moesterry, I'm with you on thinking Waze should eventually figure it out from average segment speeds. I'm not positive it always does, though.

Thanks for looking, mapcat. This specific location came up because of an update request that was just under a year old. The segment average speeds may have normalized since then and Waze may not route that way any longer. Problem is, if I add a new rest area and the ramp segments have no average speed, waze will want to route someone that way just to get the data.

Good points, again. Makes me think that ramps should have higher penalty than freeway road type. But, I'm thinking of this more simply than the devs are, probably. I don't know all the drawbacks to adjusting a penalty like that...

The penalty for ramps would not have to be significant at all, but just enough so that if given a choice between a ramp segment and an freeway segment, and both have the same start/end point (not true in real life, but this is a simplified example), then the freeway segment wins.

1 freeway segment > 1 ramp segment, and certainly
1 freeway segment >> 4 ramp segments.

In some jurisdictions, doing things like using on-off-ramps get around a traffic jam or using a parking lot as an intersection cut-off will get you a traffic violation if caught. I don't think Waze should use ramps to route around traffic either. My example was predicated on no traffic, but even with, I say 'no.'

mapcat wrote:The main drawback I was thinking of was someone having to evaluate ramps individually and only assigning penalties under certain circumstances. I have no clue how the algorithm works, but it seemed that automatically giving all segments of one type precedence over all segments of another type could cause problems. Maybe only occasionally, but then the exit ramp routing problem is not occurring everywhere either.

Segments do already have a hierarchy of penalties. Parking Lot roads, for example, have a very high penalty.

Talisker66 wrote:Look at the GPS-Tracks. Maybe the ramp and the road are too close together AND the GPS-Track is only slightly off the map AND since the Routing is now via the Ramp Waze does not recognize that you stay on the road and "allocates" the data to the road.

I concede this may be a possibility. With the link from my OP, I saw that the ramp is right on the edge of the mass of GPS tracks. There really were very few on the other side, so I doubt that is what was happening. The average segment speed for the ramps are 8-16 MPH below the main interstate. I really can't fathom why Waze would want to route onto the ramps. However, I have adjusted the ramp a little further away from I-5, but I do have my doubts that in my original case, this was the issue.

I have not, and I do not recall anyone else writing as much either.

This is only my 2 cents comments but since version 2.0 of Waze, it seems my trust in the reliability of waze has decreased... At first it was almost impossible to reach the server to get the maps displayed but now this off/on-ramp issue.

After making a few feedbacks for wrong navigation instruction, I was told it could be that using the ramps might be quicker due to traffic and therefore Waze was behaving as 'normal'.
But not only I find people driving this way very asocial, it also asked me to leave the main motorway to for the ramp on lots of occasion at night while I'm alone or as such on the road...
As far as I can remember I didn't experience this behaviour in the past and this started a couple of months ago.

Last week I went to a totally new location and I had doubts about how I was lead there... my doubts were strong enough for me to stop and check the itinerary via the built-in Maps app of the iPhone...

I find the principle of Waze excellent and I want to keep using it while I'm driving within the borders... I would find it disappointing to have to purchase TomTom or similar app because I come to the conclusion I cant rely on the free stuff...