Catapulting Boulders

6 September 2015

In writing his guest post about the Notice of Intended Prosecution, Ralph shared with me some material he received from Surrey Police. It makes for interesting reading, and this post highlights one particular point.

I should be absolutely clear in advance that this post is not a criticism of Surrey Police (though the points in Ralph’s post about antiquated IT policies and processes, and the notable remark that video is not considered independent evidence, remain valid). Their handling appears to be pragmatic and reasonable: In assessing any case in terms of prosecution, they must inevitably consider whether the CPS are likely to proceed with it. Their view in this instance was that they would not; and with reluctance and weary cynicism, even without considering the harsh realities of ever-dwindling resources, it’s hard to disagree.

The point, though, is that this situation—finding it hard to disagree with a pragmatic lack of prosecution—is incredibly frustrating. The statutory definitions of the relevant offences are already flawed, the application of those offences is being eroded, and the issues highlighted in this post add to that erosion. So let’s take a look at the video and see where these issues start to erode the likelihood of prosecution.

The video

The incident that Ralph reported occurs at 1m13s. As Ralph begins to pass a parked van, a Corsa is seen to pass extremely closely. It leaves very little room for error, is close behind another vehicle, and is passing on a hatched area approaching a traffic island which forms a pinch point. This traffic island prevents the driver from being able to overtake (it is possible to speculate that she has failed to see it because she is fairly closely following another vehicle, or it could be that she has simply misjudged the cyclist’s speed), but after having to drop back the driver sounds the horn repeatedly and then the Corsa is seen again, clearly being driven deliberately towards Ralph and slowing to stay alongside him. The driver then finally overtakes, again while approaching a traffic island, and again leaving little space.

A close pass alongside a parked van.

In the screenshot above you can clearly see the Corsa’s wing mirror in the top right of the frame. It appears to be close. Incredibly close. In the appendix below I’ve tried to show just how close.

So we have an extremely close and dangerous pass, clear misjudgment of hazards, deliberate aggression and misuse of the horn, repeated misjudgement, and another close pass. Certainly several viloations of the (non-legally binding) Highway Code, and many would argue also at least careless driving. Some would argue it constituted dangerous driving, and it is feasible that the deliberately aggressive driving could constitute assault. (The pragmatic questions are around how likely a prosecution is for each offence, and whether a prosecution for each is in the public interest.)

But enough of the evidence. Let’s take a look at three comments on the video which form a large part of the reasoning for not recommending the case to the CPS.

Close passes are bad, m’kay?


The video shows you following a blue Fiat vehicle before you then go to its offside very close and overtake it whilst that vehicle is still moving forward.

In other words: You (a person on a 10kg vehicle with no protective structure) closely passed another driver (a person encased in a 1400kg vehicle with a steel safety structure) while they were moving. This is Something That Should Not Be Done, hence it is unlikely that we could prosecute the accused (a person in a 1600kg vehicle with a steel safety structure) for closely passing you (a person on a 10kg vehicle with no protective structure) while you were moving.

Using the hatched area is bad, m’kay?


You then travel along the road centre in a hatched area bordered by broken lines overtaking a number of vehicles. The Highway Code advises that road users should not enter such areas unless it is necessary and you can see that it is safe to do so.

In other words: You entered the hatched area to pass other vehicles. This is Something That Should Not Be Done, hence it is unlikely that we could prosecute the accused, who entered the hatched area to pass you.

Leaving no room for error is bad, m’kay?


Later you are following a Peugeot vehicle very close to it. As such you do not appear to have a clear view of the road ahead of it. Should the vehicle have braked suddenly there was a very real chance that you would have collided with it.

In other words: You (a person on a 10kg vehicle with no protective structure) were sufficiently close to another car (a 1200kg vehicle with a steel safety structure) that had you needed to take evasive action you may have struck it. This is Something That Should Not Be Done, hence it is unlikely that we could prosecute the accused (a person in a 1600kg vehicle with a steel safety structure) for being sufficiently close to you (a person on a 10kg vehicle with no protective structure) that had you needed to take evasive action they may have struck you.

