Speed Limit Sign Errors

Blitzkrieger

silverside wrote:

Were the roads not vastly safer immediately after penalty points were introduced?

Vastly safer is an overstatement. Accidents went down but have since gone up again.

silverside wrote:

If everyone stuck to the speed limit all the time, do you not think that road deaths and injuries would be dramatically lower?

No.

silverside wrote:

There are other real issues, but speed, <snip> and drink account for a huge proportion of accidents.

And no. Ask any police force in the world and the official line is that speed and drink driving cause very little accidents. The <snip> is where you mentioned reckless driving. Like, well duh

silverside wrote:

The other issues (bad roads, tiredness, and stupidity) are harder to crack down on.

But this is where the real gains in road safety would be made. It's hard, so the goverment won't do it. What's more dangerous - a guy doing 70mph in a 60mph zone on a quarter mile straight stretch of dual carraigeway, or some sleepy muppet in a huge blue van trying to hurry to work cutting corners? Guess which guy I was this morning and guess which guy almost hit me. You always see gardai doing speed checks on that straight but do you think they would have bothered to stop your man for dangerous driving even if they saw him?

[aside]I think we all did statistics and probability[/aside]

unkel

Blitzkrieger wrote:

What's more dangerous - a guy doing 70mph in a 60mph zone on a quarter mile straight stretch of dual carraigeway, or some sleepy muppet in a huge blue van trying to hurry to work cutting corners?

Agree with you there, Blitzkrieger. Speed checks should only be carried out with improving safety in mind and never with collecting cash / distributing penalty points as the main motivator

silverside wrote:

I am a trained mathematician so could give chapter&verse on this

Alun wrote:

I'm not sure what exactly constitutes a "trained mathematician", but I studied probability and statistics at University too, so fire away, I'm all ears

Let's have it on then

Seriously, I'd be very interested to see some views on this. When quoting data, please provide source (link if possible)

silverside

Ok, briefly:
assuming road deaths follow a Poisson distribution (happen as one-offs, road death in Dublin doesn't affect the chance of a road death in Donegal). Now assume the 'normal level' of road deaths per month is given by X. Then the standard deviation of the number of deaths per month is given by root X. Similarly if we look at a longer period such as a year, or shorter like a day.

My hypothesis is that the introduction of penalty points caused a shift in peoples behaviour so that the level of road deaths in the 4 months following was significantly lower than it would otherwise have been. Alun's position is that there is no discernible difference (forgive me if I am putting words in your mouth).

I will consider a six-month period (4 months would show an even stronger effect), i.e. November through April, for each of the 5 winters where we have data (from http://www.boards.ie/vbulletin/showthread.php?p=1854719#post1854719 - Garda data).

The death counts are
192 189 211 141(*) 175
with a mean of 182. This would give a standard deviation of 13.

We see that the 2002-2003 period is 3 standard deviations below the mean. Using a one-sided T test with 4 degrees of freedom, this gives a P-value of <0.05. This means that, in less than 5% of random fluctuations from the mean, would there be a value this low.

One can never rule out chance, but I think a better explanation is that people drove more carefully for those months. I don't think I can be accused of picking figures to suit my argument. I'd like to hear your reply.

Seth Brundle

silverside wrote:

Ok, briefly:
assuming road deaths follow a Poisson distribution (happen as one-offs, road death in Dublin doesn't affect the chance of a road death in Donegal). Now assume the 'normal level' of road deaths per month is given by X. Then the standard deviation of the number of deaths per month is given by root X. Similarly if we look at a longer period such as a year, or shorter like a day.

My hypothesis is that the introduction of penalty points caused a shift in peoples behaviour so that the level of road deaths in the 4 months following was significantly lower than it would otherwise have been. Alun's position is that there is no discernible difference (forgive me if I am putting words in your mouth).

I will consider a six-month period (4 months would show an even stronger effect), i.e. November through April, for each of the 5 winters where we have data (from http://www.boards.ie/vbulletin/showthread.php?p=1854719#post1854719 - Garda data).

The death counts are
192 189 211 141(*) 175
with a mean of 182. This would give a standard deviation of 13.

We see that the 2002-2003 period is 3 standard deviations below the mean. Using a one-sided T test with 4 degrees of freedom, this gives a P-value of <0.05. This means that, in less than 5% of random fluctuations from the mean, would there be a value this low.

One can never rule out chance, but I think a better explanation is that people drove more carefully for those months. I don't think I can be accused of picking figures to suit my argument. I'd like to hear your reply.

zzzzzzzzzzz

silverside

kbannon wrote:

zzzzzzzzzzz

I know, but unkel and alun asked me to.

