Everything Cyclists Get Wrong About Normalized Power

If you've ever wondered what Normalized Power is actually telling you, this episode is for you!

Normalized Power has become one of the most common metrics in cycling, but it's also one of the most misunderstood.

In this episode, CTS Head Cycling Coach Adam Pulford explains what Normalized Power was designed to measure, how it differs from average power, and why context matters when using it to evaluate training and racing.

HOST

Adam Pulford has been a CTS Coach for nearly two decades and holds a B.S. in Exercise Physiology. He's participated in and coached hundreds of athletes for endurance events all around the world.

​​Free Cycling Training Assessment: https://trainright.com/cycling-training-assessment-welcome/

Interested in working with a coach? Schedule a free consult: https://trainright.com/coaching/cycling/

Self-coached athlete? Check out our TrainRight Membership: https://trainright.com/membership/

Find more free resources here: https://trainright.com/blog/

Resources:

• What Is Normalized Power And How Is It Used In Cycling Training? - CTS 
• What is Normalized Power?
• Understanding Normalized Power Calculations to Coach Cyclists
• Normalized Power®: What It Is and How to Use It | TrainerRoad Blog
• The Science of the TrainingPeaks Performance Manager 
• Estimating Training Stress Score (TSS) 

Transcript

Why Normalized Power Gets Misused

SPEAKER_00 0:00

The problem with normalized power isn't that it's wrong or that it's just a made-up number. The problem is that many cyclists misuse and misunderstand it. I've heard all the grumblings and the prophecies of this algorithm over the past 20 years, and the reality is that normalized power is awesome if you know how it works, why it was built, and how to apply to your training, racing, and data analysis. It's not awesome if you try to use it for things it wasn't intended for or rely upon it as the sole metric for performance. Finally, as with most numbers, if you get overly attached to it emotionally, you're certainly creating more problems for yourself. I'm Adam Pulford, head coach of cycling at CTS, and I love data. And I love helping athletes use data to better inform their decisions about training and racing. It's a big reason why I've been coaching for decades and make all these videos right now. Helping athletes reach their full potential is really what drives me to do what I do today. Now, cycling is a sport with a ton of data in it, and there's an opportunity to leverage rich troves of information to better understand fatigue, performance, and the trends of them. The flip side of the coin is that it's easy to get information overload or misinformation about how these algorithms and metrics work together to inform the rider about these trends. And this is both the boon and the plight for the endurance athlete. Normalized power is a very popular and recognized tool in our sport. It actually is both an algorithm and a metric at the same time. And it's both simple yet complicated while being expansive in its use across different software platforms, wearable devices, and large sets of data. To better understand all of this, let me give you a brief history of normalized power, and we'll get into the really good stuff after that. How best to use and not use this ubiquitous but powerful data tool. Let's dive into the background first. So, a

What Normalized Power Measures

SPEAKER_00 2:02

quick history of normalized power is Dr. Andy Coggan developed the normalized power algorithm in the early 2000s, and it was then popularized in his book that he wrote along with Hunter Allen. Normalized power was built as a better way to estimate the cost of a longer, non-steady effort in cycling compared to average power alone. And that's how you should think about it, or define it in your head. It's an estimated cost of an effort for longer periods of time. For example, if your FTP is around 300 watts and you do a hard crit, let's just say a 340 watts normalized and 275 watts average, that 340 watt normalized tells us how it hit your body more accurately. It's not that you could have pedaled steady at that power, but it's like how your body receives it or feels it. And contrary to what some popular belief is out there, normalized power is not just taking the zeros out of it.

The Four-Step NP Calculation

SPEAKER_00 3:02

In fact, here's the actual math of normalized power. For the math nerds out there, this may actually look overly simple to you, but for the rest of us, it may look a little like foreign language. Since I'm a big fan of knowing how and why something works, let's break it down into simple terms. I've got the four steps shown here of how the calculation actually works, and that's shown on the screen. I took this ripe from an article uh off of Training Peaks. And if you want to uh read more about it and get information about it, read this whole article. It's short, it's sweet, it's to the point, but it's also very detailed. So it shows here that step one is calculating a rolling average with a 30-second window, and that repeats basically every second. Step two is you take the value from step one and multiply it by the fourth power. Step three is you calculate the average of these values from the previous step. And then step four is you take the fourth root of the average of the previous step, and this value gives us the normalized power. Now that's a mouthful, and like I said, I displayed it there on the screen for you. If you want to read more about it, check out that article. Now you don't need to be proficient at this math. I'm just encouraging you to better understand it. And the key takeaway is that normalized power weighs the higher power efforts a little more than the lower power efforts. Why? Well, think about it. Riding in zone two costs your body less fatigue than riding in zones five to six, so all power efforts shouldn't be weighed the same because higher intensity elicits higher strain and fatigue to our bodies, generally speaking. So having a metric that does a better job of quantifying that weighted cost to help you when trying to disseminate the impacts of your system is really helpful. We then take this metric and apply it to others to start to see a more clear picture of what's going on in our bodies while training and racing using smart data to guide us along the way. Training

