From the Coaches: Pacing for Performance

Pacing for Performance

The old adage “It’s not about how you start, but how you finish” is a pretty good summary of the importance of pacing. That being said, there are plenty of nuances to consider in regard to pacing that can’t be summed up in a simple one-line quote. We can think about pacing in many different ways, and I want to talk a bit about a few of the ways and how to go about selecting a pacing strategy depending on what your goals are.

To begin with, pacing can be defined as a technique for spreading out one’s effort over the entire duration of a given task. That’s all well and good, but what does that really mean? First off, I believe that it’s important to emphasize what the goal of pacing is in order to better explain and describe it. For the sake of this article, I will focus on talking about pacing in regard to optimizing performance - or effectively to reduce the time that it takes to cover a certain distance or to maximize performing a certain amount of work in a given time.

Many of you are familiar with the importance of pacing the efforts in Full Frontal where you’re trying to produce the maximum output for each of the four test efforts, which can be summarized as follows:

5s NM Sprint efforts = All out, no pacing. (Note: determining proper gear to start in is important for maximizing power in sprint efforts.)
5 minute MAP effort = Steady pacing. (If in doubt, start very slightly conservatively and increase after the first 2-3 minutes if possible.)
20 minute FTP effort = Steady pacing. (Similar to 5 minute effort, best to start slightly below what you believe you can hold and increase after the first 10-minutes if you’ve got something left in the tank.)
1 minute AC effort = Nearly all out, minimal pacing. (You need to start strong and above what you plan on averaging - but it shouldn’t be an all out sprint to start. You should expect a fade in power over the minute.)

For both the 5-minute and 20-minute efforts, we would typically think of those kind of efforts as a negative split. Meaning that if you were riding a given distance, you’d end up going slightly faster in the 2nd half of the effort compared to the first. In many world records, we’ve seen that the first half of the race took 50.5% of the total record time, with the second half taking just 49.5% of the total time. These are textbook examples of a slight negative split. Do keep in mind, though, that many world records are set with different pacing strategies. For example, during a 100M running race the athletes achieve their peak speed around 40-60 meters into the race and most everyone is actually slowing down in the last 20 or so meters. There are no splits taken in a 100M race, though, so you’d never know that without studying it in detail (like us nerdy sports scientists like to do).

We’ve got a much more in depth look at optimizing your Full Frontal testing, including pacing recommendations, at:

The first thing to think about with regard to pacing is what kind of route are you riding? If the route or effort is relatively flat or consistent and there are minimal changes in wind strength and/or direction during your ride, then you’ve got the simplest version of pacing to think about. In cases like this, the goal is to start off at a level that will allow you to meter out your energy over the entire course leaving you basically empty and fatigued right at the end, just as you arrive at the finish. One way to pace these kinds of efforts is to look at a combination of your power, heart rate, and RPE (rating of perceived exertion) during the effort. If you’ve performed similar efforts in your past, you could review the average power and average heart rate of prior efforts to get an idea of what you’ve done in the past. As an example, if you’re preparing for a 40 kilometer time trial, and the past few times you’ve averaged between 190 and 210 watts with an average heart rate between 145-155 beats/minute in the past - then you would be well served to use those ranges as an initial guide. If your fitness is significantly higher than it was when you performed the previous efforts (as proven by Full Frontal or Half Monty test results, or other maximal or sustained test efforts), then you could consider increasing your target by the percentage of improvement that you’ve observed from your performances at a lower level of fitness.
As an example if you had an FTP of 200W in April, and performed a 40-km ITT in May at 195 watts, but have recently recorded an FTP of 215W in August as you prepare for a September 40-km ITT then you might consider targeting 210W for the September event. The opposite is true as well - if your fitness is lower than previous events, then you’d be better off starting a little more conservatively and commensurate with your current level of fitness rather than your previous PR or long term goal. The same thing goes for Full Frontal testing. If you’ve take some time off the bike, then it’s usually best to target power outputs that you’ve maintained at a similar level of fitness rather than versus your all-time bests.

