…but you can make it tell the truth.
The Performance Management Chart in TrainingPeaks is an extraordinary tool. As early as 1975, physiologists have been attempting to quantify, and ultimately predict, the relationship between training and endurance performance. Coaches and athletes wanted to know if you contributed X units of training, could you predict Y units of performance? Variations of this impulse-response model were proposed, until the late 1990s when Dr. Andrew Coggan introduced the Performance Manager. It was a brilliant, easy-to-use application to quantify the impulse-response relationship, and is used to this day within TrainingPeaks as the Performance Management Chart (PMC).
Regarding your understanding of the PMC, you probably fall into one of two categories. You’re either a die-hard daily PMC consumer, or puzzled by the beautiful colors of this strange chart in your TrainingPeaks dashboard. Without deliberating too long on how the PMC works (which often takes an entire book), I’ll give a summary. Each workout you complete is assigned a Total Stress Score (TSS). That score is relative to your threshold, where a score of 100 TSS is the output from your 1-hour time trial. All other workout scores orbit around this fitness polestar. An easy 30-minute ride might be a TSS of 25, but a marathon could be 250. In theory, that easy bike ride was 25 percent as stressful to your body as your 1-hour time trial, and that marathon was 2.5 times as stressful. Your average daily TSS over 6 weeks becomes your Chronic Training Load (CTL), and CTL is considered your quantitative fitness. In the PMC world, a higher CTL is always better. Theoretically, if you start an event with a CTL of 100, you will perform better than at a CTL of 90. Because TSS is always relative to the individual’s threshold, CTL is always relative. An individual with a CTL of 100 may not out-perform an individual with a CTL of 90 if the latter has a higher threshold to begin with.
After 15 years of using the PMC for myself and athletes, I’ve determined that by default, the PMC is lying. The good news is that you can tame this psychedelic graph once you understand its risks, limitations, and strengths.
The PMC Promotes Wrong Behavior
In many circles, the CTL is a badge of honor. Athletes are in awe of the elite triathlete with a CTL of 130 and seek to emulate them. When not used socially, it’s used as an individual “white whale” to be pursued. “If I reached a CTL of 80 last year, I have a goal for a CTL of 90 this year.” This CTL-based agenda impacts every workout, where the athlete, either consciously or unconsciously, tries to achieve as high a TSS as possible from the workout, and spends significant time at moderate intensity, which in turn provides that higher CTL. Trying to Identify this “moderate-intensity rut” as the primary roadblock to improved performance. Ironically, in many cases that higher CTL results in a decrease in real fitness. The CTL is up, but fatigue and stagnation keep the actual race-day result disappointingly low.
CTL Does Not Reflect the Fitness
To demonstrate how the CTL can be deceiving, let’s look at two hypothetical age-group athletes, Rebecca and Rachel. They are the same age, same weight, same thresholds, same experience level, and training for the same event. Both use the PMC to track their fitness. Rebecca trains by feel, but Rachel discovered and implemented 80/20 training early in her season.
In a sample week, Conor and Marc both train for 10 hours. Conor falls into the classic moderate-intensity rut and performs 50% of her training at low intensity, but Marc follows his plan closely and spends 80% of her time at low intensity. Their respective intensity distribution for the week looks like this:
Connor’s 10 Hours
2 hours in Zone 1 (80 TSS)
3 hours in Zone 2 (150 TSS)
3 hours Zone X (248 TSS)
1.5 hours Zone 3 (105 TSS)
0.5 hours Zone 4 (20 TSS)
Total: 603 TSS
Marc’s 10 Hours
2.5 hours Zone 1 (100 TSS)
5.5 hours Zone 2 (275 TSS)
0.25 hours Zone 4 (25 TSS)
0.25 hours Zone 5 (30 TSS)
Total: 570 TSS
If Conor and Marc continue a similar pattern for the remainder of their plan, Conor will have a CTL of 5-10 percent higher than Marc. Based on the PMC alone, Conor should outperform Marc at their event. Much to Conor's surprise, Marc has the better performance on race day.
The reason is simple: by spending sufficient time at low intensity using our method, Marc can recover quickly and gain new fitness from higher intensities using those fresh legs. Conor churns out more raw TSS, and therefore a higher CTL, but she remains mired in moderate-intensity training where recovery is compromised, quality high intensity is rare, and as a result, has lower actual fitness than Marc.
