Tuesday, December 30, 2008

PowerCranks – The First 6 Months and an Overly Complicated Data Analysis

Recall a few months ago I gave an “11 week report” on my use of PowerCranks. At the time I had invested some 75 hours of cycling into PowerCranks, but most of the training was in my tempo zone (no more than 245 W) or below. With the passing of November and December I have completed two months of more intense training with PowerCranks and brought my time with PCs to over 150 hours (and some additional hours on regular cranks by way of one weekly group ride). These two months had a focus on threshold (zone 4) and VO2 (zone 5) in an attempt to both raise my functional threshold power and prep myself for the upcoming racing season in Texas, which starts in January.

So did things pan out? Did PowerCranking affect my power positively and satisfy the claims of increasing power? Read on and see for yourself.

Coming in to November my power profile looked like the plot below. The black line represents my all-time best power, and the green line is my power in the one year before taking up PowerCranks. The red line is my power through October 2008. The plot is a combination of mean-maximal power and normalized power. Normalized power is used for durations of 15 minutes and up, and mean maximal (or average) is used for shorter durations. As can be seen, my use of PCs did not generate higher power for durations under 2 hours. Only above 2 hours did I see a boost in power. Was this due to PCs?


First off, the difference between the green and red lines is only about 8 watts, which represents roughly a 3 percent increase in power. It should be noted that in September 2008 my small training group, comprised of my closest pals and teammates, increased the distance of our Saturday training ride by over 8 miles. That added an additional 20 or so minutes to our route and initiated some of the bumps we see in the curve above. Most of my pals are incredible time trialists (55-57 minute range at sea-level). We hurt ourselves on our training ride, and it shows with the high normalized power at the longer durations. Personally, I believe the increase in power from 120-180 minutes is due to the change in our training ride rather than the use of PowerCranks. But honestly, the difference is only 3 percent so even if PCs are the reason for the increase, the increase is minimal.

But that was 2 months ago, and a lot can happen in that time. I approached each bike workout with PowerCranks as I would with regular cranks – I didn’t pussyfoot around. I hit my intervals full tilt. With the race season coming soon, I wasn’t about to sabotage my training. I can honestly say I used PowerCranks as hard as I could.

The plot below shows my training stress in terms of cumulative average TSS. The black curve represents a starting point of August 2007, the pink line a start of August 2006, and the red line August 2008. As can be seen from the graph, my TSS is much higher this training year than the previous two years. With this data and my focus on L4 and L5 work, I would expect to be at least as fit as past training years or even more powerful.
If we plot my power from August to October and compare it to the power over the last 2 months, we have the plot below. The good news is the focus on L4 and L5 is paying off. 20 minute power is up about 10 watts (around 4 percent) and 30 minute power is up 2-3%. So this begs the question; is this increase due to PowerCranks or the focus on L4 and L5?
Plotting my power from August to December of this year (my PC use) to past periods, we get the plot below. I’m ahead of my performance for the same time period as last year, but given the difference in TSS from last year (the plot earlier in this status report), that is somewhat expected. Regardless, I’m still well below (well, not that far) from my all-time best power numbers. I have this huge gap between 20 and 80 minutes to fill, and so far neither more intense training nor PowerCranks has filled that hole.
If I look strictly at mean maximal, or average power, things are even more dire. The plot below shows I’m actually less powerful in some respects compared to the same period as last year.
Now let’s examine 20 and 60 minute normalized power in more detail by looking at a yearly breakdown month-by-month. The next two plots show data going back to 2003. My PowerCrank use would be characterized by the red lines in months 8 to 12. The plots are busy, but one thing is clear - I’m not seeing any significant gains in either 20 or 60 minute power. As a time trialist, I wanted a bump in 60 minute power to cut precious seconds from my 40k efforts. My numbers so far can only be explained as my typical seasonal variances. Yes my 20 minute power is at its highest of the year, but by only a couple (as in single digits) of watts. For 60 minute power, my power has been consistent over the last 3 months, but the max over the 3 months is a mere 1 watt higher than February 2008, well before PowerCrank use.

