NP and TSS

tomUK said:

When calculating NP from a Powertap CSV do you take an average of each 30 second period? is there a quick way to do this without using cyclingpeaks software?

I believe NP is trademarked by TrainingPeaks. It's possible to build a spreadsheet from Dr. Coggan's public methodology, but I'd recommend the WKO+ software for many more reasons than just this.

tomUK said:

Also, in terms of TSS - have do people use this to avoid overtraining (knowing when to rest) and how much can you realistically increase this by each week? 10-15%? Also, at what point do you decide that you have peaked?

Again, the WKO+ software will help you manage this through the Performace Manager Chart. For niche software, it really is worth it to anyone who wants to train with a power meter.

also, is average power (minus zeros - as calculated on the Powertap) considered to be similar to NP?

thanks

Any similarities there would be purely coincidental. The math is completely different, although it's possible to construct a ride where the two were close/equal.

tomUK said:

also, is average power (minus zeros - as calculated on the Powertap) considered to be similar to NP?

thanks

NZAP is a pretty meaningless number.

I'd suggest changing your PT CPU to read the true average. That's what the software is showing anyway after you download.

based on that logic; how then is NP not a meaningless number? I mean, one is able to record the NZAP before the ride is downloaded and add it into the notes section. If one were to use NAZP for IF and calc. TSS using such, then surely the numbers are of use, assuming one sticks with this method across the board.

Quoted post said:

...If one were to use NAZP for IF and calc. TSS using such, then surely the numbers are of use, assuming one sticks with this method across the board....

Only if your ride characteristics in terms of time spent coasting and duration of coasting intervals stays constant in all riding situations. Non Zero Average Power has no basis in physiology, it just bumps the average up by a random amount depending on how much coasting you did or didn't do during a ride.

The NP algorithm models metabolic processes from the 30 second averaging that tracks physiological response times to the 4th power averaging that tracks blood lactate responses. You can debate the validity of that modeling approach (and folks do) but it's based on something.

NZAP in contrast isn't based on anything other than the observation that the average increases if you ignore rests and as frenchyge points out when NP and NZAP agree or come close to agreeing it's coincidence.

-Dave

tomUK said:

based on that logic; how then is NP not a meaningless number? I mean, one is able to record the NZAP before the ride is downloaded and add it into the notes section. If one were to use NAZP for IF and calc. TSS using such, then surely the numbers are of use, assuming one sticks with this method across the board.

NP takes into account variability of effort and the fact that intensity and strain are curvlinearly related.

NZAP doesn't.

NZAP will give you the same kJ as AP, hence using NZAP as an input is basically the same as saying that you can replace kJ for TSS in the impulse response model.

Of course you can choose to do that but it would be a misleading indicator of riding stress, unless your riding was of the same vanilla flavour every day. Here's a comparison of the differences between kJ and TSS over the course of a season:
Alex's Cycle Blog: TSS vs. kJ

I you swear off coasting, then you can just use AP in place of NZAP. That still wouldn't be the same as NP, but at least there wouldn't be the feeling that the zeroes were somehow skewing the ride picture. 🙂

frenchyge said:

I you swear off coasting, then you can just use AP in place of NZAP. That still wouldn't be the same as NP, but at least there wouldn't be the feeling that the zeroes were somehow skewing the ride picture. 🙂

How does one decided what is a 'weekly' TSS to aim at? For example in the book 'training with power' Coggan states that Armstrong can do a TSS of 350-400 at IF .85 for 21 days in a row and get stronger, whereas 2 days of this for a beginner may serverly over-reach him/her. I guess I'm just wondering how you assess these things and how you work out when to do your LT workout, your endurance work, etc.

For example I did a 57 TSS ride (endurance) on tuesday, a 163 TSS ride (endurace) yesterday and a 101 TSS ride (LT) today. Today's workout felt really tough to do and i just felt like I had no steam for it.

Cycling Peaks WKO+ uses an exponentially weighted average of the ride TSSs from the past few weeks to determine your current acute training load (ie, fatigue) and chronic training load (ie, fitness). Fatigue builds up quickly, and dies off quickly through rest. Fitness takes longer to build, but also remains longer.

