Sorry Mate, I didn't see you - an analysis of SMIDSY accidents
When I updated these articles in the summer and autumn of 2019, I realised that in some ways, this article is probably the most important I've ever written. It was written in response to the frequent and strident claims that 'drivers don't look properly for bikes'. And it was in this very article - written in the early 2000s - that for the first time I put down in reasonably clear terms an explanation of the need for riders to understand the visual perception issues behind the 'Sorry Mate, I Didn't See You' collision. Historically, the rider has always blamed the driver for not looking properly, but my background of SMIDSY-dodging as a London-based courier plus my increasing experience as an instructor made me wonder why the advice to drivers to "look harder for bikes" and the advice to bikers to "make yourself more conspicuous" wasn't working. The failure of the 'Think Bike' and 'Ride Bright' advice - which dates back to the mid-1970s - became very evident when I began to investigate collision statistics - the proportion of junction collisions had remained unchanged from the early 70s (when no-one used day riding lights or hi-vis clothing) to the time when I wrote the article in 2003.
I'd already read a lot of research papers as a way of developing my training courses and had discovered quite a lot about motorcycle conspicuity and the reasons for car / bike collisions when I was invited to work with Kent Fire and Rescue Service on the 'Biker Down' course.
So when, in 2011-12, I created the 'Science Of Being Seen' presentation (or SOBS for short) it was this research which formed the basis of the presentation. Perhaps not surprisingly given that motorcyclists have seen a stream of road safety campaigns all aimed at drivers telling them to 'look harder' or 'look twice' for bikes, we have tended to believe that the reason for the SMIDSY collision is because "drivers don't look properly". SOBS shows that's not true, and explains the real issues facing the driver - 'looked but COULD NOT see', 'looked but FAILED to see' and looked, saw and MISJUDGED' errors - why the conspicuity strategies we motorcyclists have employed - hi-vis clothing and day-riding lights (DRLs) have failed to have any meaningful impact on collision statistics. It's why I suggest that it's down to us riders to take responsibility for evading the driver's error when it happens.
Since then, I've continued to investigate the problem of motorcycle perception and visual perception, and the presentation which I continue to deliver at nearly every Biker Down in Kent, has been continually updated with the very latest research. And in terms of collisions, nothing much has changed since, as it happens.
But this article is where it all started, getting on for two decades ago. So although I've annotated the article in places, the basic text is left unchanged apart from a couple of minor typos I've corrected. And if you want to read the very latest thinking, then head for the SOBS website at http://scienceofbeingseen.wordpress.com.
Most bike riders these days also have a car licence and drive a car, usually as their main means of transport, using the bike for fun or sometimes commuting. Yet to listen to a lot of the discussion that goes on about "witless cagers" you'd be hard-pressed to realise that.
But given that we nearly all drive cars, and our old friend the SMIDSY accident still accounts for the majority of car/bike accidents aren't we likely at some time or another to have made exactly the mistake that we pillory drivers for? How many of us when on four wheels have done the unthinkable and pulled out on a bike?
As one honest motorcycle forum contributor admitted [after a near-miss in his car]: "Now if I can do this, what chance for the poor booger in his Mondeo who has got no idea of what we are about... what still bugs me is that, if they'd run into me, I'd have heard myself saying, in total honesty, as I helped sweep them to the side of the road: 'Sorry mate, I didn't see you'".
I've mentioned before that I nearly took out an R1 when they first came out... poor gloomy light, twin headlights apparently a long way off against a background of trees, me wanting to pull across the path of the oncoming vehicle and turn right, so all I needed was a gap sufficient to make it to the other lane.
Seemed safe enough so I started to go...
...but something wasn't quite right about the movement of the lights across the dark background and I hit the brakes again, stopping about halfway across the line.
Just as well I did! By the time I'd refocused on the oncoming vehicle, it was obviously a bike, moving at a fair lick, and MUCH, much closer than I had realised.
Two thoughts struck me at the time. The first was that the widely spaced lights on an R1 DO look like a car further off - I went home and even on the GSX-R where they are much closer together, immediately put a different coloured bulb in one headlight - technically illegal but it's my safety I'm worried about here.
The second was that the rider hadn't apparently reacted to me at all. He was just going to sail completely oblivious into the accident I was about to cause. Yes, technically my fault, but did he have to have it with me? Could he not have done something positive himself? There was no blast of high beam and/or horn, no anchoring up, no swerve to the other side of the road (it was clear, remember or I wouldn't have been about to pull out).
[NOTE - "technically my fault, but did he have to have it with me?"... echoes of my very first article for the Motorcycle Action Group newspaper in 2002 - "it takes two to tangle" - the driver may be the one setting up the crash, but the rider still has to ride into it to complete the collision. And, as in the near-miss I had with the R1 rider, the rider can nearly always see it coming.]
