New user,
Waze auto calculating speed in traffic is absolutely wrong in almost all cases.
Been using waze for a couple weeks now. The main think I’ve learned is that the “speed bulletins” which get posted in slow traffic are dead wrong, and in many cases can cost you a lot of time sitting in traffic. Today it took my 18 minute drive, told me to take a higher traffic route, because it would be faster, which took me 30 minutes. The root cause of the failure was the calculation it used in a high traffic street to determine average speed. The calculated speed on the app was 9mph in traffic. This was dead wrong. Traffic was Stopped with a real average of 1-2 mph. but as traffic slowly moves and the cars in front of you finally move, apparently when everyone races for 3 seconds at 9mph into the bumper of the stopped car in front of them to then stop again for another minute or so… Waze seems to record these lurches of impatience as an average speed, posting it to the general populous and using it in calculations. This needs to be rethought as today this defect almost doubled my commute time. Easier said than done, i’m sure, but waze needs to be calculating the time it takes for users to cover predetermined lengths of road to calculate speed. All I know is the current method doesn’t work and is posting inaccurate data in every single instance of this I have encountered.
That is exactly what it does. Or more specifically between junctions. Are there a lot of users in your area?
Well… Then it’s broken 
The road I took this morning is always stand-still as it’s under construction. This begins before 7am and continues to be stand still till after 9am. I entered this stretch of road at 7:43 am where it had been grid-lock for over an hour, and waze had the average at 9-12 mph. this stretch of road in 3/4 of a mile, and it took me 16 minutes to complete. This puts the speed at 21 min per mile, or roughly 3 miles per hour. that’s 300-400% wrong.
There are several wazers that commute through this area every morning as I’ve been seeing their alerts and avatars. there’s 6-10 groups in my neighborhood.
I think the average is over the whole colored line, is it possible that the traffic alert line is longer than the stop and go traffic, so is getting adjusted by “faster” traffic on the end of the jam?
Us forum trolls can only tell you how we think it works, and occasionally how waze told us it works at one time (not that features change behavior over time ). I would suggest contacting support to ask them to look into it. Provide them a permalink from the livemap, and date/time to investigate.
No sympathy from someone who sits in traffic 1 to 1.5 (sometimes 2) hours each way for a 20 mile commute. The price of paradise? 
Although, in relative terms, doubling your commute is not a good thing. I can appreciate that.
Did you check the segment a few minutes after you left it? Normally, the reported speed is slowly increasing to the normal average if no wazers pass the segment. I think it that is why you saw a reported speed that was higher than you experienced. It should have updated with a lower speed after you passed.
That sounds very likely, and what I would’ve said… Previous data.
Sent from my GT-I9100P using Tapatalk 2
And in case anyone who lives in a mostly rural area with few Wazers stumbles into this thread — you can’t trust the traffic speeds Waze generates at all, and it won’t be fixed unless more Wazers drive your routes. I know from experience, because I’m one of only 3-4 total Wazers in my area. Those average speed notices are never, never, never right.
But they will be eventually
Sent from my GT-I9100P using Tapatalk 2