In the real world, we tend to ride courses which have variation in elevation profile and even changes in wind speed and direction that impact our speed for a given effort. Because of these variations, you need to adapt your pacing strategies to achieve the fastest possible time. There’s a little quiz that I often like to ask riders who are preparing for hilly or rolling time trials, and it goes something like this: “If you were to ride up a 1km long hill at 10 km/hour and then go down the other side of that 1km long hill at 40 km/hour - what is your average speed for the 2 kms worth of going up and down the hill?” Before you answer 25 km/hour (the average of 10 + 40 km/hr, which is wrong) consider the fact that you’ll be spending more time going slowly up the hill at 10 km/hour (it would take 6 minutes to go 1km at 10 km/hour…whether you’re going uphill or downhill or on the flat), and spending less time at the higher speed going downhill at 40km/hr (which would take 1 minute and 30 seconds to go 1km…again, no matter whether that km is downhill or on the flat (or uphill if you were fit enough). So, in this example it would take 7 minutes and 30 seconds to cover 2 kilometers which results in an average speed of 16 km/hr for the 1km uphill at 10km/hr and 1km downhill at 40 km/hr. If you’re interested in this math, you can look up harmonic mean calculations. (Note: this example works when using equal segment lengths/distances. If the distances of uphill & down are different, then you’d have to use weighted harmonic mean calculations.) If you went up the same 1km hill at 12 km/hr and then rode downhill for 1km at 35km/hr you would still end up with a faster overall time (6 minutes 43 seconds) and average speed (17.9km/hr).

Even if you aren’t really into the math, the takeaway of this is the following: in order to increase your average speed the goal is to put more effort into increasing your speed in the slowest parts of the course and decreasing your effort during the fastest parts of the course in order to recover and ultimately cover the distance as fast as possible. One thing to keep in mind, though, is that your ability to generate energy (or power) does have limits - and that the maximum variations in energy output that would result in the fastest possible time might not be sustainable for the entire duration of your event. Below are some general rules of thumb that I’ve given to athletes preparing for time trials of various duration below:

Total Event Duration GAP=Goal Average Power (% FTP) Variation in Sustained Power*
< 30 minutes 105-115% +/- 10-20% of GAP
30-60 minutes 95-105% +/- 5-15% GAP
1-2 hours 90-100% +/- 5-10% GAP
2-5 hours 75-90% +/- 5% GAP

*Note: In some cases like fast descents, technical curves/turning, etc. the lower sustained power limit can become zero to stay safe and allow for recovery, especially on faster downhills. In other cases, short efforts (generally 20-30s) required to accelerate out of a corner or to push over a short/steep hill can exceed the upper % GAP target - but be judicious with these efforts as fatigue can accumulate over longer distance efforts.

Example 1: Joe is preparing for a 15-km time trial that is on a rolling out and back course. In the past, his FTP was set at 225W from a Full Frontal test performed less than a month before a race on the same course where he averaged 240W (107% FTP) with a steady power output ranging from 230 to 250W (+/- 4% of GAP). The next time he rides the same 15-km TT course I would encourage Joe to push 110% of GAP (~265W) on the uphills and take it easier on the downhills at 90% of GAP (~215W). This should result in a similar average power but slightly faster time.

Example 2: Ella is racing a half distance triathlon with a 90-km long bike portion. The course is a big rectangle, but typically has a headwind in the first section and a tailwind in the third section, and crosswinds in the two sections. Last year on this course, Ella averaged 150 watts for the whole race without much variation but might have overdone it a little bit as she really faded in the last 5km of the run in last year’s race. Her FTP was 180 watts last year, though this year because she spent a big block of time improving her cycling her FTP is now 195 watts. Even though Ella’s FTP is nearly 10% higher than last year, I would only recommend slightly increasing her goal average power (GAP) to 156 watts since she faded on the run last year. The other big change to her pacing strategy would be that I would recommend that she push 10% over her GAP into the initial headwind (target 171W, which is still less than 90% of her current FTP), ride steady in both cross wind sections at 156W, but to take it easy in the tailwind section at 140W and ensure that she’s taking in enough fuel and hydration to make sure that she’s got enough energy to last her into the 21.1km run to finish this race.

A great way to test what is best is to create a short course that takes 10-20 minutes to complete or ride a short virtual course with uphills and downhills and experiment with steady goal average power output for your first effort. Next ride the same test course in similar conditions/equipment with small variations in average power relative to hills/wind (+/- 5-10%). Then compare the previous results with a third effort including larger variations in power relative to the hills and wind (+/- 20%), and see how you perform in terms of time/speed as well as how you feel. Remember, the goal is to push higher power output going uphill or into headwinds, and the lower effort downhill and with a tailwind.