OK, so let's take a reality check.
Cars do pull out on bikes. Fact.
In around 90% of them, the bike is on the priority road, so technically it's the car driver's mistake. Fact.
But if we, as bikers, can STILL make that mistake when on four wheels, knowing all we know about car drivers doing it to us when we're on two, it's worth looking at in more depth.
I've previously suggested proactive strategies for dealing with SMIDSY incidents, but let's ask some questions about why drivers don't see bikes. If we can understand why things go wrong, it may make more sense as to why it's US as riders that have to deal with the situation, rather than use the "it was the other guy's fault, I had right of way, he should have seen me" excuse.
There are a whole bunch of reasons to worry when you approach a junction:
There is the driver with simple defective eyesight - plenty of them around...
There is the driver who doesn't look properly - too many in-car distractions, be it children running amok, the mobile phone demanding immediate attention or just singing along with Des O'Connor.
[NOTE - although 'defective eyesight', 'driving distracted' and generally 'not looking properly' seem likely reasons to explain the 'Looked But Failed To See' LBFTS error where drivers don't see approaching bikes, when I began to look into the issue in more depth it occurred to me that the vast majority of drivers DO see the vast majority of bikes. If they didn't, we'd not make it far past the first junction.
In total, there are around 350 motorcycle fatalities and some 3000 injuries each year, but they are the result of ALL crashes, not just those at junctions - they total around 100 per annum. But what about encounters that DON'T end in crashes? If we think about how many cars there are on the road (around 40 million) and how many bikes there are (between 1 and 2 million), consider how many junctions every biker passes on every ride, then work out how many times bikes pass through junctions where a car could turn, the number of POTENTIAL collisions that never happen is truly enormous - I don't think anyone has actually attempted to do the sum. The only conclusion we can make is that drivers DO see nearly every bike when it needs to be seen. And if that's the case, the only rational explanation is that nearly all drivers DO look properly on nearly every occasion. We can lay to bed the 'not looking properly' explanation - it's a handy myth.]
There is the driver who does see you but chooses the wrong course of action - is the driver inexperienced, merely incompetent, or not used to the vehicle being driven? Ever had a car towing a caravan pull out in front of you and wondered why? I never get anywhere near hire vans for the same kind of reason...
There is the experienced and overconfident driver who looks, thinks he has seen everything but "blanked" the bike because he only sees what he expects to see. New drivers and experienced drivers score very differently in hazard perception tests - new drivers check EVERYTHING in sight but cannot prioritise, experienced drivers check SELECTIVELY, prioritise better but often miss the unusual (ie the bike)...
[NOTE - it turns out there's an explanation for this too. It doesn't take long for all road users - bikers included - to develop a different strategy for emerging onto busy roads to the one we're taught. Rather than the search for vehicles (which is what we think we're doing), we're actually all searching for the gaps between them. We all do it, drivers and riders alike. Mostly, it works.]
There is the driver who makes a conscious decision to use you as the gap in the traffic, knowing you will give way - "the bike is softer than a 44 tonner" approach...
[NOTE - I looked into that, too. Evidence from insurance statistics - who you might expect to be looking for a reason to pin a collision on the other driver - suggests it's actually very rare for a driver to pull out deliberately. We riders tend to interpret it that way because we frequently see the driver appearing to look at us - "I made eye contract but he / she still pulled out" is a common post-crash statement. But as you'll see if you follow up the SOBS website and check out how the eye actually has a tiny zone of clear, colour vision and sharp focus, it's entirely likely that what we thought was eye contact, was actually the driver was looking in our direction but focused on the vehicle or gap behind us. I mention this below.]
Even given that the driver knows what to look for, is actively looking for it, knows what to do and isn't a chancer, doesn't mean he'll see you coming. There are a number of reasons.
Most modern cars have huge blind spots:
- take a look at the size of the A pillar alongside the windscreen on a modern car. They are designed to make the safety cage of the car rigid in an accident and stop the roof from folding up - it's no coincidence they are the size of girders!
- take a look at the pillar behind the driver's head where the doors come together - again it's huge
- take a look at the pillar behind the rear window - once again it is part of the safety cage
Depending on the angle the car takes up, it's quite possible the driver cannot see through you one of these obstructions, and there is always roadside furniture like telegraph poles, trees and letter boxes - if you can't see his eyes, he cannot see you.
[NOTE - and since I wrote this nearly two decades ago, the A and B pillars have got even thicker as a result of new crash protection requirements. It turns out that around one-in-five collisions actually fall into the 'looked but COULD NOT see' category. In the run-up to the collision, although the driver was searching for approaching vehicles, the rider simply hadn't put the bike in a place where the driver could see it.]
