Crash prone 'modern roundabouts'

[tradephoric]

I don't religiously oppose roundabouts unlike some on here. Moreover, the data is pretty clear they work. If you parse anything long enough, you can find the results you want.

I don't oppose all roundabouts.  I oppose roundabouts that have percentage increases in crashes that correlate to the bitcoin bubble. 

To expand on this, complex roundabouts with 2x2 and higher geometries have been shown time-and-time again not to reduce total crashes (in the case of Ann Arbor a complex roundabout saw a 1000% increase in crashes).  In fact, not one documented complex roundabout in America has seen a reduction in crashes when comparing the 3 years of before and after crash data.  But roundabout proponents are quick to point out that these complex roundabouts still reduce injuries.  But do they really?  Recently published research out of Minnesota found that complex roundabouts saw a 6% increase in injury crashes.  The Minnesota research is the first of its kind that specifically looks at the safety performance of complex roundabouts.  Anybody who claims that complex roundabouts reduce injury crashes can't cite research to support their claims. 

[jakeroot]

I'm far more interested in the views of an elected public official who has to deal with the consequences of his policy decisions than some random internet jockey.

That said, the data are abundantly clear about the utility of roundabouts, particularly in lower-density areas like Indiana etc.

Then why are you wasting your time on this (or any other) forum to begin with?

Because I still enjoy the discussions. But I don't know why some people seem so inclined, almost to the point of pathology, to oppose the roundabouts. I think some people are just scared of change.

He seems more opposed than most people; I think most engineers would approach the issues raised in this thread with caution, rather than the definite conclusion that tradephoric seems to think he's reached. We can't deny that he's onto something, though. I love roundabouts, just from an aesthetic point of view. But if they aren't as safe as we all thought, why are we defending them?

[kalvado]

I'm far more interested in the views of an elected public official who has to deal with the consequences of his policy decisions than some random internet jockey.

That said, the data are abundantly clear about the utility of roundabouts, particularly in lower-density areas like Indiana etc.

Then why are you wasting your time on this (or any other) forum to begin with?

Because I still enjoy the discussions. But I don't know why some people seem so inclined, almost to the point of pathology, to oppose the roundabouts. I think some people are just scared of change.

He seems more opposed than most people; I think most engineers would approach the issues raised in this thread with caution, rather than the definite conclusion that tradephoric seems to think he's reached. We can't deny that he's onto something, though. I love roundabouts, just from an aesthetic point of view. But if they aren't as safe as we all thought, why are we defending them?

These days it is really hard to relay a message until it is delivered with a good dose of doom, hysteria and end of the world message. Looking at things from that perspective, Trade is fairly mild and tries to keep emphases on facts and statistics with just thin icing on his cake.

[jakeroot]

I'm far more interested in the views of an elected public official who has to deal with the consequences of his policy decisions than some random internet jockey.

That said, the data are abundantly clear about the utility of roundabouts, particularly in lower-density areas like Indiana etc.

Then why are you wasting your time on this (or any other) forum to begin with?

Because I still enjoy the discussions. But I don't know why some people seem so inclined, almost to the point of pathology, to oppose the roundabouts. I think some people are just scared of change.

He seems more opposed than most people; I think most engineers would approach the issues raised in this thread with caution, rather than the definite conclusion that tradephoric seems to think he's reached. We can't deny that he's onto something, though. I love roundabouts, just from an aesthetic point of view. But if they aren't as safe as we all thought, why are we defending them?

These days it is really hard to relay a message until it is delivered with a good dose of doom, hysteria and end of the world message. Looking at things from that perspective, Trade is fairly mild and tries to keep emphases on facts and statistics with just thin icing on his cake.

But he's losing his credibility by posting these "think of the children" posts every other day. As I've mentioned before, you can't take traffic engineering personally. People will find a way to kill themselves using something you've designed. We need to be more rational. Trade isn't helping his cause by reporting on individual crashes. His whole point is how many crashes there are.

[kalvado]

I'm far more interested in the views of an elected public official who has to deal with the consequences of his policy decisions than some random internet jockey.

That said, the data are abundantly clear about the utility of roundabouts, particularly in lower-density areas like Indiana etc.

Then why are you wasting your time on this (or any other) forum to begin with?

Because I still enjoy the discussions. But I don't know why some people seem so inclined, almost to the point of pathology, to oppose the roundabouts. I think some people are just scared of change.