Finally, let’s look at critical power. I’ve calculated critical power in two approaches. In one case, which I’ll call CP15, I’m using 3, 5, 8, 10, and 15 minute average power to calculate my critical power and anaerobic work capacity. The other case is what I call CP10, which is the same as CP15 but without the 15 minute power number. It is well known that the critical power model can be effective at predicting threshold power (i.e., critical power). One caveat is that for this data I did not go out each month and test to maximize power for each range. It is assumed that I’m hitting a maximum through my normal training and racing. My motivation in training is to push as hard as possible, and likewise in mass start racing the physiological demands push riders to the limit. For the most part, the R^2 values were 0.996 or above (indeed, most were three 9s).

The plot below shows a monthly calculation of critical power. Notice the upswing from September to December. That’s the result of focused training. But also note that my highest critical power was in July 2008, which was the month leading up to the Texas State Time Trial Championships.
At this point I was thinking that perhaps my adaptation to PowerCranks was negatively affecting my critical power calculations. I collected the monthly data and cherry picked over a quarterly basis (that is, picked the peak values over a three-month period). The fourth quarter of 2008 would represent the best time period for PC use. The plot below shows my critical power using this quarterly basis. Using a quarterly basis narrows the gap some, but still my summer power is higher, albeit by only a couple of watts.

So when you take in all this data and puke it out, what is it saying? First, my focus on L4 and L5 is showing a rise in threshold power (but heck, the same can be said for my teammates who are also focusing on these aspects). But more importantly it shows that neither regular cranks nor PowerCranks are a magic bullet when it comes to breaking through a plateau. I consider myself a “well-trained” cyclist based on the historical power data I’ve gathered and my ability to rather quickly maximize my aerobic power to its genetic potential. The plot showing my historical 60 minute power certainly demonstrates my sweet spot is in the 265-275 W range. I had one aberration in January 2006 which I attribute to measurement error rather than really nailing 287 W for an hour. To date I have sufficient data at hand that clearly demonstrates that PowerCranks, once adapted to (and given my ability to ride them for 4+ hours at a high pace or complete critical interval sessions as if they were fixed cranks), have not been able to raise my power.

I really, really, really (you can add some more reallys if you want) want to increase my aerobic power, and I was hoping the PowerCranks experiment would given me just a 5% gain. I’m not greedy; I’m not looking for that 40% power increase or 2-3 mph claim. Just give me 5%, or about 12 watts, which would be about 1/3 mph. Alas, no gains. I have 6 months still ahead of me with PowerCranks (and will be riding a ton this week), but honestly with race season approaching my training volume with PowerCranks will begin to taper off due to race weekends on regular cranks.

Is there merit in using PowerCranks? I think there is, particularly for some riders. I feel many riders don’t know how to truly push themselves on the bike. When the going gets tough, too many riders take the easy way out. For me, VO2 intervals are the absolute worst, yet they are effective at helping get the last few watts out. I have learned to suffer during VO2 workouts, and Powercranks are pain and suffering at the beginning as you adapt to them. The perceived effort I was getting with PCs in the beginning was akin to VO2 and anaerobic intervals, but with more volume. If they can teach a rider how to suffer at a new level and open up their eyes to more intense training, they’ve done their job. You can certainly argue you can achieve the same end state with regular cranks, but I personally feel too many riders take the easy way out when the going gets tough. In that respect, the old axiom “no pain, no gain” is correct.

4 comments:

bugno said...

Thank you for such a detailed analysis!

Robert said...

"Is there merit in using PowerCranks? I think there is, particularly for some riders. I feel many riders don’t know how to truly push themselves on the bike."

That's surely true but, OTOH, truly pushing yourself for extended periods on the PCs don't appear to have helped much.

Unknown said...

After reading several PowerCrank forum discussions that have devolved into name calling, it is refreshing to read such a thorough and impartial analysis.

I am disappointed that the PCs did not raise your FTP significantly. Both for you, and because I recently bought a pair on ebay and was hoping they would be worth the time and effort. Oh well.

I think your point about their value in teaching riders to push through their pain threshold is a good one. I suspect that using PCs also help with developing core strength and are an alternative to single legged drills. Important to most racers is that these last two areas can be trained more efficiently and at less cost than PCs.

Thanks again and good luck with your upcoming season.

rmur said...

Thanks for all the info James. Very thoughtful and detailed.

Actually I need to print a couple of the last long posts & charts. I'm old and like paper :-)