Training Stress Score for Cycling: What TSS is, how to use it, and what it doesn’t tell you
Training Stress Score, or TSS, is commonly used to describe how “hard” a ride was, but that’s only part of the story. Although it is valuable, some important aspects of training are not reflected by TSS. So, as you are scanning through your workout data and planning your upcoming cycling training, here’s a guide for using Training Stress Score effectively.
What is TSS?
Training workload is the product of duration and intensity (Workload = Time x Intensity). As duration increases, the intensity an athlete can sustain decreases. Inversely, the shorter a workout or effort, the higher the intensity can be.
Training Stress Score was developed by Dr. Andy Coggan and Hunter Allen. Their idea was to create a single metric that accounted for both intensity and duration to quantify the training load of individual workouts
Cycling TSS allows athletes to compare the physiological stress created by a short, high intensity workout to the stress of a 3-hour endurance ride.
How Training Stress Score is calculated
Training Stress Score is typically calculated automatically by devices and training software, like TrainingPeaks. Nonetheless, it is good to understand what data is used to score your workouts:
TSS = (Seconds x NP X IF) / (FTP x 3600) x 100
Spelled out, this means:
Training Stress Score = (workout time in seconds x Normalized Power x Intensity Factor) / (Functional Threshold Power x number of seconds in an hour) x 100
Example: (7080 x 183 x .85) / (215 x 3600) x 100 = 142 TSS
- Functional Threshold Power: The highest average power output a cyclist can maintain for 60 minutes.
- Normalized Power: An estimate of your power output that accounts for variability (coasting, hard efforts, easy spinning). A very steady effort will have an NP close to your average power. A highly variable ride could have a high NP and lower average power.
- Intensity Factor: The ratio of NP to FTP, or the relative intensity of a ride compared to the rider’s own FTP. For instance, an easy or recovery ride would have an IF less than .65, an endurance ride is likely between .70-.80, an interval workout or group ride will often be .75-.85, and a ride close to an athlete’s FTP would have an IF of .90-1.0. Short, very high intensity workouts under an hour may be over 1.0. Note: These ranges are a bit lower than what is published on TrainingPeaks, but reflect what our coaches see in data files from amateur and masters cyclists.
What is a good Training Stress Score?
Training Stress Score is neither good nor bad, but when viewed in context of past and future training, we can create target TSS values for individual workouts and training blocks. Another common training metric, Chronic Training Load (CTL) is an athlete’s average TSS over the preceding 42 days. It is sometimes thought of as an athlete’s “fitness level”, but it may be more accurate to think of it as “the workload an athlete has been sustaining”. A high CTL doesn’t always lead to high performance, particularly if the athlete is not getting enough rest.
A beginner might have a CTL of 40 and an experienced racer might have a CTL of 100. If these two riders went out and completed a 100 TSS workout today, it would be an easy to moderate intensity ride for the racer and an extremely hard ride for the beginner. When planning or evaluating training, it helps to look at daily TSS as a percentage of CTL. A common way to categorize workouts by TSS is:
- Easy: TSS 10-25% BELOW CTL
- Moderate: TSS 25% ABOVE CTL
- Hard: TSS 50-100% ABOVE CTL
For our racer, a hard workout would be 150-200 TSS, whereas a hard workout for the beginner would be 60-80 TSS.
How TSS is used and why it matters
For an athlete to make progress, training stress must be balanced by adequate recovery time. Standardizing a “score” for each workout based on its relative intensity and duration helps coaches and athletes schedule training effectively. For instance, TSS can be used to:
Estimate the “difficulty” of a workout
As mentioned above, you can put today’s workout in context to your recent workload by comparing today’s TSS to your CTL. This is important because it can provide a TSS range that’s going to be challenging but still manageable. An occasional epic ride (TSS 4x CTL, for instance) can be good, but too many rides of disproportionately high TSS can spell trouble.
Estimate recovery time required
When workout or daily TSS is higher, athletes need more recovery time before their next training bout. The exact correlation between TSS and recommended recovery time varies by athlete and training history.
Plan weekly and monthly training load
Cumulative weekly and monthly TSS can be calculated as you schedule interval workouts, endurance rides, and recovery days into a training plan. Even as the focus of training changes, tracking cumulative TSS provides an overarching view of the athlete’s total workload.
Control training ramp rate
Overtraining and injuries frequently result from increasing training workload too rapidly. Having a standardized way to compare the training stress of different workouts means we can tell how quickly training workload is changing (in either direction).
What TSS doesn’t tell an athlete
As with so many training metrics, TSS can be misinterpreted or misused.
How the training stress was created
By itself, TSS doesn’t tell you how the training stress was created. Today’s 140 TSS ride could have been a long and steady endurance ride, a structured workout with lactate threshold intervals, or a group ride with random surges and lots of drafting. The training stimulus from a workout is not the same as the training stress.
A ride with a TSS of 140 could be a 2-hour endurance ride or an interval workout with 30 minutes time-at-intensity at FTP. If we are just looking at TSS, these workouts are equivalent in terms of training stress. In terms of training stimulus, the rider is challenging their aerobic endurance with the 2-hr steady ride and challenging their power at FTP with the interval workout.