Seth Brundle

...and if they asked you to jump off a cliff...

silverside

I don't usually walk away from arguments. It has got me into trouble in the past.

I should have known better than to hop onto the motors board though

Blitzkrieger

We're not disputing that it made a difference with you, Silverside It's how much of a difference can be attributed to the penalty points and it's long term effects that we doubt.

BrianD3

I drove along that stretch of road today and counted eight 50 mph signs on the inbound carriageway between the end of the M4 motorway and Woodies DIY. There's a similar number of signs on the outbound carriageway. So it's crystal clear. The limit is 50 mph. A 40 mph limit is unenforcable on this stretch. Nobody will have gotten penalty points for doing <50 mph on this stretch. The Gardai are not "rigorously enforcng" a 40 mph limit.

Basically the newspaper that kbannon mentioned has f*cked up with the story and someone should receive a good bollocking for this scaremongering. You'd think the incompetent ****wits would reserach their story before splashing it all over their front page :mad:

BrianD3

Alun

silverside wrote:

Ok, briefly:
assuming road deaths follow a Poisson distribution (happen as one-offs, road death in Dublin doesn't affect the chance of a road death in Donegal). Now assume the 'normal level' of road deaths per month is given by X. Then the standard deviation of the number of deaths per month is given by root X. Similarly if we look at a longer period such as a year, or shorter like a day.

My hypothesis is that the introduction of penalty points caused a shift in peoples behaviour so that the level of road deaths in the 4 months following was significantly lower than it would otherwise have been. Alun's position is that there is no discernible difference (forgive me if I am putting words in your mouth).

I will consider a six-month period (4 months would show an even stronger effect), i.e. November through April, for each of the 5 winters where we have data (from http://www.boards.ie/vbulletin/showthread.php?p=1854719#post1854719 - Garda data).

The death counts are
192 189 211 141(*) 175
with a mean of 182. This would give a standard deviation of 13.

We see that the 2002-2003 period is 3 standard deviations below the mean. Using a one-sided T test with 4 degrees of freedom, this gives a P-value of <0.05. This means that, in less than 5% of random fluctuations from the mean, would there be a value this low.

One can never rule out chance, but I think a better explanation is that people drove more carefully for those months. I don't think I can be accused of picking figures to suit my argument. I'd like to hear your reply.

Repeat after me :

Demonstrating correlation does not prove causality.

Alun

Blitzkrieger wrote:

We're not disputing that it made a difference with you, Silverside It's how much of a difference can be attributed to the penalty points and it's long term effects that we doubt.

Exactly my point. You can demonstrate all you like that there was a significant drop in accident (or more accurately, fatality) figures during that period, but he failed to demonstrate any causality between the introduction of penalty points and that decrease. In fact, short of performing a series of controlled experiments, there is no way of proving causality ... any volunteers?

Seth Brundle

BrianD3 wrote:

I drove along that stretch of road today and counted eight 50 mph signs on the inbound carriageway between the end of the M4 motorway and Woodies DIY. There's a similar number of signs on the outbound carriageway. So it's crystal clear. The limit is 50 mph. A 40 mph limit is unenforcable on this stretch. Nobody will have gotten penalty points for doing <50 mph on this stretch. The Gardai are not "rigorously enforcng" a 40 mph limit.

Basically the newspaper that kbannon mentioned has f*cked up with the story and someone should receive a good bollocking for this scaremongering. You'd think the incompetent ****wits would reserach their story before splashing it all over their front page :mad:

BrianD3

From the Lucan Gazette (no online presence) 14/8/2004

"rigorously enforced by the Gardai as a 40mph zone"
"The speed camera at the location [opposite Spa Hotel] is the biggest revenue earner of all the speed cameras in Ireland, according to Lucan Gardai"
"SDCC stressed this week that the correct speed limit for the Ballydowd junction [Woodies] to the Spa Hotel is 40mph."
Now the paper maybe telling large porkies but I have my doubts so I rang the gardai in Lucan who said that they think(!) it is 50mph - article seems to be incorrect
K.

silverside

We could look at what happened in other countries when enforcement was stepped up. We could look at the state of Victoria, Australia, which reduced accident rates significantly by stepping up enforcement. We could look at Durham county, UK, which has higher accident rates than surrounding counties, in part due to it not having any speed cameras. Short of doing repeated experiments I don't think you will be satisfied. Alun, are you still arguing that the reduction in deaths after penalty points was introduced was a 'blip'?

