More statistical nonsense

The problem with the idea of accident statistics at a single camera sit is that you can play silly buggers with the statistics as you choose. For example a number of cameras were placed along a stretch of road near me because the number of KSI incidents on that road was huge. The authorities had tried dropping the speed limit to no avail. The road is a classic of it's type. It is a county A road with lots of trees and stone walls nearby, it's in the hills so forward visibility is bad in places and it has quite a few very minor roads entering onto it and at some of those junctions visibility isn't great. It's not actually dangerous if traffic is sticking to the speed limit, but the visibility is not sufficient when the traffic is doing 70mph or more.

For those who don't get the reference about trees and walls it should be obvious that leaving the road when there is ample runoff is a lot safer than hitting a tree. In the first instance you have a minor accident with some property damage, in the latter a KSI.

The cameras on this road are mostly well placed. On the approach to blind bends, crests and some of those junctions I mentioned. There are only two with which I would disagree and they are both on straight stretches of road where forward visibility is good. However it just so happens that there have been as many KSIs on those two stretches as any where else on the road. The accidents in question were apparently down to dodgy overtaking manoevres. Probably drivers who have had to wait behind a slow moving vehicle waiting for the straight and then misjudging their maneovre.

Now you could argue that statistics would always support the placement of those speed cameras. That RTIs are by their very nature freak events and as such a KSI at a given site is unlikely to be repeated within the next few years. Therefore you might argue that statistics are likely to show that there are been no accidents at the camera site since it was placed there thus supporting the placement of speed cameras. Somebody will probably use the TLA RTM to talk about that phenomen in this very thread. The person doing so would however be confusing numbers with statistics.

If you want statistics you need to look at events over a number of years and along the entire stretch of the road. Picking on one camera site one local campaigner argued against the placement of the camera by saying there was one fatality there in 2007 and there were none before and there have been none since. Thus by his argument the accident was a freak event. His argument does not for a moment discuss whether that accident could have been avoided. That does not however take into account that there have also been several KSI incidents at that same site in the run up to the installation of the camera and there have been none since. So there are arguments both for and against that particular camera using the same numbers.

Where the statistics can prove to be very misleading is where other criteria are not taken into account. Take a camera a few miles away when KSI have reduced significantly since the camera was installed. At about the same time the camera was installed the speed limit was reduced and traffic calming measures were introduced. The official statistics would appear to support the installation of the camera, but the statistics make no mention of the reduction in speed limit or the other road safety measures. Because those measures were not given time before the camera was installed the statistics are useless because nobody can prove which of the three cameras lead to the reduction in KSI incidents.

One other problem with statistics is that modern cars are, by and large, incredibly safe. Things like ABS, EBA and EBD make accident avoidance more likely. And even if you fail to avoid the accident secondary safety measures mean you are much more likely to avoid serious injury than you were eveb twenty years ago. So what? Well there are no available statistics to show how many collisions have occured at a given site that did not result in death or serious injury. There are definitely no statistics to show how many near missed there have been at a site.

For an incident to move from being a mere bump to a KSI doesn't necessarilly take all that much. An overweight occupant in one of the vehicles (yes really, especially in smaller cars) can make the difference between bruises and serious injuries or worse. That one of the cars is older or poorly maintained or repaired (ever seen a dodgy cut and shut after a collision?) could turn an every day shunt into a fatality. The statistics also don't take that into account.

Only an idiot would deny that a given collision is more likely to cause injury or death if (all else being equal) the speeds involved were higher. However it does not follow therefore that if an accident did cause death or serious injury the speeds must have been too high. The statistics do not record whether the injured party fell outside the car designers criteria for the average person when they designed the safety systems. They do not show whether one of the vehicles was poorly maintained. They don't even show if somebody had fallen asleep at the wheel. They don't even tell us if any of the vehicles involved was exceeding the speed limit - if you are trying to use the statistics to support static speed cameras it helps greatly if you assume that they do.

All of which is why static speed cameras are difficult to support (or indeed argue against) successfully with statistics alone. If, however, you take a stretch of road a few miles long with a high rate of KSI incidents and install average speed cameras then the statistics might just be useful. However installing that sort of system would cost a lot more than throwing in a handful of static speed cameras - so for the foreseeable that practice will continue and weak statistics will be used to support it.

There are no secret speed cameras...

Indeed every "safety partnership" is required to publish the locations of all their fixed cameras and where temporary cameras are being used. You'll usually find them on t'interweb.

Not only that, but any of those stories you hear about drivers getting nicked by hidden cameras are just so much BS. I've heard all sorts of tales about hidden cameras, and people being successfully prosecuted by their use. However there are strict rules governing the placement and warnings of both temporary and permanent cameras and were you to get pinched by a camera that didn't comply it wouldn't be too hard to get the conviction overturned.

Those stories are usually just drivers making up excuses for being daft enough to get nicked by a bi bright yellow box or a van with a big speed camera sign painted on it. In over half a million miles of driving I've been pulled for speeding twice. Once I simply didn't spot the patrol car with the laser and he got me. The other I didn't spot the patrol car coming up the slip road as I passed, no I don't think for a second he was being sneaking he was just coming up the slip road and saw me pass the end of the slip road and gave chase. In the first instance I was not paying enough attention and as such should certainly have not have been travelling over the limit. In the second instance I was unlucky. However I am mature enough to accept that however the pull was made I was bang to rights. The law is simple - you adhere to the posted speed limit. If you don't want to get nicked, don't exceed the speed limit.

The only grey area in that respect is where the speed limit isn't clear. And I can't think of many cases like that, reminder signs are the norm on roads with a limit below the national limit. The only exception to this being in built up streetlit areas where the limit shall be assumed to be 30mph unless signs indicate otherwise. Again if you got nicked where there wasn't a clearly posted limit you should have little trouble avoiding conviction.

I've never done it with speeding, but have on two occasions avoided conviction by presenting photographic evidence that in one case a road sign was obsured by trees and in another that road markings were missing. You only need to employ "Mr Loophole" if you have more money than sense. The guy is famous for getting people off speeding convictions, but it tends to be less well docuemented when he fails. Sometimes you just can't get away with if you're in the wrong.