Skip to main content

Currently Strava provides the following stats about each segment: distance, elevation, and grade.

The problem is that these stats are only valid if a segment is a straight uphill or straight downhill.

But if a segment has both uphill and downhill parts the stats can be wrong or even outright misleading.

Here a good example. The segment below takes one of the steepest trails uphill to the top of a popular local summit, then goes down on another equally steep trail, then turns around and ascends the second trail, then goes down on the first trail.

The segment description shows the grade as 0%, which couldn't be further from truth. It also shows the elevation incorrectly as 2035 ft. In fact this segment has about twice the elevation gain and the average grade is closer to 25%. So not only this segment stats don't provide correct information about the segment but also it won't likely come back in a search as a hilly segment. 

I think Strava can do better than that. 

I understand how Strava gets these elevation and grade numbers, so there is no need to explain that. My point is that showing the grade as the difference between the end and the start divided by the segment length isn't going to tell me enough about the terrain in between, so that isn't useful except in simplest cases. Also, calculating the elevation as the difference between the lowest and the highest point isn't useful when segments are longer and have more complex terrain. Perhaps this is an example of using segments for something completely different that Strava originally anticipated, but this isn't very unusual as far as trail running segments go. These kinds of segments are getting increasingly popular. They often mirror FKTs. 

Hello @Silentvoyager 


Thanks for your post and nice to hear from you again.


Just so I'm clear - your point is that the % grade we show for some segments is misleading, because in segments that have a lot of ascent AND descent, net % grade is very small even though the elevation shown (in your example over 2000 feet) is significant.  Am I understanding correctly?


If that's the case, is your suggestion that we consider only ascent in a segment when calculating the % grade?  I'd be interested to know your thoughts on this.  


Yes, I think it is misleading because 0% grade suggests that the segment is flat while in fact it is mountainous - gaining or losing about 1300 feet per mile. Another point is that the actual elevation gain for that segment is twice the value shown - not over 2000 ft but over 4000 ft because there are two climbs in the segment.

I think, at the minimum the elevation fields should show the total elevation gain. I am not sure what it currently shows - I assume the difference between the lowest and the highest point. Ideally it should show both total elevation gain and the total elevation loss as two separate numbers.

Perhaps if both the total gain and the total loss are shown, then it should be OK to continue showing the net grade as Strava does now.


I assume Strava currently just finds elevation of start and end and subtracts? Is it possible to sample inside the path instead? That would allow for better total elevation change at least. As far as grade, perhaps a "median grade" or "average grade" and "max grade" would be helpful, calculated from the sampling? I agree with OP that currently the gain and grade are only useful under specific conditions and would like it to work more broadly


Reply