Two wrongs do make a right

As you may have noticed from my ever-so-subtle phrasing, there’s a pattern here: If at any point in time you can be shown to have done anything which is perceived as Something That Should Not Be Done, then it’s not possible to prosecute someone else for doing it, even if they do lots of those things at the same time. (The Corsa driver did all three of these things plus more: notice the traffic island shortly after the parked van, for instance. And let’s not even start on what appears to be a deliberate swerve towards the rider.)

But, absolutely crucially (and as I have again ever-so-subtly suggested) there’s also a key differentiator.

Stones and boulders

Take the example of being close to the Peugeot. If we accept the suggestion that the rider is close enough for there to be an appreciable risk of hitting it if the driver has to brake (which I think is fair, although I would note that the lateral positioning means that it’s likely that the rider would be able to swerve to its side while braking) then we must look at the consequences of doing so.

I confess that once, in my teenage years, I unintentionally tested this: when the driver in front of me suddenly braked to allow a pedestrian to cross the road I was too close to avoid an impact. It was at reasonable speed; enough that I ended up half on the roof of the car. We were all a little surprised, but there was no damage to me or the car. Lesson learned. Such is youth.

Here’s another example of someone doing Something That Should Not Be Done.

Again, no damage to the rider or the car.

Obviously, not all such incidents are completely devoid of consequence. But the key point is this: it is all but impossible for such an incident to cause harm to anyone inside the car. (Granted, an older convertible car with no reinforcement behind the occupants’ heads may be an exception.) It may damage the car, and it may damage the person whose fault the collision was, but that is all.

The message that needs to be heard is this: There is nothing morally or logically unsound about the notion that it should be broadly acceptable for someone on foot or on a bicycle to pass close to a heavy and incredibly strong steel box whilst it also being broadly unacceptable for someone in a heavy and incredibly strong steel box to pass close to someone on foot or on a bicycle.

This is not about people in glass houses throwing stones. That idiom makes sense only if it is clearly understood that the real problem is that the people outside the glass houses are catapulting boulders.

The potential for harm, which is a matter of basic physics, is crucial. It’s why we license firearms but not water pistols. Yet it seems that if you want someone to be prosecuted for walking down the street with a sawn-off shotgun, you’d better not walk down the street with a water pistol.

One thing seems clear: To seek the prosecution of others, one must be demonstrably flawless oneself.

But is that really true?

Demonstrably flawless

Take the example of Liz Brown. Killed when struck from behind, reports indicate that it was explicitly stated in court that she was at no fault whatsoever. Riding legally and responsibly along the road, she was hit by a van, the driver of the was found to have committed no offence.

Take the example of David Irving. Killed when struck from behind, his actions were scrutinised in court but there seems to be no implication of blame (indeed it is hard to imagine what could possibly imply it). Riding legally and responsibly along the road, he was hit by a van, the driver of which was found to have committed no offence.

And so on. There are numerous cases (not least those involving low sun) where two parties collided, the victim was deemed wholly correct and blameless, and yet the other party was either found to be also blameless or not even accused of an offence.

Social norms

Blamelessness aside, it’s tempting to wonder whether analogous situations, where the potential for harm is unilateral, would be handled similarly. Would a wife who shouted abuse at her husband’s friend and was later threatened by her husband with a kitchen knife be prevented from seeking his prosecution? Would an employee who didn’t wear a dust mask as often as he was required to be prevented from seeking the prosecution of an employer whose health and safety process failings were so severe as to have the potential to cause its employees’ deaths?

It’s hard to imagine these being treated similarly, because people see the disparities in each. Yet they both offer analogies in the context of the application of law: in one, two parties commit assault with two markedly different weapons (a voice and a knife); in the other, two parties commit health and safety violations where the injured party will likely be the same in each case.

And this is where the social norms of personal transport must be: we must recognise the potential for harm of heterogeneous vehicles, and we must recognise where harm is unilateral regardless of fault. We should be far more concerned with people’s ability to cause harm to others than with their ability to cause harm to themselves; and we should­ be more concerned with either of these than with anyone’s strict adherence to the letter of the Highway Code, which has no legal status.

So, while it’s hugely frustrating to see remarks like these from the police, those remarks simply reflect the depressing reality of the court.

And it’s hard to see how that can change without changing magistrates’ and jurors’ social norms.


To measure the gap between the Corsa and the van, we can use the Peugeot as a reference. In the frame below, I’ve traced the convergent lines of the van in pink. The lowest of these, passing the edge of the rear tyre, approximates the point on the tarmac directly below the right edge of the van. (In reality the body, which will protrude beyond the tyre, will be a little to the right of this. Thus I am erring on the side of overestimating the gap.)

