by Andrew R. Coggan, Ph.D. - People have been competing against each other on bicycles since at least 1868, when the Englishman James Moore won a 1.2 km event held in Parc de Saint-Cloud, Paris. The cycle ergometer has been around nearly as long, i.e., ever since 1896, when Elisée Bouny attached a mechanical brake to the rear wheel of an ordinary bicycle that was elevated off the ground to quantify the power output of racing cyclists. It wasn’t until the development of the SRM (and also the Balboa Instruments PowerPacer and the LOOK Max One) in the mid- to late-1980s, however, that it became possible to routinely measure a cyclist’s power when riding outdoors.
This new technology was rapidly adopted by forward-thinking professional cyclists (e.g., Greg Lemond), cycling coaches (e.g., Francesco Conconi, Paul Koechli), sports scientists (e.g., Drs. Jeff Broker, J.T. Kearney, Dave Martin, and Jim Martin), national (e.g., German, Australian, U.S.) federations, and trade teams (e.g., track cycling’s Team EDS). Among the “firsts” that occurred during this time were the first use of power meters by an entire team during a stage race (by the U.S. National team at the Tour du Pont in 1994), the first use of a power meter during a mountain bike race (at a test event held on the Atlanta Olympic mountain bike course in 1995), and the first use of power data to characterize the demands of the individual and team pursuit (by the German cycling federation and also as part of USA Cycling’s Project 96). The Australian Institute of Sport also collected extensive power data on their athletes in training and in competition from the early- to mid-1990s onward and especially during the lead-up to the Sydney Olympics in 2000. Thus, by the end of the previous century power meter use had already gained significant popularity at the elite level. The high cost of the SRM, however, prevented its use from spreading much beyond these circles. Furthermore, because those with the most experience using power meters were in the business of winning races, not freely sharing their hard-earned knowledge, at the time there was very little, if any, publicly-available information about how to best utilize power data when training cyclists.
The above situation began to change rather rapidly after the Tune Corporation introduced the original PowerTap hub in 1999. Prior to its introduction, they loaned one to the well-known cycling coach and author Joe Friel for a few months, who based on his experience produced a small manual called simply Training with Power that was included with each new PowerTap. In it he explained the similarities and differences between training with a power meter and training with a heart rate monitor, and laid out a system by which a cyclists’s efforts over various durations were referenced to their actual performance ability, i.e., their power output. To my knowledge, this was the first widely-distributed resource that not only specifically addressed the question of training with a power meter, but also “elevated” power data to a position of primacy over heart rate – previously, power data had been used primarily as a way of quantifying the demands of various events (especially on the track), as a way of controlling the intensity of intervals, and/or as a way of placing heart rate data (which was still considered the primary metric) in better context. However, a comprehensive approach to utilizing power meter data had yet to be developed, or at least shared with the world at large.
The introduction of a less-expensive alternative to the SRM helped indirectly foster another important development, which was the creation, in the summer of 2001, of the “wattage” online discussion forum, which was initially hosted on Topica.com and then later moved to Google Groups. The brainchild of Kwan Low, this forum quickly became an important nexus where power meter users could share their ideas, experiences, etc. This included yours truly, who first used an SRM for a few months in 1996, then became a pilot user for PowerTap in 1999. This experience, coupled with the free-flow of information on the wattage list and a chance conversation with a cycling coach in the infield of the velodrome in Trexlertown, PA, stimulated me to try to develop a logical approach to power-based training that was grounded in sound physiological principles. Initially, this merely consisted of a series of power-based training levels anchored to what I decided to call “functional threshold power”, which I first described on the wattage list in the fall of 2001. This was followed, however, by other analytical tools, i.e., by normalized power, intensity factor, training stress score, power profiling, and quadrant analysis in 2003 and by the Performance Manager approach in 2004 (although not revealed until 2006).