He seems more opposed than most people; I think most engineers would approach the issues raised in this thread with caution, rather than the definite conclusion that tradephoric seems to think he's reached. We can't deny that he's onto something, though. I love roundabouts, just from an aesthetic point of view. But if they aren't as safe as we all thought, why are we defending them?

These days it is really hard to relay a message until it is delivered with a good dose of doom, hysteria and end of the world message. Looking at things from that perspective, Trade is fairly mild and tries to keep emphases on facts and statistics with just thin icing on his cake.

But he's losing his credibility by posting these "think of the children" posts every other day. As I've mentioned before, you can't take traffic engineering personally. People will find a way to kill themselves using something you've designed. We need to be more rational. Trade isn't helping his cause by reporting on individual crashes. His whole point is how many crashes there are.

You didn't get it.

Then play Sim City. In the real world, political context is a major influence, always has been and always will be.

"Think about kids!" is one of those things which are perfectly winning cause for 90% of voters.  Trade uses that, but uses pretty mildly - that's my point.

[jakeroot]

But he's losing his credibility by posting these "think of the children" posts every other day. As I've mentioned before, you can't take traffic engineering personally. People will find a way to kill themselves using something you've designed. We need to be more rational. Trade isn't helping his cause by reporting on individual crashes. His whole point is how many crashes there are.

You didn't get it.

Then play Sim City. In the real world, political context is a major influence, always has been and always will be.

"Think about kids!" is one of those things which are perfectly winning cause for 90% of voters.  Trade uses that, but uses pretty mildly - that's my point.

But this thread isn't a PR campaign. I'm influenced by facts. Hence why I'm not as pro-roundabout as I used to be. Lately I'm more anti-tradephoric. 

[kphoger]

In my opinion, this graph should be weighted to reflect the difference in total miles driven by rural and urban drivers (12,264 and 9,709 respectively).  The gap still exists either way, but it should be roughly 26% narrower.

[kalvado]

But he's losing his credibility by posting these "think of the children" posts every other day. As I've mentioned before, you can't take traffic engineering personally. People will find a way to kill themselves using something you've designed. We need to be more rational. Trade isn't helping his cause by reporting on individual crashes. His whole point is how many crashes there are.

You didn't get it.

Then play Sim City. In the real world, political context is a major influence, always has been and always will be.

"Think about kids!" is one of those things which are perfectly winning cause for 90% of voters.  Trade uses that, but uses pretty mildly - that's my point.

But this thread isn't a PR campaign. I'm influenced by facts. Hence why I'm not as pro-roundabout as I used to be. Lately I'm more anti-tradephoric. 

and how many people are able to change their mind  to begin with? Usually it takes significant authority (boss, professor, president) for people to change their mind.  I remember when you started to turn roundabout-skeptical - and I was telling to myself "no way I witness THAT!" You really broke my perception of how these discussions work.   A more typical approach is

Moreover, the data is pretty clear they work. If you parse anything long enough, you can find the results you want.

This one is specifically interesting due to "data" statement.

[kalvado]

In my opinion, this graph should be weighted to reflect the difference in total miles driven by rural and urban drivers (12,264 and 9,709 respectively).  The gap still exists either way, but it should be roughly 26% narrower.

What is the reason for such weighting? You want to assign crash probabiluity to trip, as opposed to per-mile?
That may make sense if some trips are doomed to begin with, e.g. drunk trips. Which is entirely possible, quoting some random page:

Alcohol use — Drinking and driving is often a bigger problem in rural areas for several reasons. First, there are generally no taxicabs, subways, or other forms of public transportation to use as an alternative for driving while impaired. Also, because of the often large distance between homes and towns, having a designated driver often does not seem convenient in rural areas. For this reason, drunk driving occurs frequently and can result in serious accidents.

https://www.dolmanlaw.com/auto-accidents-common-rural-urban-areas/
But that justifies expected decrease of crash per-mile AND per-trip rate as population density goes up.

[tradephoric]

I remember when you started to turn roundabout-skeptical - and I was telling to myself "no way I witness THAT!" You really broke my perception of how these discussions work.   A more typical approach is

Moreover, the data is pretty clear they work. If you parse anything long enough, you can find the results you want.

This one is specifically interesting due to "data" statement.