If you’re a multisport athlete who is going to have to run after your cycling portion of an event, then you might want to reduce the overall variation to reduce fatigue and save your legs a little more for the run. Also, if you are running or swimming before riding then you might need a few minutes at a slightly lower intensity before dialing into your goal.

The next time you ride your favorite loop or try to tackle a longer Strava segment for a personal record time, think about how you can employ these pacing strategies to improve your times.

Test things out and let me know what you’re finding and experiencing!

PS - If you’re trying to set a UCI Hour record, I recommend using a steady pacing strategy after you get up to speed in the first lap…it’s a long 3,600 seconds and you can never really take a break, rest, eat or drink until it’s all over!


Great read thanks Sir Neal!
Looking at the chart of hour records, I wonder how much faster Bobridge could have gone with better pacing? And for a rider not famed for his TT ability, how perfect was Jens’ negative split (or could he have gone a bit faster overall)?!


I believe that if Bobridge had paced properly he could have beaten Brandee’s record…but it was exceedingly clear after 4 laps that he was headed for an epic fail and should have stopped, reset and restarted.

1 Like

Could you use the same idea to pace a hill climb? I’ve tried to pace our local climbs by going harder on the steeper sections and a little easier on the flatter sections, but can’t say I’ve attached any numbers to how much harder or easier to go. I’m guessing it would depend on gradient, duration and target power so could be more complicated?


Interesting. As a marathon runner, I used to blast the first 13 miles getting within in minute of my HM PB , then have time in the bank to slide.I would typically lose 10 mins over the 2nd half. A few times I paced athletes who were 20-30 mins slower than me, yet I still lost this 10 mins on the 2nd half. Perhaps it is the ability of the muscles to continue to get and use energy deep into an event that is critical. As a cyclist, I find my starting speed or effort, determines the rest of the ride. If it starts zone 3, then it is difficult to go harder at any point. Starting harder makes me loath to lose the speed, and so I seem to be able to continue it longer.

1 Like

Hi Alistair - yes, you should definitely use a similar principle of varying the effort on a longer climb to optimize your time. As an example, we have a local 9km climb that has the first 6km at about 8% grade and the upper 3km at 10% avg gradient. I usually ride the lower 6km at about 3-5% lower power than my goal for the whole climb and then push about 4-6% higher than goal average for the upper 3km. If I’m trying to average 275W for the whole 36 minutes of the climb, it’s normally about 20 minutes at at 260W and the upper 16 minutes at 285-290W.


Great read Sir Neal. I’m about 30 seconds off the best time on this local 10 minute climb. I’ve struggled to pace this as the first pitch gets pretty steep, followed by a longish flat section, then the final pitch. With no power meter I’m going by RPE. In my last attempt I pushed pretty hard on the first pitch, then took a little breather in the beginning of the second pitch before getting into TT mode and pushing fairly hard. When I got to the last pitch I just pushed with whatever I had left to get to the top. I think I’m following your philosophy, but any input would be welcome.

1 Like

Hi @Sir.Jeff.Kerr! From what I see here, on the profile - here are some thoughts:
Mile 6.5 to 7 - lower grade, so don’t push quite as hard here…maybe just below threshold.
Mile 7-.5 - steeper grade. Since you still have a ways to go after this section you need to go harder…but not completely blow up and NEED a breathing from 7.5 to 8.3 miles. My approach would be to start at an effort that feels like threshold (RPE, HR and breathing as a guide) and then steadily increase the effort every minute or so in that section so that you are hitting the highest speed at the 7.5 mile mark where it gets easier…but you’ll hit that section at a higher speed so that you can reduce your effort for the next 0.8 miles to catch your breath and drop back to just below threshold - but never completely resting/recovering.
8.3 to finish: It looks like it gets steeper here - so this is your ALL IN moment. For that, it’s VO2 max effort…above threshold…but you need to sustain the effort all the way to the end. Also, if you use a Wahoo Bolt, you can use a favorited Strava segment and see your pace relative to either the KOM or your PR so see if you’re ahead or behind of your previous best. Hope that helps - good luck getting that KOM!


I have an ELEMNT I use most of the time. I’ll look and see if I can set this up in RideWithGPS. Now I’m motivated to crush this climb. Thanks!

