CBenson wrote:

fvwazing wrote:Wazers on the blue track will report THE NEXT SEGMENT closed. As soon as that block works, Wazers coming from the same direction will be guided off the highway ONE EXIT TOO EARLY and into the roadblock - over the green track.

So if I understand this correctly, the reporting wazer will be immediately rerouted based on his/her own closure report. But if on the freeway, where the next segment is just the freeway segment between the off ramp and the on ramp, the reroute can be straight back down the on ramp to the closed freeway. It seems that is what howsmith reports here.

In my first example I assumed that at least the traffic coming down from the last-offramp-before-the-closure would be guided to the next junction after the actual accident, some km further. All traffic coming from behind would be guided as per stupid deviation around the jam reported by the original reporter (assuming there is a jam before the closure; typically there is) and would have to report some closures more.
In MY almost-perfect part of the Waze-world, off-ramps are not connected to on-ramps wherever possible - howsmith ran into a ramp that is clearly mis-configured and in his case he would have had to report another closure - of the onramp.
Anyway, you need a lot more closure-reports for a single stretch of freeway.

No, as in my typical freeway-examples above you need 3 or 4 closures for a stretch, each of these confirmed by a number of Wazers, within a reasonable time. Will not happen soon, and until every mousehole is closed expect routing to bend over backwards to send you into that sheet of ice.

jwkilgore wrote: The solution is simple:
- Expand the closeable roads to include 3-4 segments in either direction starting at the one you're on, including the segment you are on.

I am afraid the solution is not that simple. Yes, there are situations where your scheme could help in reaching at least more of the segments that you want to report closed. Even the typical #-crossing is impossible to close now. But with more options to choose from than the current "solution" it would take patience, a precise control over your fingertip and maybe even advanced gaming skills . It takes a very cool head to do that on the freeway, at the same moment slamming your bakes to avoid running into a jam, with ambulances and firetrucks whizzing by, police motorists clearing the traffic as it rolls, at night and in bad weather.

And it is not enough that YOU manage to do all that, some other Wazers have to confirm/copy your actions as well.

Curious - in the first image it looks like that short segment is "skipped" and the next (left)turn is offered.

"Others will be routed around it" is misleading at least.

Over the past months I may have seen 3-4 roadclosures (client-created) in the wild - that is very, very few for the whole of the Netherlands. And it isn't that we didn't try!

holmes4 wrote:I reported a closure the other day - a side street under construction for a week and half. Attached a photo too. Nothing shows on the map. I won't bother in the future.

It is not enough to report a closure alone - more people have to report the same closure. If reporting once is hard, reporting 3 times is hard to the power 3.

It should be possible to deduct closures from drivingpatterns, from the deviation of the proposed route. Something I call the "Reversed Ant Colony Optimization" (no need to google that yet!).

Can you see who created them?

[edit]
JDAzn and Rick, some plain Wazers probably.

Suggestion: rather than forcing the stressed user through a cascade of menus, have Waze DETECT closures from the deviation of the proposed route. Multiple users deviating at the same point make it ever more likely that a road is closed beyond that point. The part of the route between the deviation and the point where the original route is picked up again is then depreciated - until someone just drives through again.

It really isn't that simple. For one, the current system could stay as it is so that deeggo could report as he pleases (if he can reach the right buttons in time). But as soon as a Wazer leaves the proposed route a dialog could pop up (at the same moment the recalculation kicks in) asking if the deviation is a) caused by a closure b) caused by buildup of traffic ahead c) because of a maperror or d) because a better/faster/shorter route is expected that way. Or that dialog is omitted, and it would take more deviating Wazers to confirm a closure.

More complication: During the deviation the Wazer will be presented a number of times with alternatives that bring him to the original route, and these may be ignored as well. Do these deviations weigh in along with the original ones?