Yeah exactly.  They reference "data" that proves roundabouts work, only to not post any "data".  Of course if they did cite that 2000 IIHS study, which is seen as the authoritative study on roundabout safety in America, they would fail to see how flimsy that study is anyways.  Of the 24 roundabouts analyzed in the 2000 IIHS study, only 9 were double-lane roundabouts and they were all concentrated in the Colorado ski-resort towns of Avon and Vail.  Here are aerials of the 9 double-lane roundabouts analyzed in the study:

(2 roundabouts):  I-70 & Vail Road interchange roundabouts in Vail, CO
https://www.google.com/maps/@39.6444378,-106.3779447,18z/data=!3m1!1e3

(2 roundabouts): I-70 & Chamonix Rd interchange roundabouts in Vail, CO
https://www.google.com/maps/@39.6280191,-106.4215908,409m/data=!3m1!1e3

(5 roundabouts) 2500 ft section of Avon Road in Avon, CO (Avon Rd is signed for 25 mph):
https://www.google.com/maps/@39.63489,-106.52215,16z/data=!3m1!1e3

The perceived safety of double-lane roundabouts in America is based on a few interchange roundabouts in Vail and a corridor of 5 roundabouts along a 2500 feet section of Avon Road (signed for 25 mph).  Far from comprehensive (and the study itself is only 15 pages long).  Good thing double-lane roundabouts outside of ski-resort towns in Colorado were analyzed... ooh wait, they weren't.

[kphoger]

In my opinion, this graph should be weighted to reflect the difference in total miles driven by rural and urban drivers (12,264 and 9,709 respectively).  The gap still exists either way, but it should be roughly 26% narrower.

What is the reason for such weighting? You want to assign crash probabiluity to trip, as opposed to per-mile?

Because the graph was used in support of the assertion that the urbanization of Carmel has been the reason for its reduction in vehicle crash deaths as measured over a certain amount of time.  Injury crash rates are measured in incidents per year.  The number of crashes at a given intersection over the course of a given year is not related to the total number of miles the average vehicle through the intersection drives in a year.

Think about it this way.  If Bobby's daily commute in a rural area has him driving 10 miles each way, while Johnny's daily commute in an urban area has him driving 8 miles each way (I'm pulling those numbers out of my butt just because they're approximately 26% apart), and they each get into one accident over the course of ten years–who cares how many miles each one's commute was?  The rate as it relates to this thread topic would be 1 accident in 10 years for both drivers.  But, as represented on the graph provided by tradephoric, it would come out to something like 8 accidents per million miles for Bobby and 10 accidents per million miles for Freddy.  Same amount of time, same number of crashes, different rates.

[kphoger]

I remember when you started to turn roundabout-skeptical - and I was telling to myself "no way I witness THAT!" You really broke my perception of how these discussions work.   A more typical approach is

Moreover, the data is pretty clear they work. If you parse anything long enough, you can find the results you want.

This one is specifically interesting due to "data" statement.

Yeah exactly.  They reference "data" that proves roundabouts work, only to not post any "data".  Of course if they did cite that 2000 IIHS study, which is seen as the authoritative study on roundabout safety in America, they would fail to see how flimsy that study is anyways.  Of the 24 roundabouts analyzed in the 2000 IIHS study, only 9 were double-lane roundabouts and they were all concentrated in the Colorado ski-resort towns of Avon and Vail.  Here are aerials of the 9 double-lane roundabouts analyzed in the study:

(2 roundabouts):  I-70 & Vail Road interchange roundabouts in Vail, CO
https://www.google.com/maps/@39.6444378,-106.3779447,18z/data=!3m1!1e3

(2 roundabouts): I-70 & Chamonix Rd interchange roundabouts in Vail, CO
https://www.google.com/maps/@39.6280191,-106.4215908,409m/data=!3m1!1e3

(5 roundabouts) 2500 ft section of Avon Road in Avon, CO (Avon Rd is signed for 25 mph):
https://www.google.com/maps/@39.63489,-106.52215,16z/data=!3m1!1e3

The perceived safety of double-lane roundabouts in America is based on a few interchange roundabouts in Vail and a corridor of 5 roundabouts along a 2500 feet section of Avon Road (signed for 25 mph).  Far from comprehensive (and the study itself is only 15 pages long).  Good thing double-lane roundabouts outside of ski-resort towns in Colorado were analyzed... ooh wait, they weren't.

Roundabouts "working" and roundabouts being safe aren't the same thing, are they?  I interpreted the statement "the data is pretty clear they work" as meaning "the data is pretty clear they're efficient."