The PMC Cannot Reflect Specificity
Successful endurance training is based on a few universal principles. These include the principle of progressive overload, the principle of hard/easy, the principle of intensity balance, and the principle of specificity. Training specificity simply means that your movement, exercise duration, and intensity should reflect the event you are training for. There’s an infinite number of ways to come up with the same TSS score for a given workout, but that identical TSS score does not promote identical abilities. For example, here are just three run workouts that all result in a TSS of 75:
1.5 hours Zone 2 = 75 TSS
53 minutes Zone X = 75 TSS
45 minutes Zone 3 = 75 TSS
According to the PMC model, each of those workouts provided you with the same increase in fitness. But, which workout provides the specificity for a 10K? Which is best for half-marathon training? Replicate these workouts over a training season, and you’ll end up with multiple athletes with the same CTL, but with completely different abilities. I can rack up an impressive 120 CTL if I did nothing but Zone 2 cycling for a year, but I’d then pass out trying to complete a half marathon at low Zone 3. CTL can only capture the quantity of training, it can’t capture the quality of sport-specific intensity and duration.
CTL Does Not Matter. Speed Does
As demonstrated in the two sections above, CTL can be misleading. Ultimately, CTL is also irrelevant. None of us arrive on race day, show the race director our 130 CTL, and request a podium award before the race even begins. CTL may be an indicator of our fitness relative to a previous period, but in the end, we have to earn results on the asphalt. The ultimate measure of the effectiveness of a training plan is demonstrated in the field with racing and periodic testing. Regardless of what the CTL says, if you are getting faster, your plan is working.
The PMC Requires Significant Maintenance
Even Dr. Coggan would agree that the PMC is a harsh boss. Every workout must be recorded. Every workout must be recorded accurately. Your threshold is constantly changing (as it should in an effective plan) which means you must constantly change your threshold in the PMC. The device runs out of battery on a 3-hour ride? Good luck in guessing the TSS. Using HR for TSS one month and Pace the next? A recipe for PMC disaster. Do you include strength training? How do you enter a TSS for that? (I use 40 TSS per hour, a colleague uses 70) I once spent 45 minutes calculating a TSS score for an athlete who participated in a tennis match. An accurate PMC requires a complete understanding of the methodology and daily commitment to ensuring the inputs are accurate. In this scenario, it’s not a matter of the PMC lying to you, it’s a matter of you lying to the PMC.
The biggest mistake athletes make with the PMC is to just let TrainingPeaks automatically manage it for them. TrainingPeaks will guess your threshold, and then it will guess your future TSS to guess your future CTL. It’s a guess from an estimate wrapped in a presumption. The only accurate PMC is a micro-managed PMC requiring daily, workout-by-workout verification that the TSS recorded, and the multiple inputs that affect that TSS, is accurate.
CTL is Always Relative: Even To You
As mentioned earlier, TSS is always relative, and therefore so is CTL. Rachel’s brother-in-law, Brad, has also started 80/20 training, and he will be participating in the same event as Rachel. Both Rachel and Brad adhere to their 80/20 training plan, and both arrive on race day with a CTL of 80. Why would Rachel still outperform Brad if they have the same CTL? Instinctively, we know why, that these are very different athletes, but let’s quantify it anyway.
Brad has a running threshold pace of 7:30 per mile. Rachel has a threshold pace of 7:00 per mile. Brad runs one hour in mid-Zone 2 and achieves a TSS of 60. Rachel runs one hour in mid-Zone 2 and also achieves a TSS of 60. But Brad’s mid-Zone 2 is 9:10 per mile and Rachel’s is 8:30 per mile. Rachel is just faster than Brad, even though they achieve the same TSS for each workout, and ultimately, the same CTL.
