Safe Speed Forums

There's plenty of news about this today. Here's the Direct Line PR:

(apologies for formatting - I really don't have time to clean it up.)

> 94% Motorists Admit to Breaking the Speed Limits

> Speeding is one of the biggest dangers on our nation's roads with
> one in three road deaths being attributed to speed (1000 people every
> year*) and injuring thousands more each year.
> Yet in a recent survey of more than 1,500 UK drivers by Direct Line
> car insurance, 94 per cent of motorists still admitted to speeding, with a
> third (33 per cent) of young drivers (18-29 year-olds) admitting to doing
> so regularly.
> Motorists questioned gave a number of reasons for speeding, but
> close to six in 10 (57 per cent) drivers think it is okay to speed on
> empty roads and a quarter (24 per cent) think it's acceptable to speed
> when weather conditions are good.
> Drivers also think it's acceptable to speed when:
> * it's an emergency (33 per cent)
> * they think the limit is set too low (24 per cent)
> * there's less traffic around, i.e., at night (21 per cent)
> The research also revealed a surprisingly lack of knowledge about
> speed limits - despite motorists generally thinking they are experts.
> Almost all (93 per cent) of motorists said they have a sound knowledge of
> speed limits, when in relatively half of UK motorists (50 per cent) are
> unable to correctly identify the limit on a dual carriageway (70mph).
> It is worth drivers knowing the speed limits as the fines incurred
> if motorists do accidentally speed can be significant.
> For example, if motorists are caught travelling 15mph over the speed
> limit on a motorway they could end up with a £2500 fine. Motorists caught
> doing over 100mph on a motorway will find themselves with an automatic
> ban.
> Drivers in North Scotland are the worst at flouting speed limits,
> with one in three (31 per cent) doing so regularly.
> In contrast, motorists in the South West are the most likely to obey
> speed limits with just one in 10 (11 per cent) regularly exceeding the
> limit.
> Direct Line's Motor spokeswoman, Emma Holyer, said:
> "The lack of knowledge amongst motorists is very surprising - particularly
> that so many don't even know the speed limits on some UK roads. Not only
> is speeding dangerous but as many motorists know to their expense speed
> cameras exist across the UK and the points on your licence and fine
> incurred can be hefty.
> "In today's high powered cars, it's easy to lose concentration and go over
> the limits and it may be that many motorists are speeding without
> realising it. This is further backed up by research we conducted with
> Brake, the road safety charity, that found that 96% of drivers believe
> that speeding outside a school is very dangerous with a further 72%
> believing that speeding in a 30 zone was dangerous in itself - so it is in
> your best interests to keep an eye on your speedometer and keeping speeds
> down."
> Further information on Direct Line go to www.directline.com, or call 0845
> 246 8888.
> -ends-
> Notes to Editors:
> The research was carried out by You Gov from March 18th-21 March
> 2005. A total of 2,059 UK adults aged 18 and above were questioned.
>
> * Source: Department for Transport, September 30, 2004.
> ** There are 33.8 million drivers in the UK (Census data, 2001) and
> according to the You Gov findings one
> per cent of motorists have had an accident as a result of a driving
> too fast or over the speed limit 0.01 x 33,809,493 = 338,094
>
> *** At 35mph a driver is twice as likely to kill someone than at
> 30mph - ROSPA inappropriate speed policy statements - May 2002
>
> Direct Line Insurance plc, authorised and regulated by the Financial
> Services Authority. Calls may be recorded. Conditions apply.
> NATIONAL DATA
> Question % of motorists who answered the q INCORRECTLY Correct
> answer
> What is the national speed limit for a car, in miles 50 per cent 70
> mph
> per hour, on a dual carriageway?
> How big a gap should you leave between you and 68 per cent Two
> seconds
> the vehicle in front in normal conditions?
> How big a gap should you leave between you and 82 per cent Four
> seconds
> the vehicle in front in WET weather?
> What is the maximum financial penalty for 99 per cent £2,500
> travelling 15mph above the speed limit on the motorway in a car?
>
>
> REGIONAL SPEEDERS
> Region Percentage of motorists who regularly drive over the speed
> limit
> East Anglia 20 per cent
> Wales 16 per cent
> South West 11 per cent
> South 23 per cent
> Lancashire 15 per cent
> North East 18 per cent
> Yorkshire 16 per cent
> Central Scotland 19 per cent
> North Scotland 31 per cent