[kalvado]

Think about it this way.  If Bobby's daily commute in a rural area has him driving 10 miles each way, while Johnny's daily commute in an urban area has him driving 8 miles each way (I'm pulling those numbers out of my butt just because they're approximately 26% apart), and they each get into one accident over the course of ten years–who cares how many miles each one's commute was?  The rate as it relates to this thread topic would be 1 accident in 10 years for both drivers.  But, as represented on the graph provided by tradephoric, it would come out to something like 8 accidents per million miles for Bobby and 10 accidents per million miles for Freddy.  Same amount of time, same number of crashes, different rates.

I have hard time buying that logic.

[kalvado]

Roundabouts "working" and roundabouts being safe aren't the same thing, are they?  I interpreted the statement "the data is pretty clear they work" as meaning "the data is pretty clear they're efficient."

Then what do you mean by "work"? Have some throughput? Yes, they do. So does an all-way stop and traffic light and grade-separated intersection.
So what?

[kphoger]

Think about it this way.  If Bobby's daily commute in a rural area has him driving 10 miles each way, while Johnny's daily commute in an urban area has him driving 8 miles each way (I'm pulling those numbers out of my butt just because they're approximately 26% apart), and they each get into one accident over the course of ten years–who cares how many miles each one's commute was?  The rate as it relates to this thread topic would be 1 accident in 10 years for both drivers.  But, as represented on the graph provided by tradephoric, it would come out to something like 8 accidents per million miles for Bobby and 10 accidents per million miles for Freddy.  Same amount of time, same number of crashes, different rates.

I have hard time buying that logic.

I'm not a statistician.  Please point out my flaw.

[jakeroot]

But he's losing his credibility by posting these "think of the children" posts every other day. As I've mentioned before, you can't take traffic engineering personally. People will find a way to kill themselves using something you've designed. We need to be more rational. Trade isn't helping his cause by reporting on individual crashes. His whole point is how many crashes there are.

You didn't get it.

Then play Sim City. In the real world, political context is a major influence, always has been and always will be.

"Think about kids!" is one of those things which are perfectly winning cause for 90% of voters.  Trade uses that, but uses pretty mildly - that's my point.

But this thread isn't a PR campaign. I'm influenced by facts. Hence why I'm not as pro-roundabout as I used to be. Lately I'm more anti-tradephoric. 

and how many people are able to change their mind  to begin with? Usually it takes significant authority (boss, professor, president) for people to change their mind.  I remember when you started to turn roundabout-skeptical - and I was telling to myself "no way I witness THAT!" You really broke my perception of how these discussions work.   A more typical approach is

Short of Jim Brainard becoming dictator of the US, and forcing all intersections to become roundabouts, people are able to think freely. Most people are fine with signals because they're used to them. And just as many are fine with roundabouts after getting used to them.

Most engineers are easily persuaded by facts. They're mostly all math geeks anyway. They live day to day based on logic and reason. It's when they start playing with roundabouts and DDIs that their emotions get involved ("less conflict points!"). The problem, as I've pointed out previously, is that not all engineers are privy to the same data. Hence why I wish tradephoric would start working on an independent report that he could present at a TRB conference (pay someone under the table to let you speak), instead of posting these meaningless stories about one-off crashes.

You can tell how much I like good-old signals by how often I use them in Cities: Skylines. Many of my old cities had roundabouts all over the place, but they often got backed up, so my current city only has two (although both are 2x2).

[kphoger]

Roundabouts "working" and roundabouts being safe aren't the same thing, are they?  I interpreted the statement "the data is pretty clear they work" as meaning "the data is pretty clear they're efficient."

Then what do you mean by "work"? Have some throughput? Yes, they do. So does an all-way stop and traffic light and grade-separated intersection.
So what?

I'm not the one who made the original statement.  But "working" could mean increasing throughput over a four-way stop, reducing tailbacks, etc.  You should ask AlexandriaVA what he meant, not me.

[kalvado]

Roundabouts "working" and roundabouts being safe aren't the same thing, are they?  I interpreted the statement "the data is pretty clear they work" as meaning "the data is pretty clear they're efficient."

Then what do you mean by "work"? Have some throughput? Yes, they do. So does an all-way stop and traffic light and grade-separated intersection.
So what?

I'm not the one who made the original statement.  But "working" could mean increasing throughput over a four-way stop, reducing tailbacks, etc.  You should ask AlexandriaVA what he meant, not me.

again, basic Trade's message is that it is OK to replace 4-way with roundabout, it is not OK to replace a traffic light with one (rephrasing a bit - 2x2 lane arrangement not working with stop signs)

[kalvado]

Short of Jim Brainard becoming dictator of the US, and forcing all intersections to become roundabouts, people are able to think freely.