Another seminal event in the very short history of power-based training was a meeting devoted to the topic that was held in conjunction with the US Professional Championship/Liberty Classic in Philadelphia in the summer of 2002. Organized by Sam Callan, the Director of Coaching Education for USA Cycling, the three speakers were Dr. Allen Lim, Dean Golich, and myself. Allen, whose dissertation research at the University of Colorado revolved around power meter use, presented an overview of how he felt power data best fit within the context of preparing cyclists for competition, especially the concept of power as a measure of applied stress and heart rate and perceived exertion as important measures of the resultant physiological strain. He also described his experience using the PowerTap as director and coach of the women’s Celestial Seasonings professional cycling team, including using it to conduct field tests to determine aerodynamic drag. Somewhat similarly, Dean shared his experience, garnered first as a physiologist for USA Cycling during the early 1990s and subsequently as a coach for Carmichael Training Systems, in using SRM data in the preparation of elite athletes such as World Champions Mari Holden and Alison Dunlap. In particular, Dean described how he used power data to target specific energy systems in training, and emphasized the usefulness of power data in determining when an athlete had “overreached” enough to induce improvements in performance. He also advocated the use of block training (i.e., multiple hard days in a row) not only for its effectiveness, but also for its time-efficiency. In doing so, he presaged one of the most important ways in which widespread use of power meters has seemingly changed the way many amateur cyclists tend to train, that is, with less total volume but with more emphasis on specific, structured efforts. Finally, calling upon my academic background I discussed issues such as the limitations of heart rate as a measure of exercise intensity (due to, e.g., cardiac drift), the cardiovascular and metabolic ramifications of the highly variable (“stochastic”) nature of power when cycling outdoors, the average effective pedal force-circumferential pedal velocity relationship and how this relates to fiber type recruitment and fatigue, the Monod/Scherrer critical power paradigm, etc. I finished my talk with a list of the top 10 things that I felt I had learned as a result of training and racing with a power meter, the last three of which were:
Only about 30 people were present in the overheated Philadelphia conference room in which this meeting took place, such that its impact could well have proved to be relatively inconsequential. Among the attendees, however, were Hunter Allen, a professional cyclist-turned-coach, and Kevin Williams, one of his clients and a skilled computer programmer. After listening to the speakers discuss various ways in which power meter data could be manipulated to advantage, they recognized the need for more sophisticated software that would make such calculations easier and more accessible to everyday users, and over lunch that day they quite literally sketched out their plans to produce such a program on the back of a napkin. After further development, Hunter invited me to join the effort, with the final result, a program called CyclingPeaks (now TrainingPeaks WKO+), being released in the fall of 2003. This actually made it the 2nd such aftermarket program for analyzing power meter data (after Paul Koechli’s PowerCoach, which was released in 1996), but Hunter and Kevin’s program achieved much greater popularity due to its features, ease-of-use, lower price, and the fact that it ran under the most popular operating system, i.e., Windows. (A Windows-based alternative to the original SRM DOS software had previously been developed as part of Project 96, but was not sold to the general public.) In turn, the success of CyclingPeaks/WKO+ inspired the development of a number of other quite similar commercial and open-source programs (e.g., RaceDay, Golden Cheetah, SportsTracks. As well, our (i.e., Hunter and my) association led to the development of many of the analytical tools listed two paragraphs above, as well as our co-authoring of the book Training and Racing with a Power Meter, the 1st edition of which was published in 2006 and the 2nd in 2010.
Yet another consequence of the confluence of events represented by the Philadelphia meeting was the decision by Sam Callan to offer a special certification in power-based training to coaches licensed by USA Cycling. Due to the burgeoning popularity of such devices among both coaches and athletes alike, this program quickly grew from its inception in 2005 to where approximately 100 coaches now attend these two day seminars every year.
Today, power meter use has grown to the point that there are approximately a half-dozen such commercial devices on the market, and a number of enlightened professional teams (e.g., Garmin-Transitions, Cervelo TestTeam) equip all of their riders with one and hire consultants who specialize in interpreting the data they provide. Even at the local level, up to a third of the riders in some fields can be seen racing with a power meter. Indeed, in December of 2009 VeloNews listed the growth and development of training with power as the 3rd most significant story of the last decade. Thus, approximately 20 y after Uli Schoberer’s development of the SRM (and a little over 100 y after the invention of the stationary cycle ergometer), the use of power data in the preparation of racing cyclists can truly be said to have become mainstream.
The author would like to thank Sam Callan, Dean Golich, and Dr. Jim Martin for their feedback during the preparation of this article.