As we know, Waze collects trafficdata per half hour, which means that for the most-travelled roads there exist a kind of profile that would probably show one or two dips in the average speed. These dips are ofcourse the rushhours. Driving through a road during rushhour will almost never result in a trafficreport, because the speed is expected to be slow around that time anyway - but there is no visible feedback for that.
Couldn’t Waze show those roads as experiencing light or even heavy traffic on the map, showing them yellow, orange or even red? That would explain to anyone driving there why no trafficreport is generated - the speed is as slow as it is.
If someone manages to drive there way faster than expected, the colour could go back to normal - for maybe half an hour, where it could show traffic again, if that still is the expected speed. Always in relation to the profile!
As far as we know and can deduct from the information available to us (in the Wiki), it works (more or less) as described. Can be that the window is not half an hour, and can be that only very few roads have that much data collected for them that these profiles exist. But if they do: Waze, show them!
Furthermore, the comparison of current road speed to average road speed, for the purposes of traffic congestion reports, may not be using a time-window average speed, but a net average speed for the road.
We only get one average for the road map. Even if the routing is using a time-window, it may be too much of a server load to use a time-window average for traffic congestion.
By observation I’d be certain that traffic congestion is based on a total average road speed at all times.
That is, 25 km/hr is slow, based on the average 100 km/hr and will generate a heavy traffic report.
Even though the average speed +/- 2 hours of the time is 18 km/hr.
That would be nice, but I doubt it. As well as subject to change, I expect they might consider it a competitive advantage.
What do you mean by “traffic jam”. A traffic jam road report is manually generated. I would expect a complete standstill road report will always be generated, as the average road speed will never fall to zero. And in the event of a complete standstill Waze does all it can to avoid that road. So even if it is a t a complete standstill every day at that time I’d still expect to get a standstill report? Is that what you mean by “traffic jam”?
To put it another way, are there any roads where you see “heavy traffic” congestion reports at the same time every day for a month?
I see some congestion reports - indeed that’s what I meant - reoccur every day. It’s not that I’m keeping a list, but yeah, they’re there practically every day.
Well I do keep a list. So this morning, I see a traffic congestion (heavy traffic 25 km/hr ) at 0906 hours, as the traffic was heaveir than usual and so extended further in time. And I see moderate traffic on another stretch (18 km/hr) where the speeds are always slower at this time.
So what is heavy and moderate does change according to what is expected at the time of day. And traffic congestion reports have reduced at times that we expect to be congested.
It’s impossible to judge these things without a record. All human’s have selective recall bias. What I see from actual records is exactly what I would expect based on my mental model of how it should work. Over time Waze learns expected traffic density at expected times, and reports if it is abnormal from expectations.
I like exception reporting. Life is too cluttered with information. Just tell me the unexpected problems. Not the expected problems.
OK, so now I am confused as to the value of Waze to report realtime traffic conditions.
Google reports heavy traffic, but is frequently out of date by the time I get there. I need to know at least 30mins before if the UK M25 is jammed up (for example), to enable any chance to route around it - and frequently it shows a jam and when I get there there is none.
I hoped that with Waze, the wazer speed would be pretty instant and reported on the map. I mean, there’s no point taking an average of 1 or 2 wazers on a particular road in 1 day, over 30 mins. In time, with 1000’s of wazers an average makes sense, but I DON’T want Waze estimating that traffic will be 10mph at a certain time based on a sketchy previous profile. I want to know what wazers are experiencing at the time.
I noticed this morning for example that Waze was showing roads with a much faster speed(>30mph), and when I get there they are crawling along with rush hour traffic at <10mph. This is presumably because there is no real wazer speed to go by at that time, or it is a profiled speed based on previous days?
How can I tell if the map is showing a ‘guess’ speed, or a real speed?
Waze doesn’t show guess-speed but still uses it. And your only chance for useful realtime data is to have many many wazers around… (look at the netherlands on a mondaymorning). My proposal was, as a transitional solution while there are not enough wazers, to show the guess-speed by lack of the better - real time data. One could, indeed, give them a slightly different look…
Thanks fvwazing.
That’s what I would like to see - to know if the road speed being reported is an estimated one, or one based on current wazer data.
Lets face it - on a commute or any regular journey, you already KNOW where the likely traffic will be bad - but you really want to know if its gridlocked or accident-jammed, so a real wazer trace and speed is gold dust compared to a ‘projected speed’.