Waze в деталях

We are told to expect some changes in few weeks, so it is perfect time to prepare one more  sample route  to so we can see if anything actually changed or not. Let's examine Estonia-Russia border near Narva. Trying to get same rote at different days of week and time of day, we can see that on russian side, speeds never change, while on estonian side speeds vary alright. Note that place where speeds (un)freeze is not exactly the state border, but is shifted somewhat west of the border. Let's try another route nearby . Speeds, again, are fixed (I mean frozen, not repaired) near Russia, and from some point start to behave normally. One segment on estonian side keeps fixed estimation of 43 km/h - but it is not a default speed for Minor Highway road type. Is it another issue? Anyway, we'll see what happens later.

I did not check each and every country in the world to see if speeds history is absent. It seemed functional everywhere except Russia. In Belarus it was fine, but previously I examined Minsk only . Now, trying to figure where is the line between working and broken historical data let's see some roustes around Russia-Belarus border here  https://www.waze.com/editor?env=row&lon=30.07857&lat=55.83259&zoom=0 Some animated gifs are below. We see speed estimations remain default all week in Russia; then there is some life in Belarus near the border, but then stops and is still for quite a while; then gets to life again. So, if we say that traffic history broken in Belarus (or Uzbekistan, or in other country) we have to specify where exactly it is broken.

About a year ago, May 2019, Waze editors in Russia found out that calculated routes and time estimations show no sign of speeds history anymore. All historical speed data gone for good, and does not grow back. It would it was deleted, but it does not. So we conclude it got broken in a way so that speeds became inaccessible. Only historical data is broken. Realtime traffic data was still functional. Only happens in Russia. Neighbouring countries are ok. Some claim it happened much earlier, at 2019 beginning, or even fall 2018, they've noticed something was wrong but were not sure enough to explain. Historical data loss was somewhat disguised by realtime data making it not so evident. What they saw: weird routing, over-optimistic time evaluations in places where it used to be fine -- all of this occasionally happenned before, now became a tendency. It has becamer clear what's wrong when we tried to 'Plan a drive' - histogram had some bulge at present time, then quickly