Seth Brundle

was enforcement of road traffic offences actually stepped up (and maintained at at least that level) after the introduction of penalty points?
It is possible that the reduction in accident rates was purely down to the fear that you may be caught but people slowly realised that they are unlikely to be caught so offences returned to pre-points levels

silverside

I think that is what happened. The gardai made a show of enforcement, and I did notice a lot of speed checks the first few months, but I think people and the Gardai have gone back to their old ways.

I think bringing in a dedicated traffic corps would do wonders, along with a lot of speed cameras, but that is another argument.

Alun

silverside wrote:

We could look at Durham county, UK, which has higher accident rates than surrounding counties, in part due to it not having any speed cameras.

Wherer did you get that gem of (mis)-information from?! A look at any speed camera database for the UK will reveal that there are indeed speed cameras, both stationary and mobile, in Co. Durham. And in any case, there could be many more factors that could influence these figures, such as road densities, types of roads (Motorways vs. single lane roads etc. ). There is no reason on earth why every county in the UK should have identical accident figures, speed cameras or not.

Alun

kbannon wrote:

was enforcement of road traffic offences actually stepped up (and maintained at at least that level) after the introduction of penalty points?
It is possible that the reduction in accident rates was purely down to the fear that you may be caught but people slowly realised that they are unlikely to be caught so offences returned to pre-points levels

I agree. Although I didn't notice people driving any more safely after the introduction of penalty points, I did notice an increased level of "nervousness" if they even caught sight of a Garda car in the distance, or stopped by the roadside dealing with another car. By this I mean cars in the outside line of a dual carriage way or motorway, who previous to spotting the Garda were cruising along quite legally at 60 or 70 mph suddenly slamming on the brakes to coast past at 45 or 50 and then speed up again afterwards.

silverside

Alun, have a look at the following link

It claims that the UK government has shown an average reduction of 40% in accidents where it has introduced speed cameras.

http://www.paconsulting.com/news/about_pa/2004/About_PA_Speed_Cameras.htm

What do you think?

and OK, maybe Durham isn't the best example. But their Chief Constable *has* refused to bring in more speed cameras.

Alun

I've seen that report before, and what bothers me about it is that there's (as far as I can see) no analysis of what the trend in collisions and casualties is at non-camera sites, i.e. is what we're seeing a general downward trend unrelated to the presence of cameras at a particular site. Also some of the "headline figures" are a little unbelievable ... A1, Ellesley, Notts / 13.7 / 0.7 for example, or was there a coach crash there one year, and virtually nothing the next, and the "massive" reductions in speed of an average of 2.4 mph resulting in a 71% reduction in accidents ... pull the other one!

Mind you I've worked on a project together with PA Consulting in the past, so I'm not entirely surprised at their impartiality (not!).

unkel

silverside wrote:

the UK government has shown an average reduction of 40% in accidents where it has introduced speed cameras

From the sultans of spin :rolleyes:

I consider myself to be a safe driver, but I know for a fact that I'd be safer on a motorway without speed checks than on one with speed checks without knowing if / where they are. I'm afraid of penalty points alright and the extra attention my speedo gets now and then, albeit only for a fraction of a second, is not good

One thing that has to be said for the UK though is that fines / points are not distributed on the motorway unless you do over 85mph (where the limit is 70mph). That sure is a reasonable margin. I've been told the margin here is only a few miles...

Our man in Havana

The threshold here would be 79mph. Speed limit + 10% + 2mph=prosecution threshold.

Blitzkrieger

This argument is going around in circles.

I think we all acknowledge that Silverside has a point, but we all feel he's missing the core issues. I think if you ask just about anybody they'll agree that we need more driver training. Hibernian give a 30% discount on top of your no claims discount if you pass an advanced driving test. That's how much they believe in increased driver training and they'd be the one's to be hit in the pocket* if it wasn't a problem. Several companies in the UK give a discount too. You see horror stories on the TV like britain's worst driver, where a guy with a full license has written off 36 cars, or a girl who just passed her driving test can't handle roundabouts and almost crashes at every one they encounter while filming. The current testing and licenseing system just isn't good enough.



* Who would then, of course, pass the cost back onto us.

unkel

Bond-007 wrote:

The threshold here would be 79mph. Speed limit + 10% + 2mph=prosecution threshold.

That seems reasonable enough. It conflicts with what I've been told though. Any backup / link for this Bond-007?

Also, does this apply to all speed measurement technologies, i.e. gatso, fixed camera, laser gun, etc?