Safe Speed Forums

Gentlemen,

Can we please examine this faintly ridiculous argument in some detail:

Since around 60% of vehicles are speeding under free flowing conditions at
sample sites, yet even the wildest DfT claims suggest that 30% of crashes involve excessive speed, can we not conclude that speeding is safer than not speeding?

Of course I quite understand the reality that squares this circle - crashes take place in areas of danger where the vast majority have moderated their speed, but hey, that's the point isn't it?

What do you say?

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

Would it not be simpler merely to state that there is a clear statistical correlation between the numbers of drivers who routinely speed and the number of drivers who aren't routinely involved in accidents...?

_________________
CSCP Latin for beginners...
Ticketo ergo sum : I scam therefore I am!

It could be that those of us creeping over the limit are basically paying attention to the road where it matters.

The terms are important I think. As we know, speeding does not necessarily equal excessive speed and vice versa. It's the old 29mph on ice scenario isn't it?

Looking at it another way, we could perhaps say that there are really only three speeds - too slow, safe and too fast. I suspect that the 60% of vehicles speeding are simply those with drivers who have learnt to be good at working out what is too fast on a second by second basis, and constantly adjusting to stay below that. On the other hand, the 30% ( yeah, right ) of crashes involving excessive speed are the reponsibility of drivers who are poor at working out what too fast is, or possibly know very well that it's too fast but are risk junkies (I'm thinking of car thieves etc there). They frequently drive too fast for the situation and it's likely to go pear shaped on them sooner or later. Even though they may well have a crash for an entirely different reason, say failing to give way, it's quite likely that they will also have been driving too fast at the time which will justify it as a crash involving excess speed as well.

_________________
Make everything as simple as possible, but not simpler - Einstein

Those who routinely and safely speed are more likely to be practiced at identifying the safest speed for the conditions. Those who rigidly obey the speed limit are more likely to fall into the trap of "you will be safe if you obey the limit" and so be more likely to fail to recognise when the speed limit is too fast (just a hypothesis with absolutely no proof)

_________________
Will

I concur with Will. I worry about the dumbing down aspect as much as any other.

Surely a key factor is that many crashes happen at junctions or at roundabouts where people will already be travelling relatively slowly.

Relatively few crashes happen to vehicles travelling in free-flowing conditions between junctions.

Obviously this underlines the point that observation at potentially hazardous locations is the key to avoiding crashes.

_________________
"Show me someone who says that they have never exceeded a speed limit, and I'll show you a liar, or a menace." (Austin Williams - Director, Transport Research Group)

Any views expressed in this post are personal opinions and may not represent the views of Safe Speed

I think that most of those probably involve a junction as well if speed camera locations are anything to go by.

Does not mention that though.

_________________
Speed camera policy Kills

Yep. While on the other hand speed camera effects tend to accrue to free flowing sections. Hello Ratty. My aren't you smelly today? (No Peter, I don't mean you! )

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

I say "I'm stupid. I don't understant what you're asking"

Actually that's surprisingly useful because it's forced me to turn the idea over in my head a few more times.

We've got two pieces of data that appear to have some relationship with one another.

The first is the proportion of vehicles speeding at sample sites in free flowing conditions. The DfT publishes an annual report based on speed surveys. As a broad generalisation, they find that around 60% of vehicles are speeding at sample sites.

The second is analysis of crash records. We generally find that about 11% of crashes have 'excessive speed' recorded as a crash causation factor. But we also know that (in sample surveys) 70% of excessive speed crashes are inappropriate speed within the speed limit with 30% of these are in excess of the speed limit. 11%*30% is 3.3%, so our best estimate is that 3.3% of crashes involve exceeding a speed limit as a contributory factor.

Now if 'speeding is dangerous' we'd expect to find that the proportion of crashes involving speeding would be greater than the proportion of vehicles speeding. But it isn't. In fact the exact opposite is true - speeding is commonplace and crashes involving speeding are rare.

Putting this in numerical terms, IF 60% of vehicles are speeding at sample sites AND speeding is dangerous THEN we would guess that the proportion of crashes involving speeding should be greater than 60%.

But in reality, crashes involving speeding are about 20 times less common than speeding itself. (60%/3.3%)

Of course we (Safe Speed folk) know the reason for the apparent 'inversion' - drivers slow in areas of danger, and crashes are more common where speeds are lower.

