Average Speeds in Strava and Garmins & Cheating Segments

Average speed can a good way to track performance over time, and also compare between riders. After all, how many of us are not a little bit competitive?

But average speed, if calculated in certain ways, can be very misleading. It can make you look a lot faster or a lot slower than you really are.

Example 1: Record only the fast parts
A group I ride with, which incidentally is very fast, likes to make themselves look even faster by only recording the core (fast) sections of their rides.

They don't start their Garmins until the ride has really begun - all the time coasting out of their driveways and around stop signs to get to the main road gets excluded.

Same goes for the end of the rides - they stop their devices before any cool-down, right after the final sufferfest and final sprints.

The result is high average speeds but low(er) total distances.

Don't get me wrong - there's nothing wrong with this. It's a matter of preference: Do you want faster or longer rides overall?

A real example. The ending is where the rider starts his slow section - the cool down.

These rides become apparent on Strava as the start and end are usually the beginning of the fast parts of a ride, not anybody's house. If the red line meanders into some small residential roads (the starting or ending of the ride) he's probably recording the entire ride, straight from home.

Or, like in the example above, the start and ending are at different places.

Example 2: Enable auto-pause
If you use a Garmin, you'll find the auto-stop feature under Menu -> Training -> Auto Pause/Lap. By setting "Auto Timer Pause" to "When Stopped", the timer will momentarialy stop when you are at red lights or stop signs.

The Auto Timer Setting in the Garmin Edge 705

With this setting off, at stop signs you will observe two things:
1. Your timer will continue
2. Your average speed will slowly erode. So if you are stopped for long enough, your average speed will eventually hit zero, or very close, even if you've been hammering it at 45 km/h for the whole ride.

Or, instead of selecting "When Stopped", you could use the "Custom Speed" setting, which will stop the timer whenever your speed is lower than the custom speed. For example, if you set it to 5 km/h, the timer will stop when you are going 4 km/h, for example. This is to eliminate any erroneous timing as a result of any minor movements while essentially stopped.

Protip: Some people claim is possible to cheat Strava by setting a custom speed and then completing a segment by moving slower than that custom speed. By doing so, you can get the KOM.

However, Strava's take differs: 

"...stopped time is included in segment time to keep leaderboards competitive. So using autopause does not allow for cheating on segments. We time segments based on the time stamp from when you enter a segment and when you finish it, it is not based on the "ride-time" that occurred on a segment but rather the elapsed time."

Source: Strava Knowledge Base

Personally, I record my entire ride, including the slow parts just getting to the main roads, and the cool-down coming home. This gives me a lower average speed.

I do enable the auto-pause; I don't think it's accurate to be recording stops, whether they are breaks or red lights.

However, if you are using Strava, most of this really doesn't matter. Strava will automatically crop out your stops, giving you that faster average speed. So you could have an average speed recording of 28.0 km/h on your Garmin, but 30.0 in Strava.

This seems consistent with the idea that you can cheat and get KOMs by completing a segment slower than your custom threshold speed. In fact, you can even set your custom speed to 99.9 km/h....

So next time you see a 40km/h average find out how it was recorded, and next time you see the slowest guy around getting KOMs which you know he's not capable of, ask him what his threshold speed is set to.