How safe are your roads ?

[bobins]

In a recent email newsletter to me from Direct Line insurance, they had a feature on the safety of all the main roads in the UK. You can find out just how bad the roads are around you from this (little bit clunky) website:
http://www.roadcrashindex.org/
It turns out that I'm surrounded by bad roads. Of the four roads that lead from here, only one of them is rated 'Medium Risk', the rest are 'Medium-High Risk'. West Sussex as a whole is rated 66 out of 78 counties i.e. somewhere down the bottom of the table for safe roads. I think I might lock the front door and stay at home.

[myglaren]

Although the county (Tyne & Wear) is listed as one of the worst - 20/78 the area around me (Washington) is quite low risk and I rarely travel the higher risk road anyway.

[bobins]

20/78 puts you in one of the better counties. The further down the list the county is, the less safe the roads are.

[myglaren]

The site said we were one of the worst and risked being nuked from orbit if we didn't pull our socks up.

[bobins]

I think the only roads that are considered safe are the little go-kart ones for the kids at Butlins

Butlins go-karts - Bump-to-baby.com, fair use

Us in West Sussex have been put on notice that we'll have to go and stand on the naughty step unless we improve our driving skills...... or at the very least gain some driving skills

[andy5]

20/78 puts you in one of the better counties. The further down the list the county is, the less safe the roads are.

The lower number positions are better, but the ranking seems to be about tendency, in order of whether things are getting better or worse, not an absolute comparison of current safety or risk.

So the top of the list has 32% improvement, and the bottom is 27% worse, comparing the period 2013-15 and 2010-12

I'm not a statistician, but I wonder if the results can be said to be statistically significant, will there be too much variation in such a short timespan or would it better to look at 7 or 10 year periods.

Useful to compare relative safety of local roads though, though we possibly already know by reputation some of the worse ones around us.

[bobins]

Ah, I see ! I was going on the background colour - the higher numbers seemed to have a red background colour, lower numbers a neutral colour.
I think you're right about the statistical relevance - it'd only take a particularly nasty single incident involving several fatalities to skew the results and make a whole county look bad. Similarly, if there was an unusual (freak) but nasty incident involving several fatalities / serious injuries several years ago but there hasn't been one since then that county's roads will appear to have got safer.