The Startle Effect – SOBS, meet No Surprise!

I’ve spent the last few Fridays looking at an unusual court case where a judge withdrew the case from the jury, after rejecting the prosecution’s case that a rider involved in a fatal collision with a pedestrian had to time react when the pedestrian stopped unexpectedly.

The incident reminded me of the words of Chesley Sullenberger – the Miracle on the Hudson pilot – “the startle factor is real, and it’s huge”. He was talking to a US Congressional hearing into the two aviation accidents involving Boeing’s 737 Max, and refuting claims that an alert and properly trained pilot could have dealt with the issues the plane was throwing at them.

Sound familiar?

Think Bike? Think again!

Now, take a look at the photo. You may remember it as one of the long-running series of ‘Think Bike’ products, aimed at the driver.

The idea is, given the target, to try to make a driver aware of just how hard a bike can be to spot.

As soon as I saw it, my thought was that the message should be ‘Biker, Think’. And that’s because it’s a perfect illustration of the point that I regularly make when discussing the Science Of Being Seen (#SOBS); the effect of any CONSPICUITY AID – in this case, the bike’s headlight – depends entirely on WHAT’S BEHIND the rider.

It’s not the lightness of clothing, or – as in this case – the brightness of the headlight, it’s the CONTRAST against the BACKGROUND.

And that’s the message that is so difficult to get over to riders, despite my best efforts and the inclusion of SOBS as a module of the Biker Down courses that have been run by so many fire services in the years since this campaign. When the photo reappeared the other day, the riders’ responses were predictable:

:: drivers don’t look hard enough for bikes

:: the rider should be wearing hi-vis clothing

They both miss the point.


Survival Skills has adapted! Have you?
Ride2Work & Ride4Fun courses!

Online video coaching AVAILABLE!

Drivers fail to spot motorcycles for well-documented reasons – in this case, it’s the camouflage effect of the bike lights against the wintery background behind the rider.

And the belief that conspicuity aids stop ‘looked but failed to see’ incidents is mistaken. There’s little compelling evidence from crash statistics – junction collisions are as common as they ever were. It’s easy to check.

So we end up with a double whammy…

a driver who either never sees the bike before pulling out – or possibly spots it halfway through pulling out and SURPRISED! stops dead, blocking the rider’s path…

and a rider who expects to be seen (“because I had my lights on and drivers are more likely to see bikes with lights”) and caught by SURPRISE! only reacts at the very last second when he / she realises the car’s pulling into the bike’s path.

Only by understanding BOTH the Science Of Being Seen AND the No Surprise? No Accident! approach to riding do we get a full understanding of the issues thrown up by this simple photograph of a bike blending into the background.

And we’ll only begin to reduce junction collisions by understanding BOTH sides of the collision – why the driver makes the error that puts the biker at risk, and why the biker fails to predict a highly predictable error. This is what’s known as ‘INSIGHT’ – seeking to offer understanding of the relationship between a specific cause and effect within a particular context.

It’s a type of learning that revolves around problem-solving through understanding the relationships between our own abilities (self-awareness) and the ‘system’ in which we’re operating. Insight is the basis for all my training, incidentally.

If we focus on simplistic and reductionist explanations alone, we may know WHAT went wrong, but without looking for embracing, holistic explanations we’ll never know WHY it went wrong.

For more information on both projects:

The Startle Effect – expect the unexpected

Holism is the belief that the parts of something are intimately interconnected and that a full understanding is only achievable in the context of the entire system. Ecology is a classic example, being a science which is based on the principle that the systems in which organisms exist are created by the interaction of those organisms with each other and their environment.

The opposite of the holistic approach is reductionism, aiming at achieving understanding by breaking things down into their smallest parts. It’s an approach that has been followed by experimental psychology which often assumes that human behaviour can be studied effectively in relatively simple experiments, where complex behaviour is reduced to isolated variables.

The examination of motorcycle crashes is often reductionist. For example, we look at riding errors and seek to pigeonhole them – I do it myself when I explain that the three main crash types are at junctions, in bends and when overtaking.

And when we study a single crash, we try to determine – via a forensic analysis – the individual steps that were taken, and specifically identify the failures believed to have led to the crash. “The rider should / shouldn’t have…”, “the driver should / shouldn’t have…”.

The trouble is that whilst we can often identify an error in the specific place things went wrong (“the rider entered the corner too fast, and ended up offline and on the wrong side of the centre line”, “the driver failed to see the motorcycle and emerged into its path”), finding the error itself doesn’t tell us just WHY the rider or the driver made the mistake JUST THERE and nowhere else.

As I often point out, the rider MUST have made it round plenty of other corners to crash where he or she did, and the driver MUST have emerged safely from many other junctions when there were bikes around.