How NP Drives TSS And IF

SPEAKER_00 5:03

Peaks adopted normalized power and uses it deep in their platform to assist other metrics that help us quantify training load and use it to calculate things like intensity factor, training stress score, variability index, and has its fingers and other things too. One thing I will point out here is to show how TSS is made because it helps shed some light on how deeply involved normalized power is in our training journey. Training stress score or TSS is calculated by using how long you rode, how hard you rode it, all relative to what your current FTP is. And normalized power is how hard you rode it, as you can see in this equation. By definition, 100 TSS points equals 60 minutes at your functional threshold power. You can ride longer and less intense and rack up some points. You can ride shorter and more intense in rack up points, but the anchor of all of this training load that we call TSS is generated from normalized power and duration. Not everybody knows that, so I wanted to point it out here. Now I kind of don't want to take all your time talking about the super detailed inner workings of things like this because it's easy to look up. And actually, Jim Ruppberg, a fellow CTS coach and editor of our CTS newsletter, has a great article out there of how normalized power is nested in the Training Peaks suite of things. So read that in the links below. And if you like what you're hearing so far, hit the like button. If not, stick around because next, I want to spend the majority of the time on the real-world applications of how best to use and not use normalized power.

The Biggest NP Mistakes

SPEAKER_00 6:40

Let's start first with how not to use normalized power. And I'm just going to rip through a bunch of these here and give you a quick summary of it at the end. First one is using it as an end-all, be all. Next, using it because it's just a bigger number, using it for durations less than 10 minutes, using it to estimate your FTP, or using it to set training ranges. The quick summary of all of this is that normalized power needs context. So looking at your average power along with intensity factor, TSS, duration, etc., is the best way to gauge how it all went on training day or race day. Not just that you hit a bigger number. The algo wasn't built for anything less than 10 minutes, and in fact, the longer you go, the better. I tell my athletes actually not even to look at it until you're past 20 minutes for the best information that you can have out there. You need to do actual field testing of key power durations to use your average power to set training zones and to know your FTP. Normalized power just helps to tell us how hard the days were, better than the average power alone, and on a more holistic view. One final way athletes misuse normalized power is thinking that it is performance itself. And it's not. I mean, I track normalized power for 20 minutes and longer for all my athletes, and I use it to give better insights on how they're holding up in the races and in handling anaerobic efforts for sure. But in a race, you can pump out the biggest NP of the day and still lose. Not that that's ever happened to me before. But big NPs usually give some big fatigue. So if you're going for the win, my advice is to do as little as it takes to win. That might be giving your everything at the proper time, but it also might be less effort than you thought you actually had to give in order to win the bike race. Now, there's probably more ways that athletes are misusing it, but now let's look at how to properly use normalized power.

Using NP Live In Training

SPEAKER_00 8:43

Now, here I'll focus on some fun aspect of it, actually using it in real time in training. First step to this is you need to display it on your screen. So there is a function of normalized power on your Wahoo Garmin or Hammerhead, whatever you're using, but have normalized power displayed. A lot of people don't actually even do this, and it's kind of wild to me. And if you don't have normalized power displayed, then you're missing out on some really valuable information. My challenge is if you're not doing that, start doing that today. Go to your device and put normalized power on there and hell, while you're at it, put on kilojoules too, and I'll probably do a whole video about that later. Now, having it displayed, this gets you familiar with what 200, 250, 300 or X wattage, whatever normalized power starts to feel like. You can then associate what hard and hilly is versus steady threshold riding versus a punchy sprint day from a normalized power standpoint and see a better representation of the cost to the whole body. Remember, NP gives us better insight as to how your body feels when the power is up and down, on and off, hard and punchy. The first step is getting aware of how these normalized powers are feeling, and then you can take it to the next level. Then you want to look back on some of your previous Harv rides, so your peak 1 hour, 2 hour, 3 hour, and 5 hour normalized powers and get a sense of your limits for these durations. This will help you design some specific riding and training sessions to prep for the next long mountain bike race, road race, or gravel race. Now there's some good ways of weaving in the specificity of training using normalized power. When you look at the past hard group rides or races that really challenged you, identify the normalized power of what these were for the whole duration, then try to replicate that on a training ride within reason. This is good for people who hit peak normalized powers in races but never do it in an actual training session. You need to work on hitting higher normalized power solo, and you'll be way more durable and way more comfortable in the next race. Additionally, for all you gravel racers out there, this can help you on pacing for five hours or more. Check to see what you have done in the past, then try to replicate it, or maybe even try to beat it. I have key hard training sessions like this with my athletes that may be harder than racing itself, so that when they get into the race, we can focus on the game of it, not just the effort of it. We're more resilient, more durable, and we know that we can handle it because we've been there before in training. These are actually some of my favorite workouts. It's less structured, it's hard, and kind of in the moment type of riding. It's excellent prep work for group rides and races. Let's take a

