Monday 9 August 2010

Is this just regression to the mean?

I hope that it's obvious that I'm not a fan of speed cameras. I regard them as pointless, thoughtless enforcers of a stupid law. But I wonder if people aren't setting themselves up for a fall here:

Accident data shows that in the first nine months after the devices were scrapped in Swindon, there were 315 road casualties in the area as a whole, compared with 327 in the same period the previous year.

In total there were two fatalities – compared with four in the same period previously – and 44 serious injuries, down from 48.


And the fact of the matter is, this is no different from the equivalent falls crowed over by advocates for speed cameras when the cameras are put up. At some point the numbers will crawl up, and then they will fall again. Put a camera up, take it down, the real thing to remember is: shit happens.

13 comments:

patently said...

Yes, definitely. Far too soon and far too small a sample to reach anything approaching a conclusion.

The significance, though, is that the doomsayers predicted a slaughter of the innocents on the roads the moment that the infernal devices were switched off. Which they had to, having spent years telling us that they were essential to road safety as we know it. So if (over time) the accident rate doesn't leap dramatically, then the devices will be damned.

marksany said...

Oh yes. Regression to the mean should be on the exam paper you have to take before you are allowed to vote.

AntiCitizenOne said...

I'd prefer that people sit an exam showing they can perform compund interest calculations before they are allowed to borrow at interest bearing rates above inflation.

Anonymous said...

OK, maybe you're right, maybe you're wrong. But arriving at your decision with only two years of data and seemingly generalising this decision is pretty dumb. I hope you're not somewhere making important public decisions.

patently said...

Anon - yes, that would be dumb. As dumb, in fact, as completely misreading the original post?

Anonymous said...

Yes, sorry. I just noticed the title of the post. :">

Anonymous said...

I suppose there's no way of knowing whether it's regression to the mean or not unless many more samples are taken.

You would expect the data to be contrary to the trend implied. I know I tend to unconsciously slow down when I see a sign for speed cameras. Assuming, I suppose, that you can correlate speeds with accident rates.

Obnoxio The Clown said...

Hahahahahaha ... Jeez. Sometimes I wonder why I bother to write this shit if people don't read it! :o)

Anonymous said...

That was dumb. I feel like a dick. I'll be more careful in future Mr Obo.

patently said...

Top tip: Obo is not the person to tell that you "feel like a dick".

We've all done it. We'll all do it again some time. If you can't laugh at yourself, who can you laugh at? (apart from Obo, of course)

Man with Many Chins said...

As a resident of Swindon, I think it's wonderful that the ecilops finally have to get out of their shiny new PFI station and get into their shiny high end 5 series and X5's and actually POLICE the fucking roads for the terrible driving round here. Speed is not and never has been the issue, it's drivers who are just shit, unobservant, distracted etc. Safe speed for the conditions is the order of the day. For example, in a residential area you will rarely find me above 20mph and busy watching for kids. Half the other cunters are steaming around on their phones, fiddling with their stereo/twatnav/heating arguing with kids in the back seat, looking at their partner etc whilst driving. I think ALL car drivers should be made to spend 2 years on a 125 scooter before they are allowed behind the wheel. Observation would drastically increase, and the general "I am invincible in my cage" would decrease.
That's my tuppence worth.

patently said...

ALL car drivers should be made to spend 2 years on a 125 scooter before they are allowed behind the wheel.

I agree with the principle, but prefer my implementation ;-)

Roger Thornhill said...

Driving in an original Mini might be just as good. Make a mistake and 300kg of iron lands on your legs, but until then, oodles of active safety by nature of its superb handling.