The reality is that seemingly unique events are entwined within a much more complex matrix – faulty decisions that rider or driver took in the moments before a crash were not taken in isolation, but influenced by training, aptitude, attitude, the circumstances as they appeared to be and the assumptions made about what would happen, as well as being based on prior experience. “It worked last time.”

Personally, I’ve always thought we need a much more holistic approach to rider safety and that the mistakes and errors riders and drivers make are greater then merely the sum of parts.

Over the last two Fridays I’ve looked at what seemed to me a very unusual outcome of a police prosecution. The case was withdrew from the jury by the judge when he allowed that a rider involved in a fatal collision would have been unable to react in time when a pedestrian crossing the road spotted the bike and stopped unexpectedly.

The prosecution applied a reductionist approach, pinpointing the single error that led to the fatal collision – the rider should have reacted faster when the pedestrian stopped moving – whilst making the assumption that an ‘alert’ rider would have been able to react in such a short time.

But the judge took a more holistic approach and allowed that the rider had seen and reacted to the pedestrian crossing the road ahead of him, and that the rider’s prior experience would have been such that he wouldn’t have expected the pedestrian to stop in front of him, and that the steps taken to that point would have prevented a collision. Only when the pedestrian suddenly stopped unexpectedly did everything go wrong. Thanks to the ‘Startle Effect’ delaying an immediate reaction, it would have taken the rider more distance to stop than he had available.

This is not to say we should ignore rider error. We can certainly work to improve skills.

But it’s only by understanding the context in which a collision like this happened that we can train riders to be better prepared for similar events. Hindsight often makes a sequence of events quite obvious. The trick is to learn from incidents like this to give us the foresight to be prepared when the same situation happens to us.

Expecting the unexpected we’re far more likely to overcome the Startle Effect. Then we’ve a better chance of dealing with an emergency as it develops.

The Startle Effect part 2 “it’s real”

On Friday last week I covered at some length what seemed to me a very unusual outcome of a police prosecution of a rider after a judge withdrew the case from the jury. I was reminded of the words of Chesley Sullenberger – the Miracle on the Hudson pilot – “the startle factor is real, and it’s huge”.

Very briefly, pensioner Ghulam Rasul, aged 74, was badly injured when he was hit by a motorcycle ridden by David Hurst, 52. The collision happened when Mr Rasul was crossing the road in Leigh, in the Wigan area, on September 10, 2019. Mr Rasul did not gain consciousness after suffering severe head injuries and died two days later in hospital. Mr Hurst spent three days in hospital after suffering neck injuries and a broken nose.

Mr Hurst was charged with causing death through careless driving, he pleaded not guilty, and the case finally came to court last week.

The prosecution alleged that the rider had only reduced his speed ‘at the point of no return’. Without a full transcript and only the paper’s report to go on, the key point seems to be that the prosecution set out to prove that

“Mr Hurst put himself in a position where the collision became unavoidable.”

To understand that, you have to put yourself into the shoes of a prudent motorcyclist and argue that such a rider would have done differently.

But – and here I have to confess my utter astonishment – having heard evidence for the defence, Judge Paul Lawton withdrew the case from the jury and gave a not guilty verdict.

His reasoning – as told to the jury – was that he had concluded it was “impossible” for them to put themselves in the defendant’s position with the evidence put before them. He cited the ‘dynamics of the accident’.

As I also mentioned on Friday, unlike most collisions where much of the evidence is based on eye witness statements or forensic evidence such as skid marks, in this case the entire incident appears to have been visible on CCTV.

And this means the rider’s responses to the pedestrian’s movements could be closely followed.

Expert witnesses said that nothing he’d done was inappropriate. The road is a 30 limit and CCTV showed that Mr Hurst saw the pedestrian and braked almost forty metres away – way beyond ’emergency stopping distance’ at 30 mph.

The prosecution’s case hinged on the rider’s inability to respond IMMEDIATELY when the pedestrian unexpectedly hesitated in his path.

Judge Lawton said “At 14.6 metres Mr Rasul froze which was an understandable response but entirely unpredictable.”

The judge continued by saying that Mr Hurst could have stopped with a 0.5 second reaction time – and that suggests to me that the rider had slowed to somewhere around 20 mph.

The crucial factor is that’s ‘best case scenario’ reaction time. It’s not just about being ‘alert’ or ‘careful’ or ‘prudent’.

We actually have to KNOW exactly what is going to happen next in order to hit the brakes to pull of that kind of reaction time.

As we do when we’ve reached the part of the car or bike test where are told that we have to perform an emergency stop.

I don’t have the full court transcript but the implication is that the prosecution’s case was that the ‘prudent rider’ should have known he would HAVE TO STOP and therefore SHOULD have reacted that fast.

