Originally Posted by
Bacciagalupe
Wait, personal experience trumps empirical scientific findings? Why didn't I get the memo?
This is not based on statistical surveys or conjecture. What they've done is put a power meter into the actual pedal itself, so these conclusions are drawn from
direct observation of the actual forces applied during pedal strokes. Not only do these types of power meters measure when, where and in what direction the forces are applied, it also measures each leg independently. These studies pulled from a diverse pool of cyclists, including professional cyclists, MTB, track etc. The author of this article started these tests in 1996 and routinely works with a wide range of athletes, including elite riders.
The data gathered by this method typically looks like this, courtesy of Metrigear:
("Vector" is the name of their upcoming commercialized pedal-based power meter.) On the bottom graph, the gray (flat) line in the center is an average of the forces applied during the pedal rotation. Note how the forces applied drop to zero or less on the upstrokes -- and how the result is a consistent application of force.
Other data collections, by the way, point out how most of the forces applied during the stroke are
tangential. If my understanding is correct: On the down stroke, you are not successfully putting every ounce of force into pushing the pedal down; you're actually pushing the spindle forward as well, which reduces the efficiency of the pedal stroke. So when you are pulling hard on the upstroke, most of that time you are actually just pulling the spindle
back rather than
up. (This shouldn't be too surprising, since the pedals use a radial motion rather than an up-and-down pumping action.)
Subjective assertions and anecdotes may be persuasive for some people. Ultimately though, without an objective measuring device, we humans have no better idea what our legs are doing during a pedal stroke than, say, whether we are working aerobically or anaerobically. There is now an empirical, objective, direct, repeatable and testable means to measure exactly how much force is applied to the pedal itself, and it's far more accurate than guessing what your legs do or watching videos of pros.
On a positive note, the emphasis is still on developing a smooth stroke, so you haven't wound up doing anything counter-productive. It just turns out that your underlying assumptions about what your legs are actually doing are incorrect, but you've wound up doing what you need to do anyway.
I don't have to guess what my legs are doing for three reasons - one, I do a lot of one-legged pedalling on my rollers and know exactly what it feels like to keep a steady rim rpm; two, when pedalling with both legs I can feel the muscles contract, driving the pedals; and three, as I said, I can pedal with hams alone so I
know I'm not powering on the downstroke.
The funny thing about your second graph is that the black, "average" line is obviously software generated and does average the power output, but over a much longer term than the individual pedal meters. Looking at the graph, one can see that both pedals drop to zero power at the same time, yet the black line doesn't move. It should also drop to zero power if it's showing what you think it's showing. There is simply no other way to interpret the second graph.
My pedalling looks exactly like the second graph when I stand, but when I sit it looks like the graph I created for my post. I'm definitely not alone in my pedalling style! I ride with many people who can pedal circles. It's easy to tell: the upper body never moves on a rider who is putting a constant torque on the bottom bracket. Riders who bob or move from side to side under heavy effort are applying torque as shown by your graph. Something to do with Newton's laws of motion, I believe. I can climb at cadences down to about 45 without moving my upper body. Below that, I have to start pushing down hard and do bob.
The first graph is a clear demonstration of what I'm saying! Look at that cadence bounce with the change of pedal angle! When I pedal rollers, i.e. almost no averaging from momentum, the rollers make a constant noise. My cadence does not vary with pedal angle, which is what I'm talking about. So when riding in the saddle, no momentary accelerations.
Your "scientific" graphs, while undoubtedly accurate, are based on the pedaling of an unknown person, of unknown ability. We don't even know if they were sitting or standing! It's also unknown how much better the subject might be if they pedaled properly. The wattage they are producing is that of a fairly strong pro at LT. If they are not a such a pro, then they are producing wattage far exceeding their normal output, and it would be normal for their pedaling style to deteriorate. Mine probably would too, though I'm fairly good at 200 watts.