Cycling UK Forum

George Riches wrote:There's also the factor of how well lit. There are quite a number of feeble lights out there.

Didn't the CTC run a big campaign back in the 1930's claiming that making back lights compulsory on bicycles wouldn't make a difference to the number of crashes?

Feeble lights doesn't fully explain it. If all lights were so feeble that they made no difference, you'd expect an equal chance of lit and unlit cyclists being hit. You still wouldn't get lit ones being more likely to be victims.

The CTC campaign is often misunderstood. AIUI, it went like this.

Until WWII, only front lights were required. With the bombing, these were a risk because they gave away the location of roads and hence navigational references. Therefore lights were shaded and reduced. With the reduced levels, rear lights were made compulsory to reduce risk.

With the bombing over, near the end of the war, proposals were made to make rear lights compulsory permanently. The CTC opposed the legislation as framed because it was putting the responsibility on the victim. Without accompanying it with continued enforcement of existing standards, the bar would be raised on visibility, but the benefit would go to drivers, who could (subconsciously) reckon to go faster while still seeing cyclists in time.

This is pretty much what happened. If lights were introduced but the safety standards stayed the same, you'd expect drivers to be prosecuted equally whether or not the cyclist had lights, since drivers should still, as they had always been, be expected to see unlit riders. Of course, sensible riders would still increase their chances by using lights, and I don't think the CTC thought otherwise. However, you wouldn't hear "It was his own fault, he had no lights" because that wouldn't be enough to affect the blame.

Now in practice, on today's roads, life is more difficult for cyclist and driver alike, and the great majority of us who are drivers may be nervous at the prospect of unlit vehicles buzzing around for us to see. However, the underlying attitude shift has still happened, and the responsibility is still more on the road user who is struck than it would have been had the CTC succeeded.

And of course, to get back on topic, it's not clear exactly what or how much effect lighting has...

Thanks for such an excellent summary of the much misunderstood CTC campaign against compusory rear lights.

One more thing. It wasn't just for pedal cycles that the law was changed and to which CTC objected. Before WW2 no vehicle was required to have rear lights, and the possibility of running into the back of big, black, slow-moving coal wagon, made any self-respecting motorist concentrate wonderfully!

There's not enough info to reach any conclusion and there is a flaw in the logic of the original argument. We need to know how many night time riders there are and of these how many ride with and without lights. (plus a whole lot of other things already mentioned.)

But the main point is: what proportion of non-lit riders does that 10% represent? It could be as high as 100% or as low as 0.01% or lower. And what proportion of lit riders does the 90% represent? Until we know at least that then no conclusion about relative safety can even be guessed at.

We've been here before in this thread. There isn't enough information to reach a conclusion, but there's easily enough to ask some very serious questions.

The percentages of lit cyclists involved in accidents are massively, massively greater than most people, especially critics of cyclist behaviour, would allow to be representative of the cycling population in general. You're perfectly right that, if most lit cyclists were keen riders doing 100 times the average night-time mileage, that would distort probabilities strongly in favour of lit riders getting hit. However, in practice it seems to me that most night-time mileage is likely to be done by commuters, and whilst those who frequent this discussion area probably do use lights, commuters as a whole fall somewhere between those here present and the cycling population in general.

In any case, any effect from increased mileage will also affect the chances of your seeing a lit vs an unlit cyclist riding along the street. In other words, if your argument were true, hardly anyone would notice and complain about the unlit cyclists because you'd be so much more likely to encounter lit ones.

I said much earlier in the thread that I wouldn't be getting rid of my lights just yet. I mean it. However, it seems to me that anyone remotely seriously interested in cyclist safety would be pretty keen to see a proper, well-thought out explanation of this very counter-intuitive result. There's been a good debate here, but we're no nearer an answer.

I'm sorry, but there's nothing counterintuitive here because you don't know the relative proportions. If the 10% of unlit accidentees represent, say 50% of all those riding unlit, while the 90% of accidentees represent 1% of all those riding lit then that would demonstrate it's massively safer to ride lit than not. Without these stats it is impossible to draw any conclusion whatsoever.

You need to know the probability of being in an accident riding unlit vs the probability of being in an accident riding lit, factoring out for road types, etc and to do that you need to know how many riders there are in each category.

Auchmill wrote:you need to know how many riders there are in each category.

Supposedly that comes from the size of the impression which unlit cyclists leave in our memories compared to the impression left by lit cyclists.

Traveling 6 miles across Coventry at rush hour last Thursday I distinctly remember seeing an unlit cyclist. I can't remember a lit cyclist, nor a bus, come to that a van. Yet the roads must have been packed with motor vehicles!

