Crash prone 'modern roundabouts'

[kphoger]

As much as we love to say that life is "invaluable", there is a number to use - and it's not very high. Same with injuries, more granular data on injuries would be interesting to flatten into financial value.  Same with extent of car damage.

And, see, I've been assuming that, when a department is tasked with "fixing an intersection", the reason is primarily to reduce the number of deaths and serious injuries that are occurring there.  So, if they do that, then I figure it's a success—even if other types of crashes go up.  That is to say...

— Too many traffic fatalities are occurring here.  Fix that.
— There.  Fixed.  Installed a roundabout.  Fatalities have gone down.
— But non-fatal accidents have gone up.
— That's what they told us to fix.

[kphoger]

Last, but not the least for me personally... I do have to cross the road near roundabout these days (thanks to certain people messing with parking arrangements, but whatever). And it IS tough. I am paying attention, I do make eye contacts, I can move fast. But it's just a high risk exercise.

This is a great point kalvado makes.  Looking at the safety within 250 feet of an intersection (whether it be a traffic signal or a roundabout) is great, but how does the intersection impact safety a 1/4 mile down the road for a driver trying to pull out of their unsignalized sub street? There are several traffic signals in the county i live that cycle 24/7 to create gaps in traffic to allow drivers to more easily exit nearby sub streets. A free-flowing roundabout corridor could potentially lead to more injury accidents mid-block.

First of all, that sounds awfully hypothetical—which is surprising to hear from you, who are usually interested in actual statistics.  I'd love to see an actual study of this sort but, failing that, let's just leave it as pure imagination.

Secondly, I agree that free-flowing traffic exiting a roundabout makes it more difficult for cross-traffic nearby.  You mentioned vehicular traffic coming out from a side-street, but I was thinking even more about pedestrian traffic trying to cross at a point mid-block nearby—especially those not crossing at a marked (and actually yielded to by drivers) crosswalk.

[tradephoric]

First of all, that sounds awfully hypothetical—which is surprising to hear from you, who are usually interested in actual statistics.  I'd love to see an actual study of this sort but, failing that, let's just leave it as pure imagination.

I agree.  I have had the thought that roundabouts may lead to more accidents downstream due to the free-flowing nature of roundabouts leading to fewer gaps in traffic, but i haven't dwelled on it seeing that it is so difficult to quantify.  A study that attempts to look at the impact would probably have some pretty big holes in it.

[kalvado]

As much as we love to say that life is "invaluable", there is a number to use - and it's not very high. Same with injuries, more granular data on injuries would be interesting to flatten into financial value.  Same with extent of car damage.

And, see, I've been assuming that, when a department is tasked with "fixing an intersection", the reason is primarily to reduce the number of deaths and serious injuries that are occurring there.  So, if they do that, then I figure it's a success—even if other types of crashes go up.  That is to say...

— Too many traffic fatalities are occurring here.  Fix that.
— There.  Fixed.  Installed a roundabout.  Fatalities have gone down.
— But non-fatal accidents have gone up.
— That's what they told us to fix.

Are there a lot of intersections in roundabout throughput range where fatal accidents are more or less regular events, not one-off-in-decade?