Fortunately for the rider, real life intruded. The judge told the court that a normal reaction time to an unexpected event was 0.7 to 2 seconds, and to my astonishment said:

“The court is to award him the two second reaction time.”

OK, so that’s what happened in the court.

So where does this two second reaction time come from?

The answer is what’s been called ‘The Startle Effect’.

The conventional approach to motorcycle training has always been to train riders in emergency manoeuvres. Anyone who’s taken bike training in the last 30+ years knows how to perform and emergency stop, and the swerve exercise has been part of the motorcycle test here in the UK for almost a decade now.

But in an emergency, things rarely go to plan.

We’ve known for a long time that the direct cause of many motorcycle crashes is an inappropriate response to an emergency. Typically we over-react (for example, by grabbing a big handful of brake – which explains why ABS is now mandatory) or target fixate (by looking at the hazard, not the way out of trouble).

Keith Code identified these inappropriate responses years ago, and called them ‘Survival Reactions’.

Code also identified a third Survival Reaction – in a significant number of crashes, the rider simply freezes and completely fails to make any meaningful response – including braking when something or someone pulls out or steps out into the rider’s path.

I’ll call this the Startle Effect for reasons which will become obvious.

So if Code identified the ISSUES US researcher James Ouellet got his tape measure out to put some numbers on car / bike collisions at junctions. And he found that if things went wrong when the rider was three seconds or more from the collision point, there was almost always a good outcome – the rider was able to take effective evasive action.

If things went wrong with less than three seconds to the collision point, many riders who COULD have braked to a halt failed to do so.

Something derailed that rider’s response mid-emergency.

So the question we have to ask is “Why don’t we perform in the way we’ve been trained? Why do Survival Reactions kick-in and what triggers the Survival Reactions themselves? “

No Surprise? No Accident believes it’s SURPRISE! We’re surprised by unexpected events, when the situation begins to develop in a way we hadn’t anticipated.

As with so much safety research, the airline industry is well ahead of road safety.

Here’s something snipped from a study from Griffiths University in Australia:

“When a sudden upset occurs – such as icing or powerful air currents from a storm – even the best pilots can experience a “startle effect” and may struggle to recall manual flying skills for that rare situation… a person’s ability to process information is significantly impaired for 30 seconds after being startled…”

As Duncan McKillop commented on the original post:

“You will remember that Chesley Sullenberger ditched his Airbus in the Hudson because of the delay in recognising the unfolding situation for what it was and then doing something (skilfully) about it. In the simulator a primed pilot could have made Teterboro airport, but only with zero delay between event and reaction.”

It’s that ’emergency stop scenario’ again.

When pilots were told where they were (height, vector, speed, and directions and distances to the various airfields) and THEN the engines were cut, it was JUST possible to glide the plane to, and set down safely at Teterboro.

Sullenberger and Jeffrey Skiles didn’t have that advantage. They had to deal with the emergency developing in real time, and find an entirely novel solution. The entire flight lasted just six minutes.

Sullenberger was also called as a witness to a US Congressional hearing into the two aviation accidents involving Boeing’s 737 Max.

Not surprisingly, with the impact of the two crashes on Boeing, there was an effort to shift at least some of the blame for the crashes onto the crew, with one congressman saying that “facts in the preliminary report reveal pilot error as a factor”.

I wrote about this at the time. Sullenberger strongly disagreed and called for pilot training on the model saying that a review of the modifications to the aircraft on a computer was insufficient. He said:

“I can tell you first hand that the startle-factor is real, and it’s huge. And it absolutely and it quickly interferes with one’s ability to quickly analyse the crisis and take effective action.”

“Pilots must develop the muscle memory to be able to quickly and effectively respond to a sudden emergency. Reading about it on an iPad is not even close to sufficient; pilots must experience it physically, firsthand.”

Now, that’s a highly-trained pilot. We know full well that road users are nowhere near as well-trained.

And read Sullenberger’s thinking about the need for FIRSTHAND training – he’s talking about simulator time.

And put that in context with bike training in emergency braking and swerving.

All that training is delivered in a sterile, off-road and HAZARD-FREE environment. We may learn the technical skills to stop or swerve but there’s no context and no connection to a real emergency. We’re simply swerving round cones.

It’s so sterile many riders who pass the test never even make the connect between the swerve manoeuvre and the situations on the road where a sudden change of direction might be needed – to avoid a pothole, to dodge an emerging car, to correct for an unexpectedly tight bend.

How do we make that connection between ‘skill’ and ’emergency’? Simulators would be a huge help.

There’s a second issue. We can be told that the roads are risky places, but our experience teaches us the opposite – that as a general rule, “things don’t go wrong”…

…until they do.