See if this helps: http://home.trainingpeaks.com/media/68517/pmc_summit.pdf

Short answer (for me) is that I limit my average daily TSS to roughly my current CTL (fitness score) + 30. In this way, as my fitness score increases, representing a greater adaptation to the current training load, I am simultaneously increasing my average training load. That keeps a nice steady training stimulus on the body without overdoing it.

If you are dead-set against the WKO+ software package, you could use an average of the last 7 days TSS scores to estimate your current fatigue score, and an average of the last 6 weeks TSS scores to estimate your current fitness score.

tomUK said:

For example I did a 57 TSS ride (endurance) on tuesday, a 163 TSS ride (endurace) yesterday and a 101 TSS ride (LT) today. Today's workout felt really tough to do and i just felt like I had no steam for it.

In your case, your fatigue has built up to the point that additional workouts would be unproductive. A rest day will let the fatigue die down and the fitness rise somewhat, so that you can continue your training. The software really makes it all very simple with charts, predictive functions, etc.

tomUK said:

How does one decided what is a 'weekly' TSS to aim at? For example in the book 'training with power' Coggan states that Armstrong can do a TSS of 350-400 at IF .85 for 21 days in a row and get stronger, whereas 2 days of this for a beginner may serverly over-reach him/her. I guess I'm just wondering how you assess these things and how you work out when to do your LT workout, your endurance work, etc.

Good question. It seems like a pre-built training plan might address this. Which power based training plan or plans on trainingpeaks.com would encompass a periodized approach to training with power?

frenchyge said:

Cycling Peaks WKO+ uses an exponentially weighted average of the ride TSSs from the past few weeks to determine your current acute training load (ie, fatigue) and chronic training load (ie, fitness). Fatigue builds up quickly, and dies off quickly through rest. Fitness takes longer to build, but also remains longer.

See if this helps: http://home.trainingpeaks.com/media/68517/pmc_summit.pdf

Short answer (for me) is that I limit my average daily TSS to roughly my current CTL (fitness score) + 30. In this way, as my fitness score increases, representing a greater adaptation to the current training load, I am simultaneously increasing my average training load. That keeps a nice steady training stimulus on the body without overdoing it.

If you are dead-set against the WKO+ software package, you could use an average of the last 7 days TSS scores to estimate your current fatigue score, and an average of the last 6 weeks TSS scores to estimate your current fitness score.

In your case, your fatigue has built up to the point that additional workouts would be unproductive. A rest day will let the fatigue die down and the fitness rise somewhat, so that you can continue your training. The software really makes it all very simple with charts, predictive functions, etc.

I'm not sure I understand. Right now I'm using the trial version of WKO+ and my current CTL is 27. Does that mean you suggest that I should limit my daily workouts to not exceed 57 on the TSS score and let it build slowly? My ATL is showing as 65 right now and TSB is 3.4.

Thanks for taking the time to reply.

correction - TSB= -34

tomUK said:

I'm not sure I understand. Right now I'm using the trial version of WKO+ and my current CTL is 27. Does that mean you suggest that I should limit my daily workouts to not exceed 57 on the TSS score and let it build slowly? My ATL is showing as 65 right now and TSB is -34.

No reason to limit the daily rides, just space them out a bit. You are reading correctly, though, that if your CTL is 27, then your daily TSS would *average* ~57, or 400/wk if you don't ride every day.

Now, if you're just starting the program you should seed the CTL and ATL values rather than letting the start at zero. Otherwise your CTL will be artificially low for a while until the math catches up.

Seed values:

Quoted post said:

Your CTL and ATL has to begin somewhere(unless you have 'just' started training), therefore you need to set some starting values. A good rule of thumb is to look at the average number of hours you train per day and multiple that by 50 to 75 (for more intense 'trainers'😉. This will give you an average of how many TSS you score per day. This would give you a good starting point for your CTL and ATL values.

Source: What is the Performance Management Chart

mcdelroy said:

Good question. It seems like a pre-built training plan might address this. Which power based training plan or plans on trainingpeaks.com would encompass a periodized approach to training with power?

That I don't know (although I guess the TP plans have progressive workload built in but must assume your starting point).

The plans we provide do take into account your current training load and fitness (as well as goals/type of riding/racing, a couple of key event dates and days/time available to train). See here:

RST Training Plans

[quote="frenchyge"]No reason to limit the daily rides, just space them out a bit. You are reading correctly, though, that if your CTL is 27, then your daily TSS would *average* ~57, or 400/wk if you don't ride every day.

