Wednesday, December 21, 2011
Monday, December 19, 2011
More Data, More Doubt
Greg Lemond, Taylor Phinney and I have the same problem - long femurs. Very few slack seat tube bikes out there, so we end up running wild setback plus long-railed saddles.
I've always been curious about my pedalling dynamics. By nature a masher, but force myself to practice high cadence frequently. Right leg definitely a bit gimpy, left leg always feels perfect. One of the freaks with valgus (bow-legged) pigeon-toed feet. I've never felt fully comfortable. Always too far back, but with a couple of foot wedges, no-float cleats and a well designed insole I've been able to eliminate knee problems successfully.
I work hard on spinning, but when it's time to make a bridge or drop someone, nothing for it but to press on the pedals as hard as possible on the downstroke. That's what works for me.
Finally got around to getting the Computrainer SpinScan program going at the club training centre. Excited to find out what I could.
Some interesting data, most of it subject to wild misinterpretation.
Spinscan measures the torque on the wheel as you pedal. Cool thing is that it does it at a high enough rate that you can see how the torque changes with crank angle, and you can relate it to left and right cranks.
Right - Left power balance:
At low - medium power output (up to 50% of threshold), the right foot was putting out 60-65% of the power.
At higher output, 70-80% of threshold, the balance was better - 55%-45% being typical.
At higher outputs, threshold, VO2max and above, it hit 52-48.
Standing, the balance was 50-50. When standing, one's body tends to find the optimum position. Surprised that it was so even.
Dunno why the difference was so large at lower power. One should always expect a difference, legs are never the same size, but this was worrying.
Just as surprising that the balance was so even at higher output. That's reassuring.
Higher outputs are where it matters, so it looks like both legs are in good shape for that. Unlikely that I'm throwing away much energy.
I experimented with moving fore and aft on the saddle. Didn't make much difference. Higher cadence equalized things a bit. Moving a little bit to the left or right on the saddle could change the balance as well, but not in a consistent repeatable way. Plenty of room to experiment with shims and cleat position.
You spend a lot of time in road races just spinning lower power. With this imbalance, it begs the question: am I going to over-fatigue my right leg? Not going to worry much about it. I think that the motion of the bike on the road will smear out the differences, and fatigue is mostly induce by the hard efforts. No idea how accurate the spinscan is at lower power either.
Pedalling efficiency - Spinscan number:
The software also supplies a Spinscan number related to pedalling efficiency. It measure the circularity of your pedal stroke; that is, how smoothly you apply force over the cranks rotation - 100% is a perfect circle.
There is zero evidence to show that a circular stroke is any better for cycling. While it's important to utilise all your leg muscles, a glance at the figure at the top of this post shows that it's the quads and glutes that supply the big power and they contribute most near the middle of the downstroke.
The CT manual states that road riders tend to have most circular strokes, with scores in the 70-85 range, and MTBers are much more jerky, but the best known studies of Coyle and Broker (summarised in High-Tech Cycling, Ed. E. Burke) show the opposite. The best cyclists tend to mash hard around the 9 O'clock angle and apply barely any force on the upstroke, while the best MTBers have the smoothest stroke - to retain traction on climbs. The CT manual recommends to use spinscan to help fill in the dead spots at top and bottom of the pedal stroke. Sounds plausible, but it might be better to work on just mashing in-between.
Results for me were all in the low 70s. Typically 72% at 80 rpm and rising to about 75% at high cadence. Not much dependence on power or fore-aft position. Surprisingly, even out of the saddle, the 'efficiency' didn't change.
So this identifies me as a masher, whether that's good or bad I don't know. I suspect it's strongly dependent on femur length, foot length and cleat placement as well. Not much to read into that. There's zero published work out there on spinscan studies and any google search will only lead you to someone trying to sell you something.
I strongly suspect that training to improve pedalling by raising the spinscan number, unless one has a really low score or is new to cycling, is a waste of time and could be harmful.
