First, this segment:
Originally Posted by
TrojanHorse
Perhaps you are familiar with the term "segment poaching?"
Here's a fun one near me and it's long enough and hilly enough that it's difficult to "poach"... but check out the first page and see who's dominating the stats:
http://app.strava.com/segments/830079
...has a couple sections where the gradient is 100% plus!
Second, the leaders times at about 16+ mph seem about right. If anything, they seem a bit slow for "leaders" on the segments (fast for me, but not some I think of).
Third: we have a weekly training ride (Montrose Ride) that, if there's less than 100-riders riding it, that's a small turnout. Yet Strava only has about 20-30 people reporting having ridden it that week.
In conclusion: Strava isn't accurate nor reliable because a) it's data/interpretation is faulty, and b) not everyone reports/uses Strava, and c) those that do aren't always the fastest/strongest.
Therefore, don't draw any conclusions from Strava.