Is there any way we could get a Quadrant analysis chart.
I know that i can’t get CPV vs average torque, but we could plot (as a scatter chart) torque vs cadence.
(BTW i am not an expert or anything like that, but i know enough to be dangerous, but maybe not useful)
Vince
Hi David,
We all know that you’re extremely busy, and was wondering if the quadrant analysis will be a feature sooner than later, please?
For those that don’t understand it,:
I copy the formulas from wko5 in case it can help in its implementation.
Q1
sum(if(cpv >= tcadencecranklengthpi/30000 and aepf >= bikeftp/tcadence/pi/cranklength*30000, deltatime, 0))/sum(if(aepf>=0 and cpv>=0, deltatime))
Q2
sum(if(cpv < tcadencecranklengthpi/30000 and aepf >= bikeftp/tcadence/pi/cranklength*30000, deltatime, 0))/sum(if(aepf>=0 and cpv>=0, deltatime))
Q3
sum(if(cpv < tcadencecranklengthpi/30000 and aepf < bikeftp/tcadence/pi/cranklength*30000, deltatime, 0))/sum(if(aepf>=0 and cpv>=0, deltatime))
Q4
sum(if(cpv >= tcadencecranklengthpi/30000 and aepf < bikeftp/tcadence/pi/cranklength*30000, deltatime, 0))/sum(if(aepf>=0 and cpv>=0, deltatime))
Not to be a PITA but it strikes me as something the folks at Training Peaks/WKO might not be so keen on being copied since it is proprietary. Just sayin’
Hi!
I don’t know if there’s still much interest in quadrant analysis, but I decided to build one anyway. It’s not perfect yet, but it gets the job done — and it is both in light and dark mode.
Instead of creating two charts, you could incorporate the option to choose dark or light in a config setting. I have done that with some of my charts.
I also tried to implement that, but for some reason I couldn’t get it to work properly. The button for switching between light and dark mode only changed the background and axis colors of the graph, while the rest of the colors stayed the same. I’ll keep experimenting with the code a bit more.
I’ll send you the revised code by DM.
I’ve made a few other changes, to remove the need of hard-coded settings, like crank length, FTP, max power, etc. I just want to go through everything before sending to you. It’s a great looking chart, and hope to help you make it better.
Thanks Gerald, I really appreciate it.
I’ve never understood how the quadrants are chosen. Can you clarify? I asked Andy Coggan about that back in 2003. I can understand how one might identify quadrants for a single ride but I don’t understand how the quadrants defined by one ride should apply to another ride.
You should only compare one ride to another when the session objective/s is/are similar. You could have average power, say 200W. 200W steady for an hour is similar to 300W and 100W for 30 minutes each. Same result, but different activities.
As for the quadrant analysis, at a quick view, you can see one of four things on how you produce power. Power is calculated, not measured. It comes from knowing how much force you apply to the pedals, and how quickly you move the pedal/foot around in a circular motion (circumferential pedal velocity). Along with crank length, you calculate your power.
So the four quadrants are:
Hunter Allen explains it in this video. I first read about it in Training and Racing with a Power Meter, so figured he’s better off explaining it.
Ah, yes. That video illustrates the point I made to Andy in 2003. Others had been plotting cadence vs. crank torque (or pedal speed vs. pedal force) for decades, including me. In fact, if you go back to the Topica Wattage archives, I had posted a cadence-crank torque plot and then Andy responded by posting the first “QA plot.” Andy’s contribution was FTP, so rather than just do a cadence-torque plot, he introduced a horizontal and vertical line. As Hunter says, the vertical line was set at a pedal speed equivalent to 90 rpm, and the horizontal line was set to where the vertical line intersects with FTP. But That puzzled me in 2003, and it still puzzles me, because my “preferred cadence” has never been 90 rpm. I had been examining cadence and crank torque for a while, and I knew that preferred cadence is a dependent variable, not an independent one, so fixing the vertical line at a particular cadence across many rides seemed puzzling to me.
Excellent and exciting question @RChung (just like in the old days!)
For years I simply assumed this was an arbitrary choice due to the lack of a better option. I would love to have better tools (and knowledge) to analyze torque, because perhaps the horizontal line OR the “FTP line” should actually relate to critical torque (or some fraction of it).
Cadence, being such a multifactorial variable, would probably also need to be represented multidimensionally (considering intensity, gradient, and fatigue), and I can hardly imagine how that could even be visualized in a chart.
Anyway, the main (and almost the only) practical use I made of this chart was to assess the execution of torque workouts. For that purpose, I don’t even need the quadrants divisions. In any case, in Intervals I prefer to do that simply using a torque histogram.
@Gabriel_Vargas We know that, among other things, preferred cadence varies with power, gradient, and fatigue: that’s why it has always kind of bothered me that riders think that they can find their “optimal” cadence on a trainer indoors, and then transfer that to their riding outdoors, as if cadence were a magic independent variable.
We can see watts on the cadence-torque (or pedal speed-pedal force) plot–you don’t need quadrant lines to see that, you only need additional isoquant lines “parallel” to the FTP line. So years ago I used to color code the points according to gradient (which I calculated from speed and power). Although I didn’t do this, color coding could also be used to identify broad categories of a ride, like first third, middle third, and last third to assess how your preferred cadence and torque vary with fatigue.
I’ve also zoomed into specific sections of a ride, like the sprints, or a particular hill. In sprint zooms, I’m usually looking only at maybe 30 seconds before the sprint and the sprint itself, so perhaps 40 seconds in total. In that case, I number each dot from 1 to 40 and display that on cadence-torque axes. That lets me see how quickly I was able to amp up my power. That’s what I call a “power expansion path.” If the PEP is very vertical, I amped up torque faster than cadence; if the PEP is relatively flat, I amped up cadence faster than torque; if the PEP is relatively diagonal, I increased each of them relatively evenly. I’ve noticed that my sprints attain higher power if my legs are already going fast and I increase the torque a lot, very quickly. Occasionally I also look at power contraction paths, but what I learned from them is that I was letting go right at the top of a hill and losing acceleration: I should be pedaling through the top of the hill and past it before easing off.