1 Like

Not sure if RideWithGPS has that feature or if it’s exclusive to Strava live segments…


Brilliant! I’ve never thought of varying my effort for a big climb but this makes perfect sense! So thanks for writing this Sir @Coach.Neal.H I’ve vEverested Alpe d’Huez in that other place and am planning on doing the same with Ventoux. For the Alpe I just tried to keep a steady Low Zone 2 but for Ventoux, I’m gonna study the gradient chart and see about going up a bit more, still in Zone 2 but higher, for the steeper bits and lower for the less steep. Of course losing 5-10 kg would make it all a whole lot quicker :thinking::stuck_out_tongue_winking_eye: This doesn’t change how I feel about you by the way. #everybodyhatessir@coach.neal.h


Everesting … jings … this seems like a challenge so out of reach as it’s not a ‘time’ challenge … long time on bike = fine, but this one needs power and lots of it !! how long did it take you sir?

1 Like

For the Alpe I did 9 full reps (including descents) as I was hoping to get a 10K done. Had to bail at 9400 metres. I was just spent. It was 16.5 hrs moving and almost 20 elapsed. When I did it in November 2019, you still had to stay on the bike during the descents so my breaks had to be after the full descents and I NEEDED them to stretch, change kit, refill water etc. Sooooo satisfying to complete. There’s a few Sufferlandrians that have taken to Everesting and at least 2 I know of that have done them outdoors as well (my Mt. SUF for 2021)


Yeah, RWGPS doesn’t have this. I’m too cheap to buy a power meter and too cheap to spring for the premium Strava memership too. :wink:

While I wait for the right time for my local KOM attempt, how about dissecting last year’s Mt Washington climb? I’ve put my HR and the gradient overlayed on the elevation profile. With your custom training plan (and a borrowed lightweight wheelset) I was able to target 1200 VAM and hit my target time of 1:25. The last pitch of 50m or so is 28%. It is a mental as well as physical challenge. I may be able to have held a higher effort from mile 5.5 on, but we had a pretty good headwind to deal with on the day, and I was trying to hold a little in reserve for the last pitch.

Any thoughts on improvement? The announcers always say “don’t go out too hard and blow up”, but it’s hard to find that sweet spot of hard but not too hard.

1 Like

Absolutely impressive Sir @Glen.Coutts! I’ve managed over 4,000m on a couple of rides…hard to imagine more than doubling that in one ride! With Everesting (as far as I’d recommend), the variations in pacing could still occur…but the goal average power would be much lower, and therefore the variations should also be a tighter range overall since it’s all about maintaining the effort over the long haul. This May I had visions of riding 1/3 of the laps it would take to Everest on one of our local climbs, Flagstaff, on the day that Ruther Winder (current US PRO champion) set her Everesting record…which would have been 5 laps (Ruth did 15). I managed 4 laps and was completely undone. Who knows, though, maybe I’ll get the itch sometime in the future to give it a whirl!


Hello @Sir.Jeff.Kerr - looking at this graph, if I assume your heart rate is the blue…it overall looks like a pretty solid pacing, though it seems like you’ve got a bit of a drop from the 5.5 mile mark until just before the end. If that’s the case, then riding those first 5.5 miles just slightly easier (likely 2-4 beats/minute - or likely about 5-10 watts) might be helpful next time…though keep in mind that there’s day to day variation in heart rate to consider relative to heat/hydration/anxiety/caffeine/etc. that makes it hard to completely pin down. A powermeter would for sure give much better insight into exactly how you paced things relative to your output. Congrats on a great ride, though - Mt. Washington HC sure is a bucket list ride/race!


I may have to ask Santa if he can get me a pedal based power meter that works with SPD-SL cleats… I think it would take some magic to pull that off


Nice. Sounds like a similar boat to me then time wise. I know a fair few full in ‘club level’ folk and of course they get it done in 12 hours … then I looked at it and though more like 18 :slight_smile:

1 Like

I did a v10K in 14:40 and IRL 5K in 8:30 ten days apart. My pacing was essentially “ride as easy as possible” where the slope, gears and cadence dictated the speed.
The 10K was long, but I finished strong. I can’t bring myself to say it was too hard riding it in an air conditioned room, watching TV and strolling to the kitchen on the decents…hats off to IRL Keepers Of The Cloud.
The 5k IRL was difficult in its own way, mainly because of the heat and humidity, and the effort to descend. There are also logistics involved and these prevented a full Everesting that time, but it’s on my list…