But drivers still don't see you when they are looking straight at you and you are in clear view. Why not? Two possible causes. An accident analysis I saw the other day suggested that a contributory factor was "visual clutter" - there was so much going on in the direction the driver was looking that she simply didn't see the bike. The brain was incapable of processing all the information being thrown at it in the time available and bits went astray. Unfortunately, amongst that lost info was the bike.
[NOTE - this phenomenon of processing information has been investigated in other fields - notable aviation - and the sum of the tasks that have to be performed is known as 'workload'. Only recent has research in workload in driving been carried out, and it shows that in typical driving situations, there's too much for the human brain to process all at once. So we 'task-shed' and focus on only part of the driving task. The very latest research (September 2019) suggests the more that's going on, the more likely drivers are to forget what they saw a moment earlier. Motorcycles seem to be particularly prone to going missing. This is not carelessness or 'not looking properly' either. It's simply the way the human brain evolved which limits our ability to process complex information.]
The second possibility is down to the way the eye and the brain work in tandem to process visual information. It may mean we see things which aren't there or be blind to things that are. Even a conscientious driver, looking carefully, may misinterpret what he sees.
The central part of the retina is what sees detail in sharp focus - it's why you have to look directly at a piece of paper to read what is written on it, but both this and the zone outside this is very sensitive to movement. Try this simple experiment - your eyes will have to move word by word to read this sentence, but if you move the mouse you can see it move over the whole of the screen wherever your eyes are focussed.
[NOTE - and there's a bit more to this than I realised at the time. As our eyes move to points of interest (the words in this case), they move in jumps and pauses - saccades and fixations. The fixations allow us to focus and pick out the detail of the letters so we read the word. What's not obvious - although we've known about it since the 19th century - is that as the eyes move in a saccade, the visual system shuts down. We're effectively blind as our eyes move between fixations. It's known as saccadic masking and is now at last being recognised as a real problem when drivers are turning their heads and looking left and right at junctions.]
As you ride, you'll often spot motion out of the corner of your eye (a plastic bag flapping in a hedge or a car approaching in a side road) whereas the driver looking back at you is using the sharp focus part of the eye and may not see you because you don't appear to be moving.
[NOTE - the brain is good at picking out movement in peripheral vision - it's how our visual system is designed - but approaching a junction on the bike, we're on a near-collision course with the driver looking in our direction. That means we're virtually motionless with respect to the background scene, and that means we create no lateral movement to trigger the brain's motion detection system. It's known as 'motion camouflage'. Interestingly this phenomenon has been known about for decades by animal scientists, sailors and fighter pilots, but only recently does it seem to have been realised it applies to drivers too.]
How might he miss seeing you? The brain spots familiar objects by using pattern recognition - as social animals we are very good at recognising faces. As drivers/riders we've trained our brain to recognise other important shapes - the silhouettes of another vehicle, the outline of a pedestrian, the pattern of a road sign. The problem is that we learn to recognise these patterns as whole - break up the outline and it vanishes - try recognising a face which is missing the eyes or the mouth! One VD contributor posted an excellent picture of a 'dazzle-camouflaged' ship painted in bold strips of grey and blue - it was invisible not because it blended into the background but because the strips gave the eye false outlines to try to make sense of, none of which said 'SHIP'.
[NOTE - once again, this is 'old news' in science but the effects of 'disruptive camouflage' is only now beginning to be recognised as an issue. It's particularly a problem for motorcyclists because our bikes and clothing are often multi-coloured. It's likely it's a significant factor in 'looked but FAILED TO SEE' errors. Even supposedly hi-vis clothing often fails to create a recognisable silhouette for the driver to see, which may well explain why there's little evidence that hi-vis clothing has had any positive effect in reducing the proportion of junction collisions.]
When approaching a waiting driver, in certain lights conditions or against certain backgrounds, part of your 'bike plus rider' outline may vanish - so the shape that reaches the part of the brain busily processing this information doesn't shout 'BIKE' to the driver's conscious reactions. If you are approaching head on, without adding movement across the background, there is nothing to alert them to the fact they have missed a vital clue until you get very close and the angle of view starts to change.
[NOTE - this problem of foreground and background colours blending is known as 'contrast camouflage'. Guess what? We've known about it and exploited it for military purposes for a couple of centuries. But road safety has focused entirely on the false premise that if riders wear bright colours they'll be more visible. Put your yellow hi-vis vest on, then stand in front of a field of oil seed rape in flower, or a yellowing autumn hedge, and see if you stand out. Two of the most important pieces of understanding are:
it's the CONTRAST that matters, not the colour
the background changes moment by moment and so does our conspicuity
If you want a daytime hi-vis colour that works reasonably well in most environments, it's not Saturn yellow, but pink! I have been suggesting this for well over a decade, so when I took a BikeSafe day with the Met last year, I was mildly amused to the team suggesting pink hi-vis. I wonder where they got that idea?]