High-intensity interval training provides another example of this. A 60-minute VO2 max interval workout with five 4-minute intervals at 120% of FTP would be a hard session. However, only 20 minutes of the workout are at a high power output. The rest of the time (warmup, easy spinning between intervals, cooldown) will be at a low to very low output. As a result, these workouts can yield a low TSS value despite being creating a lot of stress and fatigue.
When the training stress was created
Training Stress Score attempts to combine duration and intensity into one metric, but stress applied to the body does not increase linearly with time. In other words, if you rode at a constant intensity (75% of FTP) for four hours, the actual training stress from Hour 1 is less than the training stress from Hour 4. This is because the stress required to reach Hour 4 affects your ability to perform in Hour 4. That’s not a bad thing, it’s part of the reason we do longer endurance rides. However, it means calculated TSS from a long ride may underestimate the actual training stress. For a great explanation of this concept, see this video from Dr. Stephen Seiler.
The effect of non-exercise stress
Neither your TSS nor your CTL (42-day average of daily TSS) takes lifestyle stress into account. TSS doesn’t reflect that you’ve been sleep deprived for the past two weeks because of work or a new baby. It doesn’t consider the quality of your nutrition choices or whether you are eating enough to supporting your energy expenditure. This is why it is important to record subjective feedback in your training logs, not just the device data.
Non-exercise stress becomes important when you are trying to figure out what training worked and what didn’t. If you previously struggled to recover from some high TSS workouts, the workouts may not have been the problem. The issues may have been from inadequate nutrition, poor sleep, and high lifestyle stress. Instead of avoiding workouts that hard, fix the recovery problems and next time around you may be able to cope with those same high TSS workouts.
Some related reading you may enjoy: Cycling Training Terms and Acronyms Explained
By Chris Carmichael
Founder and Head Coach of CTS
Training Stress Score®, Intensity Factor®, and Normalized Power® are registered trademark of Peaksware, creators of TrainingPeaks software.
Pingback: How to Increase Ultramarathon Training Without Injury - CTS
If Bestbikesplit is estimating an ultra race is going to have a TSS of 1100. I should build my CTL to something like 600? Is that correct? Thanks in advance.
Holy moly, no! You’d need to be on your bike 80 +/- hours/week for several weeks to develop a CTL of 600. TSS estimate of 1100 sounds like, maybe, 30 hours at 60% effort???
I’m racing RAAM this year and I’ll top out with a CTL around 130, which is probably on the low side for what most people do for that race – but I’ve got a great coach and every pedal stroke is high value so fitness won’t be a limiting issue.
You should probably build in a couple of days (or even weekends) with TSS in the 500-600 range – i.e. very long, hard days in the saddle – so that you have the experience of what ultra is going to feel like.
I guess I just don’t understand that point in the article or maybe it doesn’t pertain to races… Just trying to train smarter these days.
Have fun on RAAM. I competed in ’05 on a 4 man team, DNFed solo attempts in ’06 & ’07, and crewed for a solo in ’08. I have more stories from the ’05 race than all the others combined. I stick to shorter races these days like 24 hour races.
Pingback: How Many Hours Should Cyclists Ride for Fitness and Performance? - Chris Carmichael
Pingback: Essential Guide to Cycling Training with a Power Meter - CTS
Pingback: Training Burnout: Prevention, recovery, and how it's different from Overtraining - CTS
Pingback: Cycling Training Terms and Acronyms Explained - CTS
Pingback: What is Normalized Power® and How is it Used in Cycling Training? - CTS
Pingback: How to Shift Training from Build Phase to Specialization/Racing Phase - CTS
Pingback: Essential Guide to Training with a Power Meter - CTS
Just another great article. These are so informative. It makes it obvious to me how helpful your coaching would be. Obviously, a person has to be willing to not just spend the money, but dedicate themselves to the effort, i.e. nutrition, rest, discipline to follow the training.
ITS WORTH IT! I’ve seen GREAT results, in the two plus years of being CTS “coached”.
Crazy formula. It is essentially time in hours times (NP/FTP) squared times 100
Similarly I wonder if you get 80% of the info from watts x hours and/or average Heart rate x hours (or average minus resting HR to get a bit more sophisticated!)
Glad someone else sees that! TSS rises with the square of IF. That’s the big takeaway – intensity trumps duration, and more so when intensity is high.
Thx for explaining what my garmin keeps telling me that I had no idea what it meant.
I was watching my garmin today and noting the hr and the power output from the beginning varied inversely by the end of the ride. That is, less power and higher heart rate. Has anyone looked at average heart rate/average power as an indicator of fitness? I would think that would be pretty useful comparing a series of rides.
Great article, I think one of my favorites ever!
Thank you for this great explanation. I have been riding and racing for 50 years and training knowledge has gone so far in that time. We have moved from “Ride lots!” to power measurements to TSS and more. And so much is available to the everyday rider!
Thanks for your very helpful explanation of TSS – it has cleared up things for me also. Now could you take on ramp rate?
That was a great explanation, Chris, now I’m finally starting to understand all this (after all these years)!
And I also agree Training Peaks tends to overestimate TSS for me, particularly for lower intensity Endurance rides.