Idea for problem alleys?

[tonestertm]

OK. Thanks again. Lessons learned. Old data, don't forget to check the Livemap.... :chagrined:
Sorry to have wasted your time.

[tonestertm]

Thanks for all the great feedback! I guess the wiki needs updating/clarifying re: the penalty for junction nodes. (Not news, I know. The wiki alone is a pretty mammoth job. And there is no Add-ons tool....)

There are two spots immediately coming to mind, which are definitely not end-of-drive. Dunno about the old data thing, but I think you may be right (again) about the lack of actual User Reports, ottonomy. Good point. It does sometimes seem as though the automated Map Problems come out in batches, maybe there's a cache and the problem's still being held after the problem is solved, hence the broken display in your example, but I would doubt that.

Here's a permalink I've created for one of the usual suspects (though there's no Issue at the moment). Northbound La Cienega to Eastbound Santa Monica and I've actually seen all of these segments as Waze generated suggestion, with the user drives continuing on past on Santa Monica. Again, I suspect this usually happens at rush hour, and this whole area gets pretty busy and the route may be faster, but it's a parking lot road, fer cryin' out loud....
https://www.waze.com/editor/?lon=-118.3 ... 2,61655529
I have seen this one show up at least twice, perhaps as many as four times over a substantial length of time, without an evident problem on the map.

And, davipt, I don't think anyone responded to you; if I understand you correctly, these are not residential streets, but small access roads running parallel to main streets, usually, but not always, behind businesses. They're not meant for regular traffic, but access to rear entrances or parking lots, sometimes parking under/behind apartment buildings or even houses with a garage in the rear. Thanks for the ideas.

OK, here's another one that's live right now, and I know I've seen it quite some time ago. The difficult left turn out of the alley on to Cahuenga is even Disallowed!
https://www.waze.com/editor/?lon=-118.3 ... 0,62193649

[tonestertm]

First, let me Very Clearly state that I am not suggesting this as a general approach to alley drawing!

Around here there are a few alleys on the map which are recurrent "problem alleys", which Waze tries to route folks through (I'm guessing during heavy traffic times), and 95% of them ignore it. Solve the Map Issue and go on, you say. Let the ignorant suffer!

Like the proverbial cat, these notifications will come back again and again for these particular segments. There are no apparent problems with the map--Parking Lot type, turns all around OK, doesn't seem to matter if they are named and/or City'ed or not. In actual fact, they may be faster than taking the street, but in my experience using the alleys around here for wholesale general through-traffic is frowned upon, they're sometimes in disrepair, and most people don't follow these routes anyway (as evidenced by the repeated MP's). There have been numerous threads about this type of Map Issue, but I haven't found anything to actually solve this "problem", short of deleting the alley altogether (which is, of course, an option, given the ongoing debate about the value of even drawing alleys on the map).

So I wondered: what if we add a junction (or two) into the given segment to "slow it down" even more than the Exit-from-a-Parking-Lot-Road penalty? As much as I truly detest data sprawl, in these problem cases this might be enough to discourage Waze from trying to send folks through these places in all but the worst of traffic, thus saving weary editors from the grievous tragedy of yet another three clicks of the mouse every couple of weeks. This could work for situations of unwanted routing through parking lots as well.

Perhaps I'm just beating a dead horse up the wrong tree into the wind. Thoughts?

[Timbones]

This thread is also very old. Locked.

[pinpusher]

A few lines of code could fix the alley routing problem extremely fast. Only route to the nearest street unless the pin has been moved.

Per https://wiki.waze.com/wiki/Illinois#Alleys "As a general rule, alleys should not be mapped. They clutter the map and can cause routing issues." IMHO, Chicago alleys should be turned into linear place features or remove them, it as much quicker to remove them than to move 100,000 address pins.

[PhantomSoul]

It's sounds to me that the MP's based on old data are a side effect of the way these problem junctions are detected, and even though the problem may have been conclusively fixed, Waze will keep generating problem pins there until enough drives through that junction follow their suggested routes to counterbalance the deviated drives that triggered the MP to begin with.

I usually just mark them as solved if I can't find anything wrong with the junction. Most of these occur at the minor street and parking lot junctions of a road that approaches a known traffic-prone intersection and the deviations are people just turning to find a different way to go despite Waze wanting them to continue in the traffic, perhaps because it doesn't know about the traffic yet.

[ottonomy]

One thing to remember when looking at MPs of the type "Suggested route frequently ignored" / "Most users did not follow the suggested route" is that for some reason many of these seem to be generated by quite outdated drive and routing data, where the problem has been solved conclusively, but the Waze automated map problem routines keep putting them on the map over and over. I see these SO frequently, and yet so rarely do I see an actual user submitted map problem where Waze proposed a route over a parking lot road. In the cases where I do see such URs, I suspect that the situation is often what jemay has suggested, that we can't see the route end point in the UR, and that Waze didn't actually route PAST the parking lot or private road, but merely chose it because the destination address marker was closer to the alley than to the street segment.

There has been some debate about whether Waze is still using old data for these MPs, or if indeed people are still being routed over parking lot road segments. I submit the following exhibits as evidence that such MPs are still being generated after the problems have been solved. In this case, the modification dates of the disconnected segments were well over a month earlier than the date when I found this MP and captured the screenshots (March 9 / April 20). It's interesting to me that the Waze suggested route is broken approximately where the segments are disconnected. It's as if the automated process knows that the problem has been solved, but it can't talk to itself...

SuggestedRoute1.jpg (65.17 KiB) Viewed 1241 times

ModDate.jpg (16.07 KiB) Viewed 1241 times

I've been following this issue as it pops up in thread after thread, and I'd really appreciate a PM if any of you editors out there find examples of actual user submitted map errors where its clear that Waze has routed through a parking lot road or private road segment which is not at the end point of a route. Only with URs can we verify that the routing occurred after a segment road type has been set.

[mapcat]

Permalink?

Unless things changed recently, junctions that only connect two segments have zero effect on routing.

[jemay]

OK, here's another one that's live right now, and I know I've seen it quite some time ago. The difficult left turn out of the alley on to Cahuenga is even Disallowed!
https://www.waze.com/editor/?lon=-118.3 ... 0,62193649

Another example of OLD DATA being used for the automated map problems. In the live map, waze will only do a right turn out of that street https://www.waze.com/livemap/?lon=-118. ... g=62383329

[jemay]

The MP/UR do not have enough information for all the alley ones that I have seen, all I have seen are the route instructing them to turn down the alley but not "through" the alley. So I suspect what bigbear stated about the location (stopping point)