But here’s what’s scary: CTL is not only relative athlete to athlete, it’s relative to you as well. Let’s say Brad is fed up with Rachel scorching him at races. So, he continues to follow 80/20 training and soon finds that his threshold pace is 6:50 per mile. Brad completes a 1-hour run in mid-Zone 2, or 8:25 per mile, only to discover his TSS score is (you guessed it) still 60. That’s because (if Brad manages his PMC correctly) the relative stress for Brad running for an hour in mid-Zone 2 compared to his threshold pace has not changed. Your 1-hour time trial TSS will always be 100, even as the raw speed continues to climb. Old Brad’s stress (or TSS) from running an hour at 9:10 with a threshold pace of 7:30 is the same as New Brad’s stress running 8:25 for an hour with a new threshold pace of 6:50.
The bottom line is that with a properly managed PMC, your CTL will not change for a set amount of volume and intensity distribution even if your threshold speed does increase. CTL can only increase with an increase in total training duration or a change in intensity distribution. What often happens is that athletes don’t manage their PMC by changing their thresholds regularly. As a result, the TSS score becomes artificially inflated because the current workouts are calculated against the old threshold. This not only requires changes in thresholds for future workouts but often retroactively changing thresholds for past workouts.
TrainingPeaks Has Some Bugs
TrainingPeaks is the best platform, but like all software, it has some issues. One issue has to do with the calculations TrainingPeaks uses to predict future CTL. One of the most common questions I receive is, “Why does my predicted CTL get lower when I apply an 80/20 plan?” In addition to the issues listed above, there are some quirks with the platform itself that prevent an accurately predicted CTL.
The most significant are two known problems in TrainingPeaks for HR-based workouts. First, the predicted TSS value will only increase in units of 10. If you have one run that should have a TSS of 41 and another run that should have a TSS of 49, TrainingPeaks will round down and calculate both as a predicted TSS of 40 (that’s a 20 percent reduction in TSS right there). Second, for a given amount of planned workout time, and regardless of the planned intensity, TrainingPeaks uses a minimum value. For example, if there are two runs of 30 minutes, one performed at 75% of LTHR and the other performed at 90% of LTHR, both will have a predicted TSS of 40. Thus, TrainingPeaks systematically miscalculates predicted TSS and therefore predicted CTL for HR-based structured workout plans.
This particular issue is limited to the predicted TSS and predicted CTL for HR-based plans only, and does not impact TSS or CTL for completed workouts, nor Pace or Power-based workouts. TrainingPeaks reports they are looking into the issue.
However, even with Pace and Power-based workouts, TrainingPeaks assumes that you will be running in the upper quartile of the zone range when predicting average Pace or Power (and therefore TSS). For example, if your Zone 2 run pace was 5:00 to 7:00 minutes per km, TrainingPeaks will assume that you will run an average of 5:30 per mile for a Zone 2 segment when predicting pace when in reality you could be running as slow as 6:55 per km and still be following the workout correctly. The inevitable result of Zone training is that your intensity for the day could fall anywhere within a broad range, therefore, the predicted TSS will almost always be incorrect.
How To Use and Interpret the PMC Correctly
Despite the articulated issues, I use the PMC for myself and my athletes. There is a way to effectively take advantage of this useful tool, which requires the following steps and insights.
Change your PMC chart start date to the day you started training With us. Anything pre-that is not valid.
Don’t compare your CTL to other people.
Accept that your new CTL era is going to be lower than your previous CTL. In most cases, your CTL will be 20% lower and you’ll get a PR in your next race.
Better to have a lower CTL that reflects race-specific fitness than a higher CTL with little specificity.
Do change your thresholds in TrainingPeaks after each threshold test. This is critical in maintaining an accurate PMC.
Use results, not CTL, to gauge your improvements.
Micro-manage your PMC and threshold testing, don’t delegate to TrainingPeaks.
Remember that by properly updating your thresholds regularly, you won’t see an increase in CTL if your total workout duration remains the same.
Unless you manually override the predicted TSS of every future workout in your plan, TrainingPeaks will not provide an accurate predicted CTL.
Once you have an accurate PMC and understand how to use it, the PMC can be used to compare your fitness to previous training periods and predict performance. At this point, your individual CTL of 90 will result in better performance than your old CTL of 80. Increasing volume is the best way to achieve this CTL improvement, as your intensity balance should remain the same.
The PMC can be an essential part of your training, as is it for me and my athletes. If you apply the above principles, you can change your current PMC into a real PMC (a Properly Managed Chart).
Comments