Pedestrians crossing the road ahead of us keep going. They don’t suddenly stop and ‘shuffle backwards and forwards’ when we’re just fifteen metres away. We’re even TAUGHT that this is what happens in the DVSA theory test. Millions has been invested in changing the theory test to use CGI rather than video footage, but watch the videos.


You may have to mouse-click (there’s a skill we use all the time on a motorcycle) on the ‘developing hazard’ but what happens if you don’t? The hazard FAILS to turn into a life-threatening situation. The video does not stop with GAME OVER flashing on screen. Life – quite literally – carries on.

Rather than invest in simulators to train riders and drivers for things that WILL go wrong, we still want to blame the rider or driver for failing to cope with a one-off event, something they are highly unlikely to have experienced before, and will almost certainly never experience again.

With a few exceptions like this remarkable decision, it’s still broadly assumed that it’s ‘bad riders’ (and bad drivers) who have crashes, and that the solution is ‘better behaviour’ and ‘better skills’.

I make no apologies for saying this over and over because it’s crucial to understanding why it was your best riding buddy, the rider who was “such a safe rider and never in a rush” or that rider from your advanced group, “the one we never thought would have a bad crash”, who actually did come to grief.

The aviation world has known about the Startle Effect for a couple of decades at least. Road safety?

It’s ignored.

Or – if not ignored – known only to insiders, as a comment on my original Facebook post made clear. I was contacted by a retired police accident investigator. He said “I think that you are wide of the mark when claiming that the startle effect is not known in the world of road safety. We may not have called it that a quarter of a century ago, but we took it into account.”

And he went on to explain that: “the average ‘reaction’ time for a normal driver to realise what was going on when faced with an unusual problem, formulate a plan to deal with it, and initiate such a plan was taken to be in the region of at least 1.5-2 seconds”.

Now, put that in perspective with what I’ve just written. If we take that upper limit of 2 seconds to “realise, formulate and initiate” then subtract the 0.7 seconds which is the average physical ‘reaction’ time for a driver or rider who KNOWS what he or she is going to do already, then it’s allowing A MAXIMUM of 1.3 seconds to actually recognise and analyse the situation, AND to come up with a plan to deal with it.

He continued: “for a driver to considered as being careless, one would be looking for a ‘reaction time’ of between 2-4 seconds.” Or having deducted the physical reaction time, if it takes you or me between 1.3 to 3.3 seconds to “realise, formulate and initiate” the response to the situation, we’re deemed ‘careless’, despite this delay being entirely within the limits of the cognitive delay – the SURPRISE! factor – created by the Startle Effect.

The implication is that riders and drivers are not allowed to be startled by events! And you’ll notice that in the failed prosecution, the police who prepared the case clearly weren’t allowing the rider anything other than an instantaneous reaction to the pedestrian’s unexpected hesitation ahead of him.

In any case, I’d say that being ‘aware of’ of the Startle Effect at the forensic level of police accident investigation, and the failure to ‘factor in’ of the Startle Effect at any level of rider and driver training in order to create road users who are more prepared to deal with the unexpected and avoid the mental block caused by SURPRISE! are two very different things.

Currently, we still seem to want riders and drivers to be able to respond like machines, and if they do get caught out reacting like humans, the response is that they should be punished…

…as if that will help the next pedestrian who hesitates in front of a startled rider who wasn’t expecting to have to take emergency action.

…or the next rider who falls victim to a ‘Sorry Mate I Didn’t See You’ error.

The Startle Effect part 1

At the end of March 2021, a court verdict in the Manchester Evening News drew my immediate attention because it appears to recognise that the Highway Code’s advice on stopping distances is – as I have said for many years – inadequate and misleading and unfairly penalises riders and drivers who have crashes.

On the face of it, it appears to be a tragic but fairly standard incident.

An elderly pedestrian steps into the path of motorcyclist… the motorcyclist is unable to stop… the pedestrian is seriously injured and dies two days later… the motorcyclist spends three days in hospital but is charged with “causing death by careless driving”.

Rider David Hurst, 52, pleaded not guilty to the charge relating to the death of Ghulam Rasul, 74, at an earlier hearing in Bolton back in November 2019. The case was finally heard this week.

The prosecution case was also familiar. The prosecution compares the real-time actions of the rider with the steps that a hypothetical ‘prudent’ rider would have behaved in these circumstances. In essence, the argument put forward was this:

:: having seen what was happening the rider should have taken steps to avoid the collision
:: those steps should have been obvious given the situation developing ahead of the rider

Since the rider DID hit the pedestrian, the train of thinking is that the rider SHOULD have taken EARLIER STEPS to PREVENT the collision.

Mr David Lees, prosecuting, said “as [Mr Rasul] was crossing the road, I would have expected Mr Hurst to keep an eye on him.” He argued that Mr Hurst “should have applied his brakes earlier” and “should have slowed down to give himself time to minimise the risk”.

