I think duration has a lot to do with this discussion. Certainly at TT durations, stomping is more effective, One can test this with an HRM on rollers very simply. Pedal at a given speed, say 20 on resistance rollers, stomping only. Then pedal the same speed with smooth circles. Your HR will go up pedaling circles. Which doesn't mean stomping is better than circles, but rather that it depends on context.

But out on the road, pedal according to context. Sometimes I'm smooth, sometimes I stomp.

Again, pedal at 55 cadence at that same speed, by whatever method. Then, using the same pedaling method, increase cadence to 100. Your HR will go up, which doesn't mean that 55 is a more effective cadence than 100. Again it depends on context.

It's very observable that fast LD racers and randonneurs pedal high cadence and pedal circles, Watching a local RAAM finisher pedal, neither his back or hips move at all. Just his legs go 'round, and quickly. I think if Grand Tour stages were as long as they used to be, one wouldn't see so much bobbing during hard efforts.

Be all that as it may, I've found resistance rollers provide excellent training and have made a difference on the road. Rollers without resistance are fine for form and recovery, but aren't as versatile as rollers sets which have resistance.

But out on the road, pedal according to context. Sometimes I'm smooth, sometimes I stomp. Sometimes I pedal fast, other times more slowly.

No.
There are a number of ways energy can be expended on the bike without getting it to the wheel. The most extreme example maybe something like jumping up and down while doing a track stand. I'm sure you can imagine a number of other ways our feet can apply force to the pedal where the resistance comes from the structure of the divetrain itself rather than the road or the air. I'd suggest that rollers are inherently unforgiving of the kind of pedaling form that wastes energy and upsets equilibrium that way.
I don't know anything about the absolute magnitude of the energy losses inherent in poor form, but riding rollers teaches one that they are not negligible.

That sounds like a bit of an unquantifiable stretch.

Yep. For about four years (during winter). Rollers are nice if you have one bike and switch between indoors and outdoors quite frequently.

Riding on rollers made absolutely zero difference to riding outside. First couple of times I'd do rollers each winter, I'd be a little wobbly initially. Went away quickly.

Again, nothing crossing over to the road, though, which makes sense. None of the skills necessary for navigating corners, adjusting lines, negotiating close-quarters contact, etc., are replicated on rollers. Pedaling, as I said, is a non-sequitor to begin with.

Edited to add: I used to be a big believer in the "pedaling circles/smoothness" stuff. Even had a pair of powercranks for a winter! Then I finally started using actual data instead of anecdotal tales. Saved me a lot of time.

I thought this was interesting...
Interesting that my experience with rollers is so different than yours.

This is my experience as well

I've been doing two sessions/week spinning as fast as I can for 1.5 to 2 hours per session. There were times where I spent an hour above 95 RPM and I would try to hold 100 RPM for as long as I can.

It has resulted in higher cadence on the road and cadence is less likely to fall off on sustained 1% to 3% grades. I'm better able to maintain rapid leg turnover for long periods of time.

It has trained core muscles slightly - able to hold still rather than rocking back and forth as much. This lets my arms, shoulders and neck relax.

The effects could be felt after only four or five sessions and it has become a regular part of my routine.


-Tim-

What are you referring to? Forces applied to the pedals that are resisted only by the structure of the divetrain and not transferred to the wheel? It may be difficult to measure directly, but it's all very quantifiable. And I hope it is obvious - that it not a stretch to say - that stomping or applying any downward force on a pedal that is already at the 6:00 position (or the rest of the way around until it gets past 12:00 again on a typical road bike) will get exactly zero energy to the wheel.

In other words, stomping may be effective, but stomping through the very bottom of the stroke -stomping so forcefully that one bounces up and down - is a huge waste of energy.

It seems to me that what really works best is not so much stomping as a pulsating kind of mashing.

Stuff like this is what I'm referring to. You're going to such an extreme to make a point that it's completely beyond the scope of what is actually being discussed in the first place, i.e., that rollers are an effective tool to improving "smoothness" (whatever that is) on the road.

Nothing about any of that addresses the assertion that the "smoothness" that rollers supposedly create transfers or is useful for the road.

How did your experience transfer to actually riding on the road?

And, did such experience you think you gained come only as a result of being on the rollers?

Not a stretch at all, and it is quantifiable. You used a rather silly example, but in the real world, any force applied to the pedal which is not at right angles to the crankarm is wasted in proportion to the vector analysis of said force. Hence "pedaling circles," which doesn't mean pulling up on the backstroke but rather to applying force normal to the crankarm. Which is not really very simple, takes a lot of practice, and rollers do encourage such practice.

It is quantifiable by some pedal-based powermeters.

If you are talking about max power in a sprint than maybe but if you are suggesting average power over an entire race leg than you are totally wrong. Or maybe you sprint during an entire race, if you do than I apologize. I would like to add that it is not just about turning circles it is about learning to control your power letting up at the bottom of the stroke and applying power at the top becoming as smooth as possible.

I just happened to think of a qualitative test to see if rollers might help your riding: Get on your bike, put it in a low gear, spin it up to 150 rpm and hold it there for a few seconds. If you can do that without any bouncing, you either already have rollers, are a track cyclist, or in any case don't need them. If you can't, or bounce, rollers might be in your future.

This little test tells you if you are applying force normal to the crankarms. If you are, you simply don't bounce. If you aren't, you bounce.

Anecdote: back when I was stronger, maybe 15 years ago or so, I used to lead a mixed group of geared, SS, and fixie bikes. On my SS bike I could hold my place in the line up to exactly 137 cadence, my aerobic limit, which must have been ~25 mph. I couldn't pull that fast, only ~105 cadence on the front, weakling that I am. But I didn't bounce. Rollers.

An excellent roller drill is to pedal steadily at 115-120 cadence for quite a long period without a break, say 15-45 minutes. I do these once a week. Must have done nearly 1000 of them. That doesn't sound too bad, but the trick is to do it in HR zone 2, i.e. not much excess energy expenditure from the fast pedaling. If you have power or watch your speed, that'll track improvement at the same HR.

So then you'd say that trainer, Zwift, etc. also "Doesn't really have much to do with the road, though"?

I don't see how you get that rollers don't help with holding a line, close quarters riding.

Personal anecdotes (see sig)

Shortly after being introduced to rollers (a few weeks) a good friend noticed that I kept a much better line on the road.

I had been using rollers during the winter months to keep a minimum fitness level.

For years I would do 40-45 minutes at about 90rpm with higher cadence intervals.

Recently I decided to let a bunch of air out of the tires and up the workout to 1 hour. I kept the same avg cadence and still did the high cadence intervals. This recent change has me noticing more strength on the road and ability to sustain harder efforts. When the weather allows for a more regular outdoor riding schedule I'll keep closer track of my numbers, and just observe how well I do when riding with the stronger groups.

Anecdotal, yes.

Here is how you do it! I love my rollers.