Safe Speed issued the following PR at 12:13 Thursday:

PR226: We're still speeding

News: strict embargo: 00:01 hours on Friday 5th August 2005

A survey by Direct Line released today reveals that 94% of motorists admit to
breaking speed limits.

Paul Smith, founder of the Safe Speed road safety campaign said: "After 12
years of speed cameras drivers' speeding behaviour is unchanged. Far from
being a problem, drivers selecting an appropriate speed according to the
conditions is of fundamental importance to road safety. Sometimes they set
safe and appropriate speeds above the speed limit - to do so isn't reckless,
careless or irresponsible - they are just fulfilling their fundamental duty to
road safety by setting a speed that's safe and appropriate for the
conditions."

"We have speeding magistrates, speeding off duty senior police officers,
speeding senior politicians and speeding royals. To suggest that exceeding a
speed limit in itself is irresponsible is completely absurd - the competent
and careful actions of a majority of responsible people should obviously be
considered legal."

"Speed limit enforcement is way out of line with practical reality - far from
blaming the public, we must blame the government for obsessive and
destructive enforcement of a arbitrary law."

"I blame cameras - we're obsessed with the wrong safety factor"

<ends>

Notes for editors:
==================


Official figures indicate that speeding behaviour is unchanged after 12 years
of cameras.

12 million motorists have been fined by camera.

Fines issued by speed camera total over £700 million (not £700,000 as
incorrectly noted in our last PR - apologies for the typo)

Only about 1% of crashes involve an otherwise responsible motorists in excess
of a speed limit, yet 60% of motorists are speeding at sample sites in free
flowing conditions.

Overall road safety results in the last decade are the WORST ON RECORD in
terms of improvement. Safe Speed believes that this is due to bad road safety
policy.

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

I did wonder how many people think that the speed on a d/c is 60mph, this is quite interesting, I often pass people doing 60 and wonder whether they are just taking it easy or don't know that they can do 70.


What a stupid question, it's like asking people the length of the Forth Bridge in feet and marking it as wrong unless it is precisely correct. I didn't know the maximum was £2,500 but I knew it was in the thousands region. I doubt however that no more than a handful of people have been fined this much, in fact I have never heard of a fine of more than £1,000 being handed down for *any* offence, and the ones near £1000 were when they were doing 110+ and got away without a ban.



B*ll*cks, they should read Pepipoo and Pistonheads for plenty of examples of people getting points and a fine for quite going quite a bit over 100mph.

Gareth

From Bayes' Theorem, we can derive a formula for the relative likelihood of speeding and non-speeding motorists having an accident. The resulting formula is:

P(accident given speeding) / P(accident given not speeding)
= P(speeding given accident) x P(not speeding) / ( P(speeding) x P(not speeding given accident) )

Paul's figures above give us the numbers to plug in, and the ratio works out as 0.01 x 0.4 / (0.6 x .99) = 0.0067.

What this means is that the 40% of motorists who are not speeding in free flowing traffic are 148 times MORE likely to be involved in an accident than the 60% of motorists who are speeding in free flowing traffic. This suggests to me that the most productive area of research would be to establish why these motorists are so much more likely to have an accident.

<Pedantic Grammar Checker On>

So, if they are bad at flouting speed limits then they actually do adhere to them well then?

I'll be sure to look at the dash instead of the roads and pavements next time I drive past a school then.

