Roundabouts have better throughput than a busy 4-way stop, but less throughput than a signaled intersection if the timing and sensing is reasonable (many signaled intersections don't have reasonable sensing). Roundabouts also have some pretty nasty worst case wait times; I'm really not looking forward to the state installing one near me on the approach to a car ferry; it won't be fun to wait for 200 cars to go by before you get a turn to go, and I expect long ferry lines to result in impatient people in the ferry line blocking the roundabout. Sometimes there's two hours between ferry loadings. Going to be some fun times.
Personally, I find it challenging to both look ahead to the right to confirm I have room to enter the roundabout, look to the left to confirm there is no traffic that I need to wait for, as well as looking far left and right to ensure there are no pedestrians crossing soon. Signaled 4-way perpendicular intersections have worse outcomes when a participant doesn't follow the signalling, but indication of right of way makes it easier to confirm at a glance if it's safe to proceed.
That’s not actually true. It’s entirely possible for them to have the same footprint.
Personally, I think we could replace a LOT of stoplights with roundabouts. Way better throughput and faster travel for everyone.
There are also solutions for large vehicles where the center is raised but not impassible.
-stop in the roundabout
-stop before the roundabout and let their brain buffer for 30 seconds.
-somehow go the wrong way in the roundabout
-fail to yield to traffic in the roundabout
Is way too damn high. It makes traversing one a high stress situation since you have no idea if grandpa grunt and run in to you is about to perform a confusion based terror attack on the traffic control device.
Indiana drivers seem to much better in general with a lower incident rate of "omg that guy almost hit me".
With this said roundabouts that service a fixed area, such as a neighborhood without much cross traffic seem fine in general. Whereas roundabouts in areas that pick up new traffic are far more prone to incidents. And god help you if the roundabout is in a tourist area.
One of the problems with roundabouts in the US is there are too few of them so you're always running into someone who has never dealt with one before which increases the risk of unexpected behavior.
Don’t forget that at a roundabout the risk of injury from unexpected behavior by other drivers is _lower_ than at a signalized intersection. There’s a good reason why the injury rate goes down wherever they are built.
You yield to traffic from the left, which mean someone from a leftward entrance has priority, but they can actually be blocked by other traffic. So you have to not only consider yielding to them, but also whether they are yielding to someone else, thus giving you space to go. I see this computation mess people up all the time.
Also, judging intentions is much harder. On a multi-lane highway, it's very clear when someone is cutting across lanes to exit. And there's only one place they can be exiting. On a multi-lane roundabout, they might be taking the exit before your entrance, or the one after. Often people won't be signalling, or even giving incorrect signals.
When joining as well, if I'm emerging onto a busy road with two lanes in the direction I'm going, I will probably accept joining when the nearest lane is clear, even if the next lane is not, as long as the cars there don't look to be moving into the nearest lane. On a roundabout people can peel off at any time, and you should really wait until there's a gap in all lanes.
In some sense you have to start sometime, but there's going to be pushback from the accidents and injuries that will certainly happen in the interim.
> there's going to be pushback from the accidents and injuries that will certainly happen in the interim.
In areas that have actually built lots of roundabouts the accident and injury rate dropped immediately. There was no interim period with higher accident rates.