[jakeroot]

^^
Laughs aside, he's basically dictator of the public works department in Carmel, IN. There's very little free thinking in that office, me thinks.

[kalvado]

^^
Laughs aside, he's basically dictator of the public works department in Carmel, IN. There's very little free thinking in that office, me thinks.

Laughing aside, it takes a bit of an effort to get some understanding of a topic. Granted, any driver has at least some understanding how standard traffic control arrangements work - but you need a better idea about underlying concepts to develop opinion going beyond "i like/don't like it"
Especially visible on politically charged topics - and roundabouts also slowly drift into that category.
As for other examples... "Zisis sajens" mobs are probably latest one.

[tradephoric]

I wish tradephoric would start working on an independent report that he could present at a TRB conference (pay someone under the table to let you speak), instead of posting these meaningless stories about one-off crashes.

This would probably be the first slide in my TRB presentation.  The average crash rate of complex roundabouts in America is about 4 MEV.  This is looking at crash rates of complex roundabouts throughout the entire country, not JUST ski resort towns in Colorado.  Conversely, the average crash rate of a typical signalized intersection is below 1 MEV.  When crashes at complex roundabouts are about 4x higher than comparable signalized intersections, there is no guarantee there will be a reduction in injury crashes.  Put another way the big increase in total crashes offsets the fact that roundabouts have a lower crash severity.  Recently published research supports this, and found that injury crashes increased by 6% at complex roundabouts in Minnesota. 



Is it worth building these complex roundabouts when you see a big increase in total crashes and a wash in injury crashes?  I would point out all the roundabouts highlighted in yellow have been reconfigured to eliminate circulating lanes to simplify the configuration.  Communities are having to deal with these crash prone complex roundabouts.... Raleigh, Woodbury, Bloomfield, Farmington, Tri-Cities, Venice, Kitchener, Gahanna, and De Pere to name a few.  The roundabouts that haven't been downsized still experience significantly high crash rates.  Case in point M-5 & Pontiac Trail... constructed in 2011 it is consistently near the top of Michigan's annual top crash intersections list.   By the way 116th and Illinois should no longer be highlighted yellow... that roundabout was temporarily downsized for a construction project on 116th but it has been resized back to a 2x2.

[jeffandnicole]

I wish tradephoric would start working on an independent report that he could present at a TRB conference (pay someone under the table to let you speak), instead of posting these meaningless stories about one-off crashes.

This would probably be the first slide in my TRB presentation.  The average crash rate of complex roundabouts in America is about 4 MEV.  This is looking at crash rates of complex roundabouts throughout the entire country, not JUST ski resort towns in Colorado.  Conversely, the average crash rate of a typical signalized intersection is below 1 MEV.  When crashes at complex roundabouts are about 4x higher than comparable signalized intersections, there is no guarantee there will be a reduction in injury crashes.  Put another way the big increase in total crashes offsets the fact that roundabouts have a lower crash severity.  Recently published research supports this, and found that injury crashes increased by 6% at complex roundabouts in Minnesota. 

This is a hilariously incomplete list of 2x2 or greater roundabouts in the country.  Apparently the 'average' takes into account just high-crash roundabouts, and weeds out anything without a large crash rate.  That's like saying I was an A student in high school if you don't count any grades of B and below.

I'm especially amused that this list of American roundabouts includes Quebec and Ontario.

[kphoger]

I still want to know about the closest 2x2 roundabout to me that I'm aware of:  Sheridan/Ridgeview/Rogers in Olathe, KS.  This was constructed at least 13 years ago, because I can find mention of it from 2004.  That year, its max vehicle/hour count was 1784 in the afternoon rush.  I can't find any reports of an accident there at all, but I know my sleuthing skills don't match tradephoric's.

[tradephoric]

This is a hilariously incomplete list of 2x2 or greater roundabouts in the country.  Apparently the 'average' takes into account just high-crash roundabouts, and weeds out anything without a large crash rate.  That's like saying I was an A student in high school if you don't count any grades of B and below.

You were an "F"  grade student because you didn't show your work.  You could easily prove I ignore low-crash rate complex roundabouts by simply citing them.  Let's say any roundabout with a crash rate below 0.8 MEV is considered low (that is about the average crash rate for a signalized intersection).  You yourself said my list was hilariously incomplete.  Then there must be a hilariously high number of complex roundabouts with low-crash rates i ignored.  Humor me.