Mr Lees also stated that “he could and should have taken some action such as moving the bike away from Mr Rasul but he did not…. he was beholden that he took a number of actions. He should have slowed down well before the point that Mr Rasul stopped walking… Mr Hurst should have turned his wheel away from Mr Rasul but didn’t.”

The key claim was that the rider had only reduced his speed ‘at the point of no return’ and the prosecution’s conclusion was therefore that “Mr Hurst PUT HIMSELF in a position where the collision became unavoidable”. [My capitals.]

This is, of course, looking at the circumstances with two huge advantages:

:: the benefit of hindsight to look back at potential decision points where the rider COULD have made different choices
:: the benefit of time in which to analyse the situation and show those different choices EXISTED in the first place

Fortunately for Mr Hurst, he had two expert witnesses on hand who reviewed the CCTV footage and put forward a contrary opinion – that there was nothing to suggest that Mr Hurst was NOT alert and there was nothing INAPPROPRIATE with his speed and driving. [My capitals.]

And in what seems to have been the crucial piece of evidence, Dave Poole, a retired accident investigator for Greater Manchester Police, brought up the topic of individual reaction times.

The circumstances of the collision were that Mr Hurst and Mr Rasul would NOT have collided until Mr Rasul hesitated directly in the path of the motorcycle and in the word of the defence council “shuffled” from side to side in a period of indecision.

In his view, Mr Hurst would have been aware of Mr Rasul crossing the road ahead of him but, when he stopped and hesitated, “it would have created a new element for Mr Hurst presenting him with a new set of considerations.”

And this turned out to be crucial in Judge Paul Lawton’s view.

He said: “At 38.5 metres we know that Mr Hurst had seen Mr Rasul and his brake light came on.”

Given that the judge will go on to talking about reaction time, it’s helpful to convert distances to time.

A bit of research on Googlemaps indicates that the speed limit is 30 mph. None of the evidence suggests that Mr Hurst was travelling at more than 30 mph, so if we assume that the bike was moving at 30 mph, then the rider would have been travelling at 13.4 metres per second.

So at a speed of 30 mph, and at a distance of 38.5 metres, the rider reacted by applying the brakes a minimum of 2.87 seconds before the collision.

If we look at the Highway Code you’ll see that the stopping distance from 30 mph is given as 9 metres ‘thinking time’ + 14 metres ‘braking distance’ adding up to a total stopping distance of 23 metres.

The 9 metre ‘thinking time’ equates to a delay of 0.7 second before the brakes are applied. And some fairly straightforward trials of a modern bike would suggest that a modestly-competent rider should be able to stop in perhaps ten metres once the brakes are on, and thus pull the bike to a stop a bit sooner than the Highway Code suggests.

Back to the evidence.

At the point when Mr Hurst first braked he was well outside this ’emergency’ stopping distance. The judge observed: “No motorist has to do an emergency stop because they anticipate a hazard 40 metres in front of them. If everyone did that it would cause more accidents.”

And here’s where it all went wrong.

Judge Paul Lawton noted that “at 14.6 metres Mr Rasul froze which was an understandable response but ENTIRELY UNPREDICTABLE.” [My capitals.]

Now, we don’t know how fast Mr Hurst was travelling at this point, all we know is that the brakes had been applied, so presumably SOME speed had been lost – remember the expert witness said that there was nothing to suggest that Mr Hurst was NOT alert and there was nothing INAPPROPRIATE with his speed.

But since Judge Lawton went on to say that Mr Hurst could have stopped with a 0.5 second reaction time, the implication is that he’d probably slowed to around 20 mph at this point – the Highway Code offers a 12 metre stopping distance from this speed. If he had braked hard IMMEDIATELY he would have stopped with a couple of metres to spare.

But… and this is the crucial point… this 0.5 second reaction time is based on a event where the rider has ALREADY RECOGNISED THE SITUATION as one requiring emergency braking and REACTS INSTANTANEOUSLY.

And that’s just not how the human brain copes with suddenly-changing circumstances. We have to interpret the new data and come up with a new plan. And that all takes TIME.

As I’ve said many times for many years:

“…people still have these accidents in 30 limits where conventional wisdom re stopping distances says they should be able to stop easily and the reason is they are taken by SURPRISE!

“There’s usually a three second window from the time things start to go bad to the moment of impact. The bad news is that it can take our brains up to TWO SECONDS to realise that things ARE going wrong – what I call ‘recognition time’. This leaves a totally inadequate one second to deal with the emergency. Even at 30mph that’s not enough and is why so many riders collide with vehicles in urban areas…”

Or in this case, with the pedestrian, with the tragic consequences resulting.

