Safe Speed Forums

I'm working on some interesting stuff based on this DfT report:

http://www.dft.gov.uk/stellent/groups/d ... 031459.pdf

It tells us that "excessive speed" is recorded as a contributory factor as follows:

Fatal accidents: 28%
Serious accidents: 18%
Slight accidents: 11%
All accidents: 12%

Now there's something here I'm not quite getting me head around...

Do I believe that there's a genuine differece is contributory factors between fatal accidents and (say) other high severity crashes? Well, no, not really, I don't suppose I do. I think of the risk triangle and I reckon there are two things that really make the difference between the layers - luck and degree of human error.

So does excessive speed make a fatal accident more likely? Or is it really that a fatality - a rare and quite random event - tends to select excessive speed? Now this strikes me as quite mad, but there's something in it that's seriously niggling at me to the extent that I can't actually eliminate the idea. Neither can I quite get a grip on what's causing the doubts.

Help!

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

If you're travelling at 60 mph when you make a mistake which means you hit a solid object or another car then you are more likely to be killed than if you made the same mistake at 30 mph. It is not the speed causing the accident but the consequences of the same mistake being more serious. Plus the fatal accident stats wil contain joyriders and people without seatbelts which seriously alter the probabality of an accident being a fatal. The people causing excessive speed fatalities may also be in the more reckless (or should that be wreckful?!) end of the driving spectrum and will probably be driving around in the upper end of the speed range for a given road and be less aware of their limitations. Excessive speed + poor observation = fatal accident.

That's how I look at it anyway....

Thanks. That's the view I've always held too.

Yet if we take 100 accidents involving inattention of various seriousnesses resulting in 5 deaths (say) what's the betting that the ones with deaths are more likely to be recorded as excessive speed. Maybe they are genuinely excessive speed (more inattention, less braking) but does that REALLY mean that pre incident speeds were higher on average?

Or have we made a selection after the event?

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Perhaps its easier to attribute the incident to excessive speed than to try and quantify a parameter like 'inattention'? How do you determine the degree of inattention? Perhaps we could argue that a driver was inattentive to their speed for the situation?

Thinking aloud...

Or they were so inattentive that they hit the other object at the speed they were travelling at without braking first? Most accident investigators would assume some braking before hand and assume that the speed was greater than it actually was. I have also read that it is now getting harder to judge pre accident speeds as abs is removing skid marks and the crumple zones on cars are working in ways that it is also hard to tell impact speed.

Excessive speed is also relatively simple as it just meant they didn't stop in time and hit the other object. One could argue all accidents are excessive speed in that sense

I think I'm gradually getting a better grip on this thing and there are some real effects.

Firstly, fatal accident have more contributory factors recorded on average. In the recent Dft study they have around 7.5% more contributory factors on average. (2.28 ave for a fatal, 2.12 ave for an injury accident)

Secondly, there really is a post selection effect, as suggested above. But it's probably pretty small. (I'm not entirely comfortable with this one - there's something of the chicken and egg about it).

Thirdly there might be a reporting bias effect, where the shock and stress of dealing with a fatal tends to lead those reporting it to state the causes "more strongly".

Now that 28% figure is in for one hell of a carve up. I'm still looking into it.

But:
* many of those crashes are within the speed limit.
* many involve reckless behaviour, unlicenced drivers, stolen cars and "lawless behaviour"
* young and inexperienced drivers have a far higher risk
* some have pedestrian causes
* many (probably half) have a different primary contribution factor.
* Many have excessive speed as a "possible" factor
* Many have "excessive speed as a "probable" factor.

Even without the carve-up we have 28%*3,500 = 980 "excessive speed" deaths within a population of 32 million drivers annually. That means one caused excessive speed fatality per 32,600 driver years. Yet we're speeding every day. 1 day in 12 million there's an excessive speed fatal.

Speeding looks pretty safe then...

_________________
Paul Smith
Our scrap speed cameras petition got over 28,000 sigs
The Safe Speed campaign demands a return to intelligent road safety

Speeding is safer for some people more than others though isn't it? Someone posted some footage of some bloke driving in rush hour like an idiot. The speed he was using wasn't the scary part, it was the fact that this person wasn't looking ahead very far and would continually switch lanes into closing gaps rather than opening ones. You had his view of the road and you could clearly see a decent set of gaps to drive into but the driver seemed to choose ones right in front of him rather than plotting a course up the road. He drove into several gaps where you could see brake lights coming on and knew they were going to stop right in front of him so he nearly went up the back of the tail car each time.

I think the type of people that only look a couple of cars ahead (sometimes less than that) are the ones to be targetting. I bet a lot of these people don't exceed speed limits but they clearly don't have observation skills that will allow them even to drive at the limit in a lot of conditions anyway. Then there are others that still only look a few cars ahead but ignore the speed limits, they're the reckless ones and are likely to be the type to be represented in fatals more often. The safest drivers will be the ones looking well ahead, all around and either staying within the speed limit or choosing a safespeed without reference to the posted limit.

The safest behaviour is always match speed to vision but at the moment it just seems to be match speed to lollipop which isn't quite the same thing although in some circumstances it might be of benefit where the speed limit and safespeed happen to coincide.

Perhaps because a number of fatal accidents (by the nature of being fatal accidents) leave the driver unable to answer questions, leading to an assumption of excessive speed (e.g. can't pin it down 100% to anything specific, so by including excessive speed all the bases are covered..)? The driver is also not in a position to argue that it wasn't. In Serious Injury accidents there is a chance to get the driver's perspective on the cause of the accident.

Could there also be anything in the high proportion of fatal accidents involving motorcyclists (compared to the overall proportion of vehicles that are motorcycles) - maybe people assume that because it involved a motorcycle it must involve excessive speed? Also if there are witnesses to the accident a motorcycle can sound and appear faster than it actually is (particularly from a pedestrian perspective).

What are the criteria for recording the data? I can quite easily believe that where excessive speed causes a fatality but is not a cause of the accident it may still be recorded as such.
Does that make sense? More speed = more damage in the event of a collision but does not necessarily cause the collision. The data recorder however, gets one step ahead and because speed was contributory to the severity, includes it as causal.

Just as a matter of interest (if you have the time and inclination), if you regroup all illegal act causal factors (D&D, stolen cars, unlicensed ect) what % do they form as contributory to fatals.

Perhaps it is this that is bugging you, excessive speed is actually 2 factors that have been grouped (too fast for the conditions and exceding the speed limit ( I'd bet if the speed limit was exceeded it is an automatic inclusion even if it was not an excessive speed )). Whereas other factors are split from a common group.