The next step is to draw the two blue lines: one to measure the body width of the Peugeot and one to measure the distance from the rightmost point of the van to the edge of the dashed white line.


Why am I measuring to the white line? Because this seems to represent the leftmost position of the Corsa during the period where it is passing the van. It’s possible to see the front nearside tyre edging to the left of this just as the front of the van is reached, at 1m15s, although prior to the point of the Corsa’s contact with the road is not visible. However, since both the body and the wing mirror extend to the left of the tyre, I think it can be stated with reasonable confidence that at no point is the wing mirror significantly to the right of the point I’m measuring to. I suspect it remains to the left of it at all times, but I prefer to err on the side of overestimating the gap.

Measuring the lengths of these lines as drawn in Inkscape gives values of 220.3 and 181.9 respectively. The body width of a Peugeot 1007 is 1.61m, meaning that the gap between the van and the white line is approximately 1.33m.

So, how wide is a cyclist? Ireland’s National Transport Authority gives a width of 0.75m, and the UK’s Department for Transport uses a dynamic envelope of 1.0m (being the static width plus the normal lateral moving of pedalling, and you can see the pendular motion of the bike on the video).

So that’s between 0.33m and 0.58m to spare, depending on how much the rider is moving. But that gap is shared across both sides of the rider.

The line along which Ralph’s wheels are travelling is clearly to the right of the middle of this gap. (Ralph’s camera is mounted centrally on the bike, in front of the stem.) So, again erring on the generous side, if we assume his wheels are in the centre of the gap, that leaves between 0.16m and 0.29m between Ralph and the Corsa. Were we to estimate Ralf’s position as two-thirds of the gap, that lower figure would become negative.

Given that in those figures I’ve overestimated the distance to the van body and I’ve placed Ralph further to the left than he is, I’ll let you make your own estimate.

Personally I have no doubt whatsoever that this was a dangerously close pass, about as close as they come, even without considering the proximity of the van or the approach to the traffic island, both of which add considerably to the risks and both of which should have been entirely obvious to the driver. For this to be compounded by the subsequent aggressive use of the horn, a deliberate swerve at the rider and then another close pass approaching a traffic island, and to still not warrant any action, is somewhat dispiriting.

Is it any wonder that some people turn to social media (or worse) instead of the law?


  1. Notak 7 September 2015 10:55am #

    “The driver was found to have committed no offence”. Indeed, they had no committed no offence. It’s time we redefined some traffic offences in terms of outcomes not actions.

    • Bez 7 September 2015 11:37am #

      Well, “indeed they had” by virtue of the verdict, but not everyone would agree. The issue is really the huge subjectivity in the process.

      Outcomes are a funny old thing. Consider someone who pulls out of an urban side road at night into the path of an unlit cyclist. They see at the cyclist at the last minute and stop; the cyclist brakes but fails to avoid them, is dismounted at low speed, but sustains a statistically unlikely injury which proves fatal.

      Should this be charged and sentenced similarly to, say, someone who drinks three pints of beer before racing another driver home from the pub at well above the speed limit and careers fatally into the rear of a cyclist? And how should it be charged in relation to someone who does the same thing but only causes injury, or indeed does not even have a collision? The actions that created the danger were no different and the likelihood of a cyclist having been present and similarly unavoidable was no different; the outcome is merely a product of random events that were foreseeable by the first party.

      I don’t have a problem with the primary consideration being actions (mens rea or otherwise) rather than outcomes—I firmly believe that’s the preferable approach—but I do think there is a reasonable argument for the creation of statute phrased in terms of something like “presenting clear risk of significant likelihood and/or severity of harm to others, including the consideration of any reasonably foreseeable errors of any pertinent party”.

  2. Notak 8 September 2015 8:34am #

    Sure, there’s a large element of chance in outcomes. That’s (one of the reasons) why we get away with so much risky behaviour! In line with which, I’d say not even actions but attitudes and intentions. The trouble with “presenting clear risk… ” is it still leaves a lot open to the court’s judgment, so in practice it’s not far off “was he shitter than average?” But thenattitudes and intentions hit exactly the same problem, in addition to being virtually impossible to determine.