I wrote those two paragraphs on Facebook all the way back in September 2014, so it’s hardly ‘new thinking’.

But prosecution cases continue to be built, then delivered to courts, based on mistaken belief that riders and drivers ARE ABLE TO RESPOND INSTANTLY. And juries – not surprisingly, since they are being led to this conclusion by the prosecution’s evidence – continue to convict riders and drivers for failing to react instantly.

So having read that the judge said that Mr Hurst COULD have stopped with a 0.5 second reaction time, you may imagine just how surprised I was when Judge Lawton qualified that statement by stating that the NORMAL reaction time was 0.7 to 2 seconds. [My capitals again.]

And when I read that he stated that “The court is to award him the two second reaction time” I was absolutely gobsmacked.

And when I read his concluding words, I fell off my chair…

“The jury are being asked to put themselves in Mr Hurst’s position.

“I have concluded that it is impossible for a jury to do that with the evidence put before them.

“Considering the evidence given and the dynamics of the accident I withdraw the case from you at this stage.”

In essence, what this judgement does is accept that SHOCK AND SURPRISE exist on BOTH sides of an incident like this.

If the judge could accept that “at 14.6 metres Mr Rasul froze which was an understandable response”, then his judgement allows for a similar frozen response in the rider as Mr Rasul unpredictably stopped moving in front of the bike.

As far as I know – and I am far from a legal expert – this is something of a precedent.

I do know some people will undoubtedly see this as a terrible decision, arguing that it’s a charter for careless riders and drivers to mow down vulnerable road users.

It’s not, it’s a clearsighted acceptance that the human brain doesn’t function in the straightforward linear “see hazard, do something, avoid crash” fashion that ‘road safety’ generally assumes. Instead, in unexpected circumstances, there is a ‘Startle Effect’ which temporarily paralyses us.

And it allows that it’s wrong to blame ordinary people who find themselves caught out in exceptional circumstances and who, being forced to make a decision, make the wrong choice.

So where do we go from here?

We can try to make the system safer – that’s more 20 limits in streets busy with pedestrians, barriers to prevent pedestrians crossing other than at designated crossing points, passive safety design changes to cars.

But we can also change the way road users think about safety.

The answer is that we should be shifting the way we think about road safety generally, away from a belief that it’s the behaviour of a few ‘bad apples’ cause crashes and towards system that accepts that most crashes involve ordinary drivers doing ordinary things in everyday situations that for some reason go wrong in this particular set of circumstances.

What’s needed is not to educate drivers and riders “not to make mistakes” but to educate drivers and riders to EXPECT HUMAN ERROR in others, so that when someone does do something UNUSUAL, it’s not UNEXPECTED and the rider or driver does not suffer that cognitive delay whilst they recognise things aren’t going as planned.

If Mr Hurst had a mental picture in his head that Mr Rasul MIGHT stop in his path, he may well have reacted instantly and avoided the collision and Mr Rasul would still be with us today.

Nothing can compensate Mr Rasul for the loss of his life. Nothing can compensate his friends and family. Nor will fining, banning or even jailing Mr Hurst – there’s a potential three year jail term for the offence he was charged with – make the roads safer for anyone.

Like ✔ Comment ✔ Share ✔ Follow ✔
FIND OUT how you can get Survival Skills!

visit the website at

survivalskills #advancedriding #skillsonsaturday #tipsontuesday #focusonfriday #nosurprise #scienceofbeingseen #SOBS #CrashCourse #shinysideup2021 #RideForever #ACC

Photo Credit: Charles D P Miller CC BY 2.0,

Covid-19 Update – join our Online Sessions!

Join our Online Sessions!

As you have probably realised, my face-to-face training is on indefinite hold in response to the coronavirus outbreak.

But that doesn’t mean that I’ve shut up shop completely.

Survival Skills has gone online!

WEBCHAT – ‘Elevenses’
Join me, Kevin Williams, on Monday, Wednesday, Friday and Sunday at 11am for topic news, controversial views and two-wheel tips on my new ‘Elevenses’ LIVE show – pop along to my free-to-view Facebook page and set a reminder!
VIDEOS – YouTube channel
I’ve had a YouTube channel for years but have always been too busy to fully exploit it. But now with some time on my hands I’m starting to populate it with videos covering various aspects of biking skills. It’s a work in progress at the moment, but bookmark it and come back to it at intervals to see what’s new:
VIDEO SURGERY – the doctor is IN
I’m also offering 1:1 online video sessions from my virtual office. Got a riding issue or a question about biking skills? Drop me a line here and I’ll organise a live webinar with you. No charge, just an invitation to contribute a donation to support the work I do:
Want to get your skills in tip-top shape ready for when we can all get back out on the road? Then take a look at the brand-new Survival Skills
e-course that delivers the content of the Performance: BENDS one-day course in an online environment. View the info page with no obligation and the first module is completely free. Find it here:

For any questions on any of the Survival Skills online activities, drop me a line.

