What's Cool In Road Cycling

Toolbox: 5km Pacing Strategy

Start “slow” and under control, then get faster as the time trial unrolls. That’s been the mantra for most cyclists and coaches in the sport. While it’s certainly true that hammering out of the gates is a huge physical challenge and blowing sky-high early on in a TT is no fun, can you really afford to not go hard early on in a TT?

I Don’t Wanna Pace in Vain for Your Love
Having been in the sport of cycling since the mid-1980s, I’ve done enough time trials to know that nobody will ever confuse me with Spartacus or Wiggo. Such is my love for TTs that my last solo TT was back in 2007 with my club’s weekly series. Team time trials are another matter though, and my adoration for TTTs probably stem from the whole “misery loves company” clichй!

However, love them or hate them, time trials are a fact of life in the world of cycling. That’s true whether you ever even ride a club or formal time trial. Think about all the different “unofficial” time trials that we ride. Some happen when we test ourselves on our favourite climb, like I did back when I started cycling in Vancouver and knew that a “good” time for me heading up a 3 km stretch of Burnaby Mountain was <10 min. Now that it’s fall, cyclocross fever has nailed me hard. Competing in my first two races of the season so far, it’s become obvious that so many parallels exist between a cross race and a TT. Both disciplines require an immense mental effort to keep yourself right at the point of spontaneous combustion. Both require a steady high effort interspersed with hard bursts and recovery for hills and corners. And both place a huge emphasis on proper pacing of effort, lest you take yourself way over your threshold for too long and blow up. Exit Strategy
One big difference in cross (and crits and mountain biking too) compared to most longer time trials and road races is the emphasis of the first few minutes. Rather than building up gradually, you can be completely off the back or stuck in traffic if you don’t hammer hard out of the gates. In my first cross race this season, I went out fairly hard but under control and ended up mid-pack in the first lap, but picked off a whole bunch of riders near the end of lap 1 and ended up 4th. In my race this past weekend, I went super-hard off the line and flew into the first hill in 3rd, but dropped back to about 12th by the end of the first lap and then began picking my way through the pack again to finish 8th.

Of course, that just gets the scientist in me to start wondering if my eventual placing was an accurate reflection of my pacing strategy in that first lap? Is it beneficial to go out hard right from the start and then ramp down the effort a bit? If so, how hard should you go?

Aisbett et al. 2009
During my July travels to Edith Cowan University in Perth, Australia, I met up with Dr. Chris Abbiss. Besides dropping me like a bad habit on a hilly ride, Chris is doing post-doctoral research there and also at the Australian Institute of Sport on pacing strategy and quantification of training. One study he was involved in and just published in the September 2009 issue of Journal of Sport Sciences highlights the latest attempt to answer the question: “What is the ideal starting strategy for a middle-distance cycling time trial?”

Before delving into the particulars of this study, it’s first important to note the background and other work done in addressing this question. Much of the existing literature has not been able to discern a clear benefit to either starting slower than average or faster than average compared with keeping an even pace throughout. Aisbett argues that this may be due to some limitations of existing research:
• Small sample size in most studies (<10) can weaken the power of statistical analyses. • Many existing studies testing different starting strategies controlled the subject’s power output during the first half of a test, such that there remained too little time left in the time trial to see a significant improvement. • Some studies may have had the subjects exercise too hard during the initial segment, such that they were forced to drop to too low a self-selected power output during the latter segment. Aisbett’s project aimed to minimize some of these limitations by adopting the following protocol: • Rather than a small sample size, 26 trained cyclists were recruited. • Instead of controlling power output for the first half of the time trial, only the first quarter was controlled, with the remainder self-selected by the subject. Thus, only 20-30% of the total work was controlled. • A 5 min time trial was used in the familiarization trials, with the aim to accomplish as much work (in kJ) as possible (i.e. ride as far as possible) in that time. This was used to represent the approximate 4000 m pursuit time for a trained but non-elite rider, and the relatively short time trial also helps to minimize variability based on external (e.g. motivation) factors. • During familiarization trials, subjects averaged 102.7 kJ of work during that time, and each subject’s best familiarization performance was used to determine their “controlled” pace. • For each subject, power output was initially controlled so that a quarter of their total work was performed in either 60, 75, or 90 seconds, equating to a quarter of total work performed in 20, 25, or 30% of the total time allotment. • After the controlled phase, subjects were able to freely choose power output that was pedal rate-dependent (i.e. analogous to riding a fixed gear where power increase comes from a higher cadence rather than altering resistance or shifting gears). • Subjects were given feedback on power output and work performed, but not on elapsed time or any verbal encouragement. • The test was finished when subjects completed the external work that they achieved during the familiarization trial (e.g. an “average” subject would have completed 102.7 kJ of work in all three trials). The Time of Your Life
The primary results of this study are quite simple and clear. Namely, the slow-start condition resulted in the slowest overall time (5:09 min), average power (333 W), and relative percentage of maximal power (98%). This was slower than the even-start (5:04 min, 338 W, 99%), which in turn was slower than the fast-start (4:53 min, 350 W, 103%).

Some of this difference in performance is obvious and inherent in the design of the study, where a difference of 30 s (or 10% of typical finishing time) was “wasted” with the slow-start compared to the fast-start pacing. However, this remains practically relevant in reinforcing the notion that you simply may never make up for lost time that you “give away” early in a race.

The other thing to look at is, even taking away this “wasted time” in the first segment, what was the power profiles like in the remaining time, where subjects were free to self-select their pace. Can you go harder if you take it easier early on?

The answer seems to be yes. The 2nd and 3rd quarters in the slow-start condition had higher power outputs than both the even- and fast-start conditions, and the 4th quarter in the slow-start was also higher in power than the fast-start trial. Therefore, you can go faster later on if you started slightly slower, but it still never made up for that lost time overall.

No differences were observed in the 2nd, 3rd, or 4th quarters between the even- and the fast-start conditions. Therefore, the substantive difference between these two pacing strategies was again the “lost” time with the slower initial portion in the even-start.

All Abstract Theory?
Science is all well and good, but let’s look at this with my own example this past weekend at the Guelph CX. I spent the last 5 of 6 laps on Sunday with Jeff Ker in my sights and less than 10 s in front of me the whole time. In 4 of those 5 laps, I took 1 s and sometimes 2 s out of him, with the other lap being identical. However, the big difference was that he did the first lap in 8:01 and I did it in 8:10, with the end result that I finished 3 s behind at the end. Overall, of the top 10, I was tied for 8th in my first lap time (fastest lap 1 was 7:52). What would have happened if I trained myself to sustain a faster first lap? Considering that I placed 8th and only 59 s separated 2nd from 9th, the importance of those lost seconds early on becomes huge.

In summary, this study has provided a stronger statistical approach to answering the question of starting strategy by testing a much larger number of subjects than previous studies. In addition, I feel that the switch to controlling only 25% of the time trial made the study more capable of finding real differences.

In the end, it looks like my big focus should be on training myself to be able to sustain a much faster first lap while still maintaining and hopefully also improving my subsequent lap times.

Have fun this cross season!
Aisbett, B., P. Le Rossignol, G.K. McConell, C.R. Abbiss, and R. Snow. Effects of starting strategy on 5-min cycling time-trial performance. J. Sports Sci. 27:1201-9, 2009.

About Stephen:
Stephen Cheung is a Canada Research Chair at Brock University, with a research specialization in the effects of thermal stress on human physiology and performance. He can be reached for comments at [email protected] .

Like PEZ? Why not subscribe to our weekly newsletter to receive updates and reminders on what's cool in road cycling?

Comments are closed.