Perhaps what these findings really demonstrate is that we remember the unusual (or what has emotional impact - e.g. what generates a feeling of self-righteousness) while the commonplace quickly fades from memory.

Auchmill, sorry, but George Riches is right. I did some thinking about how all the numbers and probabilities added up on this. It got complex. However, in fact, you can forget all the business about percentages of cyclists who use lights and what mileage different groups do. The crux of it is this:

Assume that those observing what percentages of cyclists use lights do so when standing by roads that are typical of those where accidents occur. If 9 out of ten cyclists involved in accidents have lights, then the relative chances of the observers seeing cyclists should be 9:1 lit:unlit.

However, we are assuming that lights make cycling safer. So let's say that lit cyclists are only ten times less likely to be hit on average (if it's much less than that, we're back wondering why we use the things...). To compensate for this, the relative chances of the observers seeing cyclists should be 90:1 lit:unlit.

So why does everyone agree that lots of cyclists have no lights?

Anecdotage: a late relative of our's used to ride a motorcycle and sidecar for many years. He claimed nearby drivers would slow because the lighting arrangement - one main light and a fag end glow on the offside mudguard - disconcerted them. There's something in the notion. Lighting standards are based around establishing predictability, something motorists can assimilate quickly and move on. Cyclists, because of their speed and road position require illumination that defers those judgements as long as possible.

Lack of illumination does that job so long as something reflects enough light to ensure visibility. There are probably other cues that suggest a wide birth - riders with low legal appreciation may be transgressive in other ways.
I'm not surprised the unlit get a lower accident count but the sums may not be like-for-like.

Street lighting levels (if lit ) , movements of cyclists injured (right turn , changing lane etc.) local or through cycists , volume of traffic , adequate reflective markings on clothing etc. etc . all need to be considered. IMHO it's the cyclists who ought to determine if he needs more than legal requirements , many cyclists do feel safe at night as regards lighting in that there is adequate street lighting in urban areas , I suspect it's negotiating junctions that is the problem , where cyclists lighting fails due to distracted drivers . Study needed .

This is my last word on this. There seems to be an anti-science bias here. If we are to make any sense of the world we need scientific observation and analysis. It's no good talking about what we think we see because casual observation is notoriously unreliable. Without a scientific study of the numbers involved etc we have no basis for making any statements about the relative merits and risks of riding lit vs unlit vis a vis the statistics mentioned in the OP. Otherwise we might as well go back the the Muddle Ages when the sun travelled round the earth.

Auchmill, it isn't only pure science that will lead you to a conclusion. Cultural and sociological factors may be at play and there may be psychological influences. Many will know of cases where the club's 'christmas tree' is the same rider who falls foul of drivers. Simple luminance and bald stats may not reveal the whole story.

drossall wrote:So why does everyone agree that lots of cyclists have no lights?

You should be aware that there are a number of psychological processes involved here: Basic perception - seeing, interpretation - recognition, and remembering. Memory, of course, has an editing function.
So, you see a law abiding illuminated cyclist. Well its just the lowest of the low innit? Utterly unremarkable, totally forgettable.
On the other hand, a cyclist without lights, on the pavement, or doing a red light or two. This is the stuff that gripes are made of and we treasure such misdeeds so that when the weather starts to pall as a subject of conversation we have something topical to move on to - "Guess what I saw on the way to the pub..."

CJ wrote:

Ben Lovejoy wrote:

CJ wrote:If we suppose that more than 10% of cyclists were unlit, the inescapable but counter-intuitive conclusion is that it's safer like that!

Only if the mileages travelled are the same, and I would wager a very large sum of money this not the case.

I suggest that almost anyone who commutes a decent distance every day in the dark will do so with lights. Those without lights, in contrast, are more likely to be occasional cyclists making only short journeys. Those with lights are more at risk because they do many more miles on their bikes.

As a cyclist, I am biased in favour of cyclists, and yet even I would not have claimed (in the 1980s when this study was conducted) that such a high proportion of the cyclists I saw had working lights.

As others have said, we remember the unusual (especially if we have an emotional response to it), which is why we get the impression there are lots of unlit cyclists around.

In reality, I doubt I see more than one per journey - yet how many lit cyclists must I see (and forget) each journey? Almost certainly more than nine.

Ben

Yes, but remembering the unusual itself has two consequences:

* the problem of unlit cyclists is insignificant when everyone thinks otherwise

* you may well be safer (ie more noticed) if you don't use lights

Are we ready to conclude either of those things?

We're guessing either way, but I choose to bet my life that I'm safer with lights than without.

Ben