Looking forward to seeing you online!

Kevin Williams
Survival Skills Rider Training
…because it’s a jungle out there


Riding a motorbike is all about acquiring and learning motor skills, but there is always the fundamental problem that we can’t learn motor skills from a book. Good advice and descriptions of process and technique can be learnt and stored as memories in the neo-cortex thanks to the Hippocampus, but motor skills can only be learnt through acting, doing and feedback which is the role of the Basal Ganglia and the Cerebellum. When we recall things from memory to describe an action process we tend to recall that which has been stored via the Hippocampus rather than recalling the vastly more complex motor actions we actually do use when riding.

Our Basal Ganglia and Cerebellum don’t much like giving up the secrets of motor actions because they have not been stored in nice linear sequences like the language that we used to memorise advice. Instead they are stored as a vastly complex and interconnected sequence of patterns that have been formed via the input of every sense simultaneously. A good example of this is trying to describe the sequence and combinations of actions required to walk from one side of the room to the other. Most of us are lucky enough to be able to walk, but precious few of us could describe the process in anything other than the most basic and simplistic terms. Analysing complex motor actions and then breaking them down into linear sequences that can be written down or spoken so that they can be learnt by somebody else is a really difficult job and not one to be taken lightly.

Luckily we humans don’t need all that book larnin’ in order to become very good at learning motor functions. What we do need however is a very powerful goal/feedback system that tells us whether the actions we have carried out are the right ones needed to reach a goal. The more precise the goal and the more overt the feedback system then the quicker we can learn the correct motor actions and the quicker we can store them away ready for automatic use the next time we need them. In the world of motorcycling decent goal/feedback systems are noticeable by their absence due in the main to a lack of ‘secondary goals’. A secondary goal is obvious in the game of Golf for example where the primary goal of the game is to get the little ball into the hole, but the secondary goal is to get the ball into the hole in as few a strokes as possible. All sports have secondary goals of one kind or another and it is these secondary goals that are used by sports coaches and psychologists to help people become master’s of the sport.

When you have precise secondary goals and plenty of overt feedback the rate of skill acquisition starts to go off the chart, but when these secondary goals are missing, progress if any is always going to be painfully slow.

In my own sport of Moto Gymkhana we have an exercise called GP8 which is five times round a figure of eight course. This is a doddle for most people, but as soon as a secondary goal is introduced, which in this case is doing it against the clock, the entire task takes on an entirely different complexion. Thanks to this addition we know that the fastest time for doing the exercise is 26 seconds, so a rider that takes 50 seconds to do it has some idea of how much more they have yet to learn. What road riders really need then is some form of goal/feedback system that they can use to improve their motor skills, because without one they really aren’t going to become as good as they need to be.

Safety II – Things go right and things go wrong in basically the same way.

The study and practice of safety tacitly assumes that we know how things are done or should be done. Since humans are supposed to follow procedures, rules, and guidelines, accident investigation and risk assessment alike assume that compliance will always lead to successful outcomes. The purpose of safety analyses is consequently to understand why the outcome of an action or a series of actions (such as riding a motorcycle) was unacceptable (adverse) rather than acceptable (successful) – as in event investigation – or how that could possibly happen in the future – as in risk assessment.

In reality riding a motorcycle is never completely regular or orderly, except in very special cases. It is therefore inadvisable to assume that the riding task is as we imagine and that compliance guarantees success. Work-as-done will always be different from work-as-imagined because it is impossible to know in advance what the actual conditions experienced by a rider will be, not least what the demands and the resources will be, which means that it is impossible to provide instructions that are detailed enough to be followed ‘mechanically.’ A safety analysis must therefore begin by establishing how riding is actually done, how everyday performance takes place, and how things go right, as a prerequisite for understanding what has or could go wrong.

The reason why everyday performance nevertheless in most cases goes right is that people know or have learned to adjust what they do to match the actual conditions, resources, and constraints – for instance by trading off efficiency and thoroughness. The adjustments are ubiquitous and generally useful. But the very reasons that make them necessary also means that they will be approximate rather than precise. Approximate adjustments are the reason why things usually go right, but by the same token also the reason why things occasionally go wrong. Things do not generally go wrong because of outright failures, mistakes, or violations. They rather go wrong because the variability of everyday performance aggregates in an unexpected manner.