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

I do indeed think you have something that can be used as a good "soundbite/discussion-point", but it needs a little bit more work first:

Note, this is at sample sites only. These same drivers who are currently exceeding the speed limit also spend a portion of their driving time NOT exceeding the speed limit.


... and this is road-network-wide, i.e.not only relating to the sample sites where you quoted the 60% figure.

I think what is needed are figures that relate to identified stretches of road, average speed figures for each vehicle measured, and number of incidents relating to those vehicles.

Then a statement of the form "under comparable conditions (state road, conditions, time etc.):
- the risk of an incident whilst exceeding the speed limit was found to be X%
- the risk of an incident whilst NOT exceeding the speed limit was found to be Y% "
could be made.



...at those sample sites, yes.

Just want to make the argument watertight, I do think you're onto something here. Getting all the required data may be difficult though, IMO.

I was thinking of moving in a rather different direction. I was more interested in establishing the idea that speed enforcement is applied where there MUST BE a low risk of crashing through speeding.

Your point is most interesting also, but I simply don't see how to begin to get the data required.

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

Me too. I think that drivers are becoming generally less capable of recognising situations where a fair amount of speed can be used safely, but what is even worse is their declining ability to recognise when it is time to get the speed down - so they drive at 45 mph everywhere!

Best wishes all,
Dave.

Is there any available info on typical speeds in accident blackspots like the data on freeflowing traffic. Comparison of these two datasets could be quite interesting.

We might just find that far fewer than 60% exceed the speed limit in such danger areas

I'm absolutely certain that we would find exactly that from examination of crash severity data. For example in built up areas 0.4% of child pedestrians were killed in impacts - 99.6% survived. Yet we know that if we hit them at 30mph at least 20% would have died.

I can't build a model that explains this low average crash severity from braking before impact. We must also have slowing in areas of danger.

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

Paul, thanks for that, here's my take.

A couple of days ago I was surfing the web at work looking for some data (I'm going somewhere with this, trust me) when I accidentally clicked on a pop-up window, rather than the close gadget. I was taken to an IQ testing website, and started to have a go at the test. here was one of the questions....

Some wicks are snicks. Some snicks are slicks. Therefore, some wicks are definately slicks.
Is this statement
a. Definately true
b. Definately false
c. Indeterminable

The answer is of course indeterminable, we simple do not have enough information to draw a definate conclusion from two apparently linked statements.

The same is true here, even though we have the benefit of some numerical quantification of the two data sources being analysed. There is an attractive element of comparability but thats about where it ends.
Supposing the number of drivers exceeding the speed limit rose to 100%. Would we logically or practically expect the number of speed related deaths to rise or fall? How does the 'speeding is safer than not speeding' suggestion weigh up?

I'd guess that the in vast majority of impacts at least some braking would occur and also that the driver will have at least some time to react. Any speed that can be dissipated before impact is going to improve the situation dramatically.

A few quick calculations (based on stopping distances) indicates that it would take 0.68sec to slow from 30mph to 20mph and 2sec to reach standstill from 30mph (this doesnt include thinking time). If we add 0.5sec for reaction/thinking time this puts it at 1.2sec to slow from 30 to 20mph (and taking about 14-15m to do so).

How quickly do hazards materialise?

Any corrections/comments on the figures welcome.

There are some models and discussion hereabouts. See:

http://www.safespeed.org.uk/forum/viewtopic.php?t=1571
http://www.safespeed.org.uk/forum/viewtopic.php?t=1667

We also need to consider the means of estimation of impacts:

http://www.safespeed.org.uk/forum/viewtopic.php?t=1587
http://www.safespeed.org.uk/forum/viewtopic.php?t=1575

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

We're not carrying out an exercise in 'logic', rather we're looking at crude means of estimation. Even if we were out by a factor of 5 we'd still have powerful evidence that the 'speed kills' arguments are false.

After all this means of estimation suggests that speeding is 20 times safer than not speeding. That's a HUGE result - so huge that I'd suggest that it must be true. It scuppers the idea that speeding (as conducted by normal responsible motorists) is dangerous. And we haven't yet even tried to pull out those horrible high speed crashes caused by reckless behaviours.

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