And as if all that weren't enough that could go wrong, even if the driver does spot you, how does he go about judging your speed and distance?
Well, if an object is heading straight at you, it's very difficult - switch to sport for a moment. If you've ever tried to make the high steepling catch where the batsman has hit the ball straight up, you'll know that it's not that easy to judge the catch as it comes down again - even the best players make a mess of it. You have to use an estimate of distance based on what your experience tells you about the apparent size of the object, then use the rate of change of the size of that object to determine what speed you think it's approaching at, and when you need to cushion the catch.
By contrast a straightforward lob to the boundary is relatively easy to catch even if you have to run to meet it because we use the movement of the ball across the background to give us an extra angle to calculate where it is in 3D.
The driver sitting looking at a bike heading towards him is in the motoring equivalent of that up-and-down catch. At the high closing speeds possible on a motorcycle, it becomes almost impossible to judge distance, speed and time at all accurately.
[This is what I've called the 'looked, saw and MISJUDGED' error. But as well as the technical difficulty of accurately judging speed and distance, there's an extra problem. Put a bike side-by-side with a car or van at the same distance and travelling at the same speed, and observers will almost always think the van will arrive first. Looking at the bike, they think they have more time, and make the mistake of pulling out. This has become known as the 'size-arrival effect'.]
And whilst we're digesting that, another thing to consider... it's not just driver to your left you have to worry about, what about the driver turning across your path from the opposite direction? You have little time to react and are likely to add the oncoming vehicle's speed to your own, and the driver has to factor in their own speed and distance to the turning point. That accident accounts for a whopping 21% of Killed and Seriously Injured in London, despite being the minority accident. By contrast, vehicles emerging from the left account for only 7% of KSI.
[This was the big lesson I personally learned from BikeSafe. I had no idea that the oncoming driver turning across the rider's path was such a big killer. It's been something I've been flagging up ever since.]
Where's my coat, I'll think I'll take the bus instead!
POSTSCRIPT - Of course back in the early 2000s, what I wrote here kicked off a lot of negative feedback and some stinging criticism, typically suggesting I was "absolving the driver of responsibility", or "making a victim of the rider". It wasn't just motorcyclists either - I was even told by a road safety officer that I was undoing all their good work promoting hi-vis clothing. Even in 2012, my presentation was often greeted with polite disbelief and shakes of the head.
But in the eight years since the first talk, other people have picked up the message and begun to run with it. Biker Down itself has gone national, and is delivered by over half of all fire services, many of which use a version of my SOBS presentation. A year or so after the first SOBS presentations on Biker Down, an RAF pilot compiled a very good article for a London cyclists' magazine - I still reference that article regularly. And more recently, an excellent video has appeared online under the 'Fortnine' moniker on YouTube which covers much of what SOBS began explaining in 2012. Somewhat to my surprise, even BikeSafe in London has begun to cover some of the issues explained by SOBS.
As a result, riders are learning terms like 'motion camouflage' and 'saccadic masking' and the science isn't quite so much of a mystery any more. The more of us saying the same thing, the more credible the message becomes and I've seen that in the response to my presentations, how attitudes have begun to shift.
Perhaps not surprisingly, as more people become familiar with the concepts of visual perception, critics have now started to say that SOBS is nothing new - that we knew all this already. I certainly don't claim that I have contributed any original research to SOBS, but what is unique is that SOBS is most certainly the FIRST TIME anyone anywhere has attempted to assemble the research and present it ALL TOGETHER and in a form that is COMPREHENSIBLE to the average rider.
I personally have delivered the SOBS presentation to several thousand attendees on Biker Down in Kent, and many of the fire service Biker Down teams deliver a version of SOBS. Outside of Biker Down, I've personally taken SOBS to rider groups across the south of England (so drop me a line if you'd like a presentation delivered to your own group).
And in 2018 and 2019, SOBS achieved international recognition as I travelled to the other side of the world, to New Zealand. At the invitation of the NZ Transport Agency, the Ride Forever training scheme and the Accident Compensation Corporation, I was a keynote speaker on the Shiny Side Up roadshow that toured the county in both years, giving my talk to hundreds of Kiwi bikers at over a dozen venues on both trips.
Kevin Williams
Survival Skills Rider Training...because it's a jungle out there
WANT MORE?
If you have enjoyed these Survival Skills articles, you can help me stay awake and keep writing. Just click the button below to buy me a coffee!
SHARE THIS PAGE!