Whenever something is done, the intention is always to do something right and never to do something wrong. For each action, the choice of what to do is determined by many different things, including competence, understanding of the situation, experience, habit, demands, available resources, and expectations about how the situation may develop – not least about what others may do. If the expected outcome is obtained, the next action is taken, and so on. But if the outcome is unexpected, then the preceding action is re-evaluated and classified as wrong rather than right, as an error or as a mistake, using the common but fallacious post hoc ergo propter hoc argument (since event Y followed event X, event Y must have been caused by event X). With hindsight, it is pointed out what should have been done, if only people had made the necessary effort at the time. The whole argument is, however, unreasonable because the action was chosen based on the expected rather than the actual outcome. Failures and successes are equivalent in the sense that we can only say whether the preceding action was right or wrong after the outcome is known. That changes the judgement of the action, but not the action itself.

It’s surprises that count so count your surprises!

Counter 2

It’s surprises that count so count your surprises (you’ll be surprised how many surprises you count)!

As we all should be aware by now, the surprise tells us that we have failed to correctly predict the future state of the system and as that is the only job we actually do understanding what surprises can tell us is going to be critically important. It’s also important to know that not all surprises lead to accidents but all riders involved in accidents will have been surprised at one point or another during the process. Armed with this information it’s then possible to turn surprise to our advantage as part of an effective self-training/learning strategy.

All surprises represent an accident that would have happened were the circumstances prevailing at the time been only a tiny bit different.

In cases such as that we have clearly made a prediction error, been surprised by the resulting system state, but have then been able to make sufficient adjustments and adaptations to avoid any accident actually happening. It doesn’t always follow though that we can make sufficient post-surprise adjustments and adaptations because that’s how accidents finally happen. That doesn’t concern us at the moment though as it’s those surprises that we successfully manage that can help us to avoid having surprises we can’t manage.

A good question to regularly ask ourselves is “If I was to ride through the same section again and under the same circumstances would I do exactly the same things?” If the answer is “yes I would” then we would have learnt nothing from our first experience, but if we would have done things differently, then the original experience has taught us a valuable lesson. We would only do things differently the second time around because one or two events happened the first time that we didn’t predict and so we would change what we did in the light of the new knowledge. This is the essence of learning by experience and if we can accelerate this ability then we will start to rapidly acquire new expertise. Of course the problem we have as road riders is that once we have passed through a section we don’t often go back and have another go at it and even if we did the circumstances would usually be quite different to the first time we rode it. Rather than go back and have another go what we can do is to count the number of surprises that we actually had during a ride and that would give us a very good indication of any shortcoming in our ability to make accurate predictions of future states. More surprises indicates that our predictions need work and fewer surprises indicates that we are getting the hang of things.

In his book ‘The Upper Half of the Motorcycle’ (see reading list) Professor Bernt Spiegel dedicates a couple of pages to the use of what he calls an ‘error counter’ as a really beneficial learning aid and as errors also generate surprises then it’s perhaps something we can use. What he suggests we do is to mount a cheap tally counter somewhere near the left handlebar and click it every time we get surprised for any reason. At the end of the ride we can see at a glance how many surprises we had during the ride and that will give us an indication of how many potential accident situations we got ourselves into, but were able to manage our way out of. The idea is that the fewer surprises a rider has the better the rider is and that’s something all of us aspire to.

Professor Spiegel has made an extensive study of how the error/surprise counter works and he found that after a short period of using the counter the number of errors/surprises recorded actually increased! What he realised was that the number of registered events increased not because the user’s riding was getting worse, but that they had become more aware of the number of errors they were actually making. Before the use of the counter he found that most errors and surprises seemed to slip by almost un-noticed, but once a rider started to actively hunt for them even the tiniest error or surprise couldn’t slip through the net. Eventually of course the number of registered errors and surprises started to decrease as the rider stated to learn what it was in the system that precipitated the errors and surprises in the first place. The increased sensitivity to error and surprise resulted in a commensurate increase in the rider’s predictive capabilities and a noticeable decrease in the number of prediction failures they made.

Considering that the whole point of the Nosurprise campaign is to help riders to become better predictors then a simple device like the error/surprise counter is something that we would thoroughly recommend.


What the motorcycle manufacturing industry can tell us about why motorcycles crash.

Considering mechanical failure accounts for so few accidents nowadays you would have thought that the way bikes are built and the way they are ridden wouldn’t have much of a connection. You would be wrong in this assumption however because bikes are built and bikes are ridden in what are commonly known as ‘systems’ and how systems work and sometimes fail to work is critical to our understanding of accident causation. Continue reading

Why do riders crash in corners?

Why do riders crash in corners? After all, had they not successfully negotiated a lot of other corners before they reached the one that got them? Was there something about certain corners that made them more likely to be accident sites? Why did these riders select a particular corner entry speed that proved to be so spectacularly incorrect? What is the process that we use for judging the severity of a corner and selecting a suitable entry speed? Do we all use the same method, or are there a number of ways in which we can analyse a corner before we reach it? Continue reading