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

I see they are still peddling the 1/3 lie even after the DfT have quietly dropped it

Nice, but of course the populations are different. Speeding takes place where crashes are unlikely. It's Safe Speed behaviour at work in the real world saving lives. Folk slow down 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

Depends what is meant by "Only about 1% of crashes involve an otherwise responsible motorists in excess of a speed limit". If the point at which the speed is considered is at the point of impact, I agree. If it means prior to becoming aware of the hazard that led to the accident, I stand by my calculation. Either way, we're in agreement that drivers who speed in free flowing traffic are hugely under-represented in accident statistics, whereas if their behaviour were dangerous in itself, they should be over-represented.

Can well see the 94% being an accurate figure - which means accidents are not happening because of speed alone. If this were the case - accident rate would be a lot higher.

Suspect the 94% relaise they drift just over on any one journey and correct it and adjust in response to hazards.

As stated umptten times - real cops tend to go for truly dangerous and risk drivers - and base their judgement on professional training. We would not be able to pull every driver who drifted marginally above a speed limit - and so long as it is a safe and marginal drift - we tend to use discretion! (cough)

_________________
Take with a chuckle or a grain of salt
Drive without COAST and it's all your own fault!

A SMILE is a curve that sets everything straight (P Diller).

A Smiley Per post
FINES USfor our COAST!

Approach love and cooking with reckless abandon - but driving with a smile and a COAST calm mind.

The speed considered is the pre-incident speed. It is desireable to determine if the pre-incident speed contributed to the outcome.

But we're still looking at disparate populations. It's too oversimplified to neglect hazards, which is what you're doing (in effect).

Suppose we had a 1 mile perfect straight followed by a tight but well signed 25mph bend.

I don't believe there would be any relationship between the vehicles that attained high speed on the straight and the vehicles that crashed on the bend. Some folk who were fast on the straight would be so because they knew it was safe. Some folk who went fast on the straight would be reckless fools.

Crashing on the bend calls for different sorts of error - failure to observe or judge, perhaps.

Different populations entirely.

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

I still think I'm right
Take the scenario you describe. Suppose we have permanent monitoring of speeds along the road, so we can always find out what speed someone was doing on the straight bit when they crash on the bend. Suppose that only 1% of the people who crash on the bend were speeding on the preceding straight, but 60% of all vehicles speed on the preceding straight. I can't see how you could draw any conclusion other than that the people who don't speed on the preceding straight are 148 times more likely to crash on the bend than those who speed on the preceding straight (using my previous calculation). This would almost certainly imply, IMO, a causal link in terms of driver competence, with the same competence that causes a driver to feel comfortable on the straight at higher speed causing them to be more likely to safely negotiate the bend. Of course there will be some reckless fools, otherwise it would be 0% not 1%, but the general inference would be correct.

No, this is nonsense as the speed-related crashes do not typically occur in free-flowing, clear conditions.

The fact that someone exceeds the speed limit when relatively safe to do so bears no relation to whether he is likely to use inappropriate and illegal speed in hazardous situations.

_________________
"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

So how do you explain the enormous statistical correlation that shows drivers who don't speed in free flowing traffic being massively more likely to have an accident? It is true that correlation is not causation, but if it isn't causation, it means there must be some other factor causing the link between the two, and I honestly can't think of any other plausible factor than it being driver competence that is linking the two things. So I would say that the statistics show there is a relationship between people exceeding the speed limit and people using inappropriate speed in hazardous conditions, just the opposite one to that which the "speed kills" brigade would like us to believe.

Excellent thread, this one, many thanks to stevei for some very interesting contributions... and a dollop of statistical calculation which helps provide a framework for the discussion.
Personally I agree with the message that stevei is bringing, especially in the post quoted above.
Don't ask me to justify this using figures, as this is more a gut feeling based on many years of living and driving, but the implications are pretty mind blowingly positive for the anti-"speed kills" lobby.

