Determining FTP - Test # 2

Leo's trail.

The trail where I walk Leo is pretty easy to spot in daylight. Surprisingly, the trail is also not all that hard to spot in the dead of night. What helps to differentiate the trail at night are the different shades of grey that mark the trail from the rest of the ground. A similar analogy can be drawn in determining Functional Threshold Power (FTP).

A problem with using test results to determine FTP are the many variables that can affect the outcome. How "fresh" one is on the day of the test, how motivated, the test environment itself, whether it was done outside or on the trainer, whether there was adequate cooling, all affect the results of the test. Comparing test results is even more difficult when different test protocols are used across tests. So, even though two FTP tests may provide two distinct FTP estimations, sometimes recognition of different shades of grey is required to hone in on a realistic FTP value.

Case in point, here are the results from 2 tests of FTP:

Date	5 min	20 min	FTP*
Nov 7, 2011	309W	269W	255W
Dec 14-16, 2011	320W	287W	276W

* estimated

The test results shown above suggest an increase in FTP of 21 Watts in six weeks. While good for the ego, common sense suggests such an increase seems somewhat optimistic. A more likely explanation is that the first and second tests under-estimated and over-estimated FTP, respectively.

There is good reason to think that Test #1 underestimated FTP. The test followed a protocol prescribed in Training and Racing With a Power Meter, shown in the table below:

warm up

5 min all-out

10 min RI

2x1 min all-out (5 min RI)

3x20 sec all-out (3 min RI)

10 min RI

20 min FTP test

cool down

As you can see from the workout above, by the time the 20 min test is done,  the legs have been considerably "softened". In addition, that test was my first effort on the Trainer since last spring; typically indoor power tends to be lower than outdoor power at least initially (see Alex Simmon's blog on this). In this light, common sense suggests this estimate of FTP (95% of 20 min power) was probably low. I'm guessing about 5 Watts low.

Conversely, there is good reason to think that Test #2 over-estimated FTP. This test used a different protocol from Test #1 requiring a 5 min and 20 min test done on separate days and using a Monod Critical Power calculator to estimate FTP. The Monod Critical Power uses two interval readings to extrapolate one's power curve and, according to Alex Simmons, is <snip> "essentially equivalent to FTP (or at least a very good estimate of FTP)" <snip>. A pre-condition for using the Monod Critical Power estimate is that the tested intervals measure one's "best effort" at that interval. For this reason, it is important to test them on separate days so that the first effort doesn't interfere with the second effort.

According to the Monod Critical Power formula, my 5 min and 20 min test results of 320 Watts and 287 Watts, respectively, yield a critical power of 276 Watts. However, the Monod Critical Power formula falls down if either of the tests fails to measure one's true potential at that interval. For example, plugging a lower 5 minute value into the formula actually increases critical power while using a higher 5 minute value lowers it. In my case, I suspect that my 5 minute test was on the low side. Most of my training sessions have been 25 minute intervals; my 5 minute power has been comparatively untrained. Also, my legs felt still somewhat fatigued from my Egg Nog Jog effort. In this light, I'm not convinced that the Monod Critical Power calculation is a good estimate of FTP in my case at this time. A more reasonable estimation of 270 Watts makes more sense.

Date	5 min	20 min	FTP*
Nov 7, 2011	309W	269W	260W
Dec 14-16, 2011	320W	287W	270W

* revised estimate

Taken together, the revised estimates of FTP suggest an increase of 10 Watts over the six weeks; a much more reasonable increase. Both tests above provide ballpark estimates of FTP. But accounting for shades of grey such as discussed above provide more common-sense estimates.