    As to the two cases in your example, I don’t think the first one necessarily needs charging or sentencing at all. The driver just needs to be held responsible in some way for the death he’s caused, perhaps by being liable to pay compensation. Or something like that. Not necessarily points on a licence.

    Of course this wouldn’t get Ralph anywhere, because there was no “outcome” in that sense – no collision or physical injury – just some aggression and shouting. In fact, despite the driving being both initally crap and then aggressive, it seems to me more like an assault case (threatening behaviour?) for which a traffic incident was the trigger.

    • donk 17 September 2015 1:34pm #

      “it seems to me more like an assault case (threatening behaviour?) for which a traffic incident was the trigger.”
      Yes, also, I’m pretty sure if I jumped off my bike and used the bike for beating someone around the head or even just swinging it around madly, close to them then I’d be charged with some sort of assault rather than a traffic offense. Drivers do seem to be able get away with using their car as a weapon and still benefiting from the leeway given to such a dangerous* form of transport.

      Have *any* intentional collisions been charged as assault rather that dangerous driving?

      *which is accepted and mitigated for.

      • Bez 17 September 2015 1:45pm #

        Yes. There have been GBH and murder convictions where the assault has been carried out with a motor vehicle.

      • donk 17 September 2015 1:51pm #

        Ah right. Did a bit of googling but most DD + assault cases seemed to be from shit driving and then an additional physical assault after getting out of the vehicle.

      • Nick 29 September 2015 12:32pm #

        It most often seems to count as assault when the victim is a police officer. Presumably they feel more confident about the quality of the evidence in those cases.

  3. Eric D 8 September 2015 3:01pm #

    That’s the trouble with HC 163
    “give at least as much room as you would when overtaking a car”
    Drivers tend to interpret ‘room’ as ‘separation’ or ‘clearance’, not ‘distance from kerb’.

    And if we squeeze past (filter) inches from them,
    they might think they are OK to squeeze past (filter) inches from us !
    Maybe we should carry car doors on our bikes ?)

    I would be interested to know why they legalised ‘Reckless Driving’:
    the definition was (IIRC)
    ‘knowing that their action would create a risk of endangering others,
    they went ahead to take that action, regardless of the consequences”.
    I guess we could argue that ‘All driving is reckless’, by that recognition !
    Hence ‘(far) below the standard ordinarily expected from a competent driver’ …
    Hmmm … I think Hansard is online, and should record the Parliamentary debates verbatim.

    This post reminds me of the biblical definitions of hypocrisy:
    “Why do you look at the speck in your brother’s eye and
    pay no attention to the log in your own eye?”

    “Blind guides! You strain a fly out of your drink, but swallow a camel!”

    • Bez 8 September 2015 4:21pm #

      “I would be interested to know why they legalised ‘Reckless Driving’:
      the definition was (IIRC)…”

      There was no statutory definition, it was left to case law. This may have been one of the motives for changing it (though 11 years of case law must have presumably stabilised it somewhat, and the definition that replaced it is—as often noted—hardly free from subjectivity).

  4. donk 17 September 2015 1:47pm #

    “And if we squeeze past (filter) inches from them,
    they might think they are OK to squeeze past (filter) inches from us !”
    If I’m stationary or slowly slogging up a hill I don’t mind cars slowly squeezing past but individual and combined speeds should be taken into account, this is what a lot of people seem to forget. Cyclists filtering past stationary/slow traffic while doing 10mph (combined 15mph) vs cars overtaking a quick moving cyclist (15mph?) while doing 30 (combined 45 – and up!) – there’s a big difference in the actual maneuvers before you even start to consider the disparity of safety/vulnerability of parties involved.

    The higher the combined speeds the more clearance needed.

  5. Ron Reid 30 September 2015 7:34pm #

    Comparing a close pass by a cyclist to one by a motorist is fatuous. Rules are not made for their own sake but for safety so it’s a question of risk. The cyclist never placed the drivers safety at risk unless you start inventing improbable circumstances. The same can’t be said for the drivers overtaking. In most cases Hanlon’s Razor applies with close passes, but the close pass by the driver looked wilfully bad.

    How do you measure punishment for malice? The police’s response was a copout, pardon the pun.

Leave a Reply

Your email address will not be published. Required fields are marked *