Average Speed Usually Isn't
 

I know cyclists like to use average speed as a benchmark. However, benchmarks by definition are not accurate. The problem is that most use moving time to figure average speed. This invites all sorts of fudging, if not actual deception. The only way average speed has any kind of meaning is if moving time is equal to elapsed time. Only then can one claim true average speed.

I am temporarily located in an area where I can, and sometimes do, ride for 2 hours without stopping. In those cases I am moving the entire elapsed time of a ride. I don't unclip, and no, I can't do a track stand. Thus, I have been able to calculate true average speed on these rides.

Get some popcorn and discuss..

That's only one reason it is a worthless statistic. There is also grade and wind to consider.

Average speed over a specific segment is a, somewhat, useful benchmark.

I disagree that average speed while moving is not useful. Unless we are talking about true time trialing where it matters to someone else what your results are, what difference does it make if you stop in the middle of a ride or are held up by red lights or a train. Those kinds of stops aren't part of real competitive riding anyway, so having to include them in your make believe results is not relevant.

Suggestions for an inexpensive alternative?

I think average speed is generally fine if you're doing a loop so net elevation change is 0. As you go up an incline slowly it takes a while, but descending will make up for speed. I notice my average speed is usually around 17mph no matter if I do a flat ride or a hilly ride. Headwinds are a big issue if the wind direction changes during the ride. I did a loop and the wind stayed the same, so at 180 watts one way I was going 14mph but on the way back I was doing 23mph with the same effort. Again, around 17 or 18mph average.

How about a pencil? Is that cheap enough? Just divide your total elapsed time into your mileage. Wristwatch anyone? Voila! Whoops, sorry I meant, wahlah!

The difference is that time stopped is recovery time. And how long one stops for can make a huge difference. Are you stopped for 30 seconds or 30 minutes? Useless, I say..

I agree, as long as MT = ET :D

It is only useless to you if you think so. I find it quite useful indeed as do many, many other folks. Recovery, BS. I am not racing, and I am not training. I'm riding for fitness and checking to see that today's ride was in line with my history. I would rather not have random road events color that judgement. Anytime I have claimed a personal best, however, it has been by dividing distance by total elapsed time such as my solo 50 miles in 2 hours, 27 minutes record. But I don't care about that very much anymore.

Time stopped at traffic lights might be recovery time, but it also takes some time to get back to cruising speed, and that acceleration uses up more energy than just riding without having to stop for the light.

Average speed is really only useful when comparing rides on the same route, and with similar weather. Even then it's not completely accurate.

I'm 36 years old. Wonder how many miles I've accumulated...

If you're going to count time spent in the bathroom as part of average speed, why not include sleeping and going to the doctor, too?

But the downhills on a zero net elevation loop can never make up the time lost on uphills, not without putting out greater power. So your total time will always be greater on a 20 mile hilly ride than it would be on a 20 mile flat ride, given the same power, and your average speed lower. Maybe you're more disciplined on your hilly ride, and don't waste power for example.

Back to OP, I can understand wanting to include stopped time because the effort with a lot of resting isn't really the same as when you never stop. But the time you're waiting at a light doesn't say anything about your moving speed either - so I can't think of one as "better" or more accurate than the other. Just two different metrics.

You may be right. But.. if I were trying to conserve energy, I'd gladly accept the price of a 10 second accelleration if I got 90 seconds rest for it.

If you can't recover while riding the bike you're doing it wrong.

I don't think you really conserve any energy because you have to expend energy to get back up to speed, and you are traveling slower while stopping and accelerating which lowers your average speed.

Until power meters can be had by all, avg speed is a good enough metric for most. I'm sure many like me follow particular routes and can see patterns of improvement over time. I, for example, can expect to maintain an avg of about 18mph over rolling terrain, and I found that to be true in the area(s) I ride near by house and yesterday during a century in an area I've never ridden before. So I think avg speed is reasonable when you look at the patterns, like routinely having 16mph over a loop to then 17 and then 18. Bottom line is, people want to see improvement and it's motivating to see numbers go up. If they want to use avg speed, let them and don't be a bunch of killjoys lol

Again, I won't disagree with you. Surely though, it gives one a chance to allow the HR to drop off. And sometimes that's exactly what I need.

Then next time you ride stop every mile and wait for 15-30 seconds then take off again and get back to cruising speed quickly. And for some add in some slowing down like you are coming to a red light, and then pretend like it turned green so take off again. Then you will know for sure which one you prefer.

You're welcome.

And this is one of the ways one manipulates average speed. By hammering to a screeching halt, resting, and then accellerating like a jack rabbit, one can do a good job of manipulating the numbers. Not that I've ever done that... :D

I obsessed about this kind of thing for quite a while when I started tracking my rides; ultimately I came to the conclusion that ride data of this type is only useful in large aggregate, compared like-to-like on specific segments short enough to limit error from external factors - and even then, it's only useful as a general guide to your overall performance, not as a competitive rating. I think the kind of analysis VeloViewer performs on your data is a fairly good implementation of this principle.

If you want a competitive metric, the only honest option is to actually compete. I like to think I'm a moderately heavy hitter in the local Strava rankings, but actually coming out and racing timed events has shown me that I've got an awful lot to learn and to develop athletically before I'd be heavy hitter in real competition.

If you need the HR to drop off you just dial back the effort a little bit - far more efficient; if I actually needed to come to a dead stop for my HR to dip, I'd be concerned.

Manipulating them downwards, yes, but I don't think that too many of us are interested in that.

Define "most"

All the people registered in this forum?
All the people on the planet?
All the people who respond to this post?

My training stats are for my own benefit. If I'm deceiving anyone
It would be myself or ... OMG the strava community.
It means nothing to anybody but me.