QUOTE]

Do you mean 400/wk if you ride everyday? Hence if I were riding 5 times/wk i'd have 285 units to spread over workouts throughout the whole week?

Thanks, again.

tomUK said:

Do you mean 400/wk if you ride everyday? Hence if I were riding 5 times/wk i'd have 285 units to spread over workouts throughout the whole week?

400 TSS would be my weekly target. I then budget those throughout the week depending on which days I chose to ride, the workouts I wanted to accomplish, etc. So, it might be four 100 TSS rides, five 80 TSS rides, three 133 TSS rides, or any combination which results in roughly 400 total for the week. From there, I just train. If the resulting TSS from a ride differs significantly from the plan, I may revise a later workout or I may just press on. If the weather is nice one day, then I may do my long ride that day and move the shorter ride to another day. It's not going to cause a problem is you overshoot for a day or even a week. Hope that helps.

Using CTL+30 as a daily average target produces a CTL increase rate of approx 5 per week. So, if your CTL were 27 today, then after a 400 TSS week you would be at approx. 32 by the end of the week. Your daily average TSS target would then become 32+30 = 62, or (62x7)= 435 for the next week. That's how I establish and maintain a moderate training stimulus from week to week without the need for rest weeks, etc.

Early in the season, when CTL is low, I usually just train by feel using sweet-spot intensity and don't really worry if I exceed the weekly TSS goal. It's not unusual for me to see CTL bumps of 8-10 during early season training if we get good weather and I go on a couple long team rides. A sustained CTL increase rate of 3-8 per week is the range which works well for most folks over a long build period, but it's not important that every week adhere to a specific value. There's room to build schedule flexibility into the long-term plan so that you're not a slave to the numbers. 🙂

frenchyge said:

400 TSS would be my weekly target. I then budget those throughout the week depending on which days I chose to ride, the workouts I wanted to accomplish, etc. So, it might be four 100 TSS rides, five 80 TSS rides, three 133 TSS rides, or any combination which results in roughly 400 total for the week. From there, I just train. If the resulting TSS from a ride differs significantly from the plan, I may revise a later workout or I may just press on. If the weather is nice one day, then I may do my long ride that day and move the shorter ride to another day. It's not going to cause a problem is you overshoot for a day or even a week. Hope that helps.

Using CTL+30 as a daily average target produces a CTL increase rate of approx 5 per week. So, if your CTL were 27 today, then after a 400 TSS week you would be at approx. 32 by the end of the week. Your daily average TSS target would then become 32+30 = 62, or (62x7)= 435 for the next week. That's how I establish and maintain a moderate training stimulus from week to week without the need for rest weeks, etc.

Early in the season, when CTL is low, I usually just train by feel using sweet-spot intensity and don't really worry if I exceed the weekly TSS goal. It's not unusual for me to see CTL bumps of 8-10 during early season training if we get good weather and I go on a couple long team rides. A sustained CTL increase rate of 3-8 per week is the range which works well for most folks over a long build period, but it's not important that every week adhere to a specific value. There's room to build schedule flexibility into the long-term plan so that you're not a slave to the numbers. 🙂

Makes perfect sense. Thank you, again so much for taking the time to answer each of my questions. I appreciate your time.

I wanted to ask how low you let your TSB go before you deem a rest day (or two) essential so you don't dip into overtraining. How do you know, btw, if you are overtrained or over-reached?

tomUK said:

Makes perfect sense. Thank you, again so much for taking the time to answer each of my questions. I appreciate your time.

I wanted to ask how low you let your TSB go before you deem a rest day (or two) essential so you don't dip into overtraining. How do you know, btw, if you are overtrained or over-reached?

Glad to help. 🙂

One day of low TSB is not going to result in overtraining, but could result in a compromised workout and less effective training. The ability to tolerate and recover from fatigue depends upon the individual (age, diet, rest, etc.), so it's worth easing into it a little to see how your particular body will respond. For sweet-spot or lower intensities I'm usually fine down to about -30 TSB before the workout (obviously it's much lower afterwards), but for higher intensities I like to be above -20 for best results.

The math works out such that if you're keeping your average TSS at 30 pts over CTL (as described in previous posts), then you're TSB will also average around -30 for the week. If you find that that seems too low to sustain, then reduce that value down from 30 and try a smaller CTL increase rate.