbie doesn’t blame anyone for what happened. She says she could blame lots of people. She could blame Brian. Or the driver. Or the truck designers. Or TFL for creating an unprotected cycle lane which the coroner described as a “comfort blanket”. But blame, she says, gets her nowhere.

That’s understandably pragmatic and philosophical, of course. However, if certain events are to be avoided, then when they occur it is quite proper to take a look at the factors in them. Retrospective blame may, in many ways, be unhelpful. But it’s worth seeing where responsibility lies.

Not least because Channel 4 are, as with most media coverage, focusing on wholly the wrong factor.

The end game

This is like modelling the nuclear standoff between the Soviet Union and Luxembourg.

This particular tragic collision came about as a result of two actions: That of Brian Dorling violating a red light, and that of David Cox violating a red light.

On the face of it, these are the same decision or mistake: Two men, each jumping a light. But whilst in some ways these actions are the same, they are not equal, just as if two men discharge a weapon in a crowded street the error might be the same, but if one is a shotgun and the other a spud gun they are not equal.

This is reflected in licensing, of course: One needs a licence for a shotgun but not a spud gun. One needs a licence to drive an HGV, a lesser one to drive a car, and no licence to ride a bike.

Pertinently, it’s also reflected in law: Firearms offences are rather more serious than spud gun offences. Similarly, jump a light in a motor vehicle and the punishment will affect your licence to use it; do so on a bicycle and the punishment doesn’t. Move further still down the physics pyramid and exchange your bicycle for a simple pair of shoes, and jumping red lights—at a pelican crossing—is no longer an offence at all.

Yet Channel 4 have focused on the decision at the bottom of that pyramid of potential for harm. Let’s check out the game theory here.

The game is basic stuff. Two parties, two decisions. If the guy with the bike obeys the lights and so does the HGV, no-one’s harmed; if he doesn’t, the guy with the bike might die. If the guy with the bike doesn’t obey the lights, and the guy with the HGV does, the guy with the bike might die; if they both jump the lights, the guy with the bike dies.

Sure, this is way over-simplified. You only have to look at cases such as those of Mary Bowers or Rosie Wright to see that even if no-one jumps any lights, the result still only goes one way (indeed, that jumping the lights can avoid a tragic result). If this is game theory as applied to the cold war, it’s like modelling the nuclear standoff between the Soviet Union and Luxembourg. And that’s the point: any violation of the light by one party, whether intentional or not, never presents any danger at all to the other; whereas a violation by that latter pretty much always presents danger to the former.

So, given those two actions, Channel 4 have devoted their article entirely to the least important.

But, in case it’s not already clear: Those aren’t the only two actions.

Pre-game theory

Bow is effectively a real-world experiment for game theory in which deaths are appalling outcomes.

The game begins long before two men arrive at a junction. Rather, that game arises purely as a result of previous decisions.

One of these is the decision to build a vehicle with, by virtue of sheer size, significantly greater potential for harm than most, but with reduced ability for the driver to see (or even hear) around that vehicle. Another is the decision to bring that vehicle into densely-populated urban roads at peak times.

Each of these are decisions which affect the end game; decisions which increase the likelihood of negative outcomes.

But, further still up the chain, is the one decision that has the power to virtually eliminate negative outcomes: How to engineer the road layout and the signage that guides users across it.

Bow roundabout is notorious. It’s effectively a real-world experiment for game theory in which the deaths of Brian Dorling, Venera Minakhmetova and Svitlana Tereschenko were all appalling outcomes. People on bicycles and people in HGVs vie for the same patch of tarmac, guided but ultimately unrestrained by some lights and some bits of paint. Both lights and paint are known to be abused constantly, and the result of simultaneous abuse is crystal clear, yet this was designed.

The smallest game, the smallest move

This focus is all wrong. All wrong.

Channel 4 are not the first to focus on totally the wrong thing. It’s common. It’s easy to say that someone jumped a light and went under a truck and call it “the tragic consequences of one cyclist running a red light.”

It’s easy because it fits a simple, digestible, binary format: people in cars and people on bikes; normal people and people like Dave Sherry. Line up your interviews, dig up your YouTube clips; it doesn’t take much to “get an angle” on this.

And it’s easy because it challenges no-one. Drivers keep driving, highway engineers keep engineering the same highways, and everyone keeps noticing people jumping lights when they happen to be on two wheels but no more. The world turns, people die, and no-one has to make any difficult decisions.

Yet this focus is all on the people least able to make a change. The people who, even if every last one of them obeyed every last light, would still be killed by others’ negligence, others’ poor decisions, and others’ simple and inescapable nature of human error, simply because of basic physics. Simply because someone brings the Soviet Union’s arsenal to Luxembourg’s game.

“Sometimes the cyclist gets it wrong too—with catastrophic consequences,” writes Nzerem, as if those catastrophic consequences somehow don’t exist if “the cyclist” doesn’t get it wrong.

This focus is all wrong. All wrong.

For an article to list most of the factors in someone’s death—including the single factor which could have virtually guaranteed that their death could not have occurred—and to choose to focus solely on the victim’s human error is shameful. It’s particularly shameful because it’s a threadbare line of reporting, repeatedly churned out and repeatedly challenged, and because it stands in the way of implementing changes which will prevent these deaths.

As long as we keep blaming people for their own deaths, others will shrug and make a crass comment to let everyone know that they’ve heard of Darwinism but not really understood it, and we will continue to have a steady stream of dead people to blame.

This was not the tragic consequence of one cyclist running a red light. This was the tragic consequence of complete failure to understand well-documented human behaviour. This was, simply, the tragic consequence of terrible engineering.

There should be no need for game theory. Engineering’s role here is to eliminate it. And, if human lives are to be valued more than page hits or Twitter mentions, journalism’s role here should be to make that clear.