A Practical NP Workout Example

SPEAKER_00 11:36

look at an example. Maybe there is a past race where uh you hit 300 watt normalized for an hour that had short, punchy hills that lasted 30 to 40 seconds in it. A simple workout then would be to warm up for 20 to 30 minutes, then hit a hard one hour effort, throwing in 30 to 60 second efforts and aiming for around 280 to 300 watt normalized for that one hour. How hard are the surges? Not max, but they're pretty hard. Try to think of how hard you were pushing in the race and go at that slightly under. Perceived effort should be eight to nine for the RPE on that, or like I said, close to what you were doing in the race. Feel it out. Because it's a solo session and motivation may be a touch lower than in the race, you can make the target normalized power window a touch lower, especially if this is your first time doing a workout like this. Punch on the hills, throw in attacks, pedal on the downhills, steady hard in zone four, then attack again. Just make it up as you go, ride hard and watch the normalized power after 20 minutes. Attack more or go hard to drive the normalized power up. Do a little less if you're overpacing. For example, if you're seeing that like your normalized power is 330 watts after 20 minutes and you're targeting 300 watts, ride easier or coast a bit, then bring it back down to your target. Finally, come back up with some harder efforts, aiming for the desired normalized power that you intended to originally. The overall goal is to guide yourself into unstructured training to hit a target normalized power by feeling the effort as you go and learning how this number is influenced by short, hard, and long hard efforts as well as coasting in your style of riding. So you could start to develop normalized power guidelines for yourself in some sense. I do this for my athletes, and it's all very individualized based on historical past history of what they have done and what type of rider they are. In my coaching practice, I look at peak 40-minute, one hour, two hour, three hour, and five hour normalized powers, and that's the top-end target that I try to hit for these workouts. A low-end target could be maybe five to 10% lower, depending on the rider type, the history of the training, and actual duration that you're doing, giving them a window of normalized power to aim for on these types of races. So it's not just a one number or a max normalized power that you've done before. That window and that wiggle room is good from a training standpoint. Now, I can't give you a specific percentage of FTP or normalized power or anything like that here because it's too individualized. My goal is to start giving you new ideas of how to use this metric based on your past history to simulate hard training and prep you for better races and event outcomes in the future. Now, don't misunderstand me here. I'm not saying to use normalized power for your ranges or in place of your training ranges right now, or normalized power should replace average power at all. I'm not saying that, not one bit. But you can use it alongside of average power, especially on longer and longer intervals and efforts to get a sense of how hard this actually is, especially on this variable terrain. Additionally, you can make normalized power in many ways shorter, higher intensity, or longer, still hard efforts, but more paced out. So when doing a hard NP training session, you should plan ahead of time if you want the hard efforts to be more like 10 to 30 second sprints or 30 to 60 second attacks, or maybe two to five minute hard VO2 work, all with a mix of upper zone two and three throughout. That's a hard day for sure. Main idea is to get you out of this interval mode, thinking that you need recovery intervals every time after you go hard, and to teach your body to deal with the fatigue as it stacks up in your legs and in your brain. The real beauty behind this is that it can help you to teach your pacing for a stochastic or undulating efforts really, really well. And you know what? I can't think actually of a more annoying thing than someone in a group that can't pace for the on-off efforts without just like super surging hard ahead of the group. It's so freaking annoying. But NP workouts can help you feel the cost of the effort to rack and stack over time and make better decisions about pacing so that you're not always surging and causing chaos in the Peloton. On top of that, some of you commented on a past video I did of how to use group rides for training. And you said that you always ride alone because you don't have a group ride to jump into, and some of the stuff I said wasn't relevant. Well, if you don't have a group ride, you can use these normalized power training sessions as ideas to simulate some of the effects that happen in a group ride, like the on-off, hard, easy, and spontaneous time periods, and you'll be dialed by the