An interesting thing that I also noticed is that my left leg was consistently a couple % more 'efficient' than my right. My left has always felt much smoother, the mechanics of my right has always felt a bit off - twenty years of soccer and other right-foot dominant sports tend to mess things up. It might also be because my left foot is bigger and fits the shoe better.
Average Torque Angle (ATA):
ATA is difficult to explain, the manual does a lousy job of explaining it. But it's something to do with where on the pedalling circle you're producing best torque. It's stated that 90 degrees is optimum. No justification is given.
My ATA values were close to 100, hitting 95 at highest cadence, little dependence on standing or fore-aft.
I suspect that ATA is again strongly dependent on femur length and leg size. Again, probably harmful to try to train to get it close to 90 degrees.
If ATA is way off, then it's a sign that you fit is way off or you have a serious leg length or muscle discrepancy - see a physical therapist if this is so.
This data is for my training bike, with worn pedals and cleats and a less than perfect fit. I'll bring in my road bike (which fits me well) and my cross bike (which does not fit me well) and compare. If there are significant differences then it may be worth making some fit adjustments to the training and cross bikes.
The quote at the top is telling. Spinscan is an amazing program that shows a lot about your pedalling almost instantly.
Given the lack of any published data or studies to support any of the manual's assertions, I think it's dangerous to try to change your pedalling in any way based on the numbers.
If the numbers are well off the normal range, then you're either new to cycling or have some physical problems that should be looked at by a specialist.
Perhaps a better use of spinscan would be to check your pedalling style periodically and track any changes in the numbers. This could indicate changes in fitness or bike fit. It would also be useful to compare styles on your different bikes to identify fit problems compared to your favourite bike.
My thoughts:
1) There is no evidence to show that a more circular pedal stroke is more efficient. If your spinscan score is in the 80s or higher, there is a good chance that you're not working your major muscle groups enough. If you're above 90 or below 70 you're either a World champ or not trying.
2) A suggested protocol for spinscan testing
Warm up for 20 mins
Observe spinscans at 70, 80, 80 and 100 rpm cadences, each for low, medium and high power. Make note of left-right balance, spinscan #, l-r spinscan differences and ATA.
Check for sprints and out of the saddle efforts.
See if there's any change for fore-aft saddle position.
Repeat every 6 months. Make note of any changes. Could be indicative of a change in fit.
3) If you have knee pain, a spin scan may indicate a change in pedalling dynamics. Comparison to previous results might confirm this
4) Pedalling indoors is not the same as outdoors. Imbalances are likely to be smeared out when outside.
5) Training to optimise the power balance, efficiency or ATA is not advisable.
6) There is very little data out there to support any assertions that spinscan can help pedalling. A few hours on the internets will show you that the large majority of sites with spinscan info are trying to sell you a service with no data to back it up.
7) The numbers are dependent on Femur length, Femur/tibia ratio, cleat placement and muscle composition.
8) One leg will be dominant. Expect slight imbalances.
9) The numbers are most to be trusted at higher powers, near threshold and VO2max power.
10) Angle of attack is subject to an offset depending on how you set up the magnet and detector. Don't trust the absolute value. L/R differences may be more significant.
A few hours on the internets revealed a multitude of pages devoted to selling you spinscan fit services; and a multitude of badly-informed fan pages.
CT spinscan manual. See pages 24-26 and p.35
CT Manual
Some well-informed links:
Good comments. Doubt if any performance boost is real though,
Leelikes bikes
TdF winner (Ulrich?) complains of Spring knee. Docs tell him to quit mashing until pain goes away, then they write a paper about it. TDF winner has good l/r balance but lousy spinscna # - should be kicked out of triathlon class. TdF winner's knee
A (surprisingly) informed (and cynical) Slowtwitch discussion
My feeling about spinscan is that it's a fun toy. It's most useful for tracking your numbers over time and comparing fits. Any gross imbalances should be given attention as to whether they reveal muscle or fit problems or not.
That's it.
(major editing and results on other bikes to come)
Subscribe to:
Posts (Atom)