When I think of a SAFE (= low risk of accident involvement) driver, I think of words like SKILLED, CONFIDENT, ALERT TO ROAD ENVIRONMENT and FREQUENTLY OBSERVERVED EXCEEDING THE SPEED LIMIT (ALWAYS GRACEFULLY).

When I think of a LESS-SAFE (= higher risk of accident involvement) driver, I think of words like ERRATIC, UNCONFIDENT, LESS ALERT TO ROAD ENVIRONMENT and INFREQUENTLY OBSERVERVED EXCEEDING THE SPEED LIMIT (GRACEFULLY OR UNGRACEFULLY).

So, yes, in my opinion, the safer drivers tend to exceed the speed limit more frequently.

I want to add that I now see PeterE's and Paul's point about the sampled populations. The issue is that when the free flowing traffic is sampled, the sample is taken during "representative" conditions. When we sample from accident events, there may be a selection taking place, as conditions in place when accidents occur may not be representative of typical conditions. What would be needed is to sample free flowing traffic in the same conditions as when the accident occurred, so that the same selection is applied to both sets of data.

Paul - it's arguably misleading for you to word your sentence to link these two statistics if this matching of samples did not take place.

That may be in all practical senses impossible because 'the same conditions' a) always result in a crash or b) those conditions directly influence the speed of traffic in an unpredictable way. (think of crashes at a cross roads, where the speed of the driver with priority is to be considered).

[I may not have fully unravelled this problem yet... most interesting.]



My purpose in quoting the figures together was to illustrate the dichotomy in a pair of figures that the authorities would have us assume belong together. Certainly my purpose wasn't to mislead - but you were misled and you're numerate, so I really don't know. At least we're developing better understanding because of it.

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

There's some interesting detail in the large document from your FoI request about roadwork accidents. There, they had video cameras monitoring the roadworks, and whenever an accident occurred, they worked out the speed of the vehicles involved. Clearly, they could have also worked out the speeds of all vehicles on the road around that time, and this probably would have been "good enough" input data for my calculations. The specifics of the accidents also make interesting reading - I believe that the majority of the accidents were not due to some freak set of conditions, just simple mistakes on the part of drivers, and hence it would be reasonable to sample the traffic speeds for a 5 minute period, say, around the time of the accident, and use this in my calculations. One could also obtain useful data from the tapes in terms of how drivers reacted to the gradual emergence of hazards by modifying their speed over time.

But in so doing you might miss noticing that similar hazard conditions existed 10,000 times without a crash resulting... It really is quite tricky.

Crashes are rare and exceptional events, yet crash preconditions are absolutely everywhere. One huge mistake that the authorities make when examining crashes is to 'forget' to consider exactly how similar preconditions frequently DON'T result in a crash.

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

And also just how identical conditions have to be at the time, like, you either have a vehicle cutting across your path at the time you're 40' away, or you don't. A single second either way, or a few metres either way, can make all the difference in the world.

Cheers
Peter

_________________
Only when ideology, prejudice and dogma are set aside does the truth emerge - Kepler

I can see I'm struggling to explain this, but what you have described above isn't relevant to answer the question I want to answer. What I want to find out is whether people who exceed the speed limit in the same conditions as the accident occurred are more or less likely to have an accident. The "same conditions" are those prevailing on the road at the time of the accident, not the actual accident conditions.

The example of a long straight followed by a tricky bend is a good one, and I hope you can easily see that in this case, if there were a correlation between speed along the straight and likelihood of skidding off the road on the bend, then this would most likely be a meaningful relationship.

For types of accident like the one you're talking about, where a vehicle goes across the path of other vehicles, the same logic still applies. If the vehicles that actually hit such a vehicle, counted over a large number of samples of such accidents, tend to one end of the distribution of speeds in the prevailing conditions at the time, we can draw conclusions about the risk incurred by people travelling at that speed, in those conditions, even when no accident occurs. If the vehicle that hits the one going across the path of traffic is generally travelling in lower percentiles of the speed distribution, or higher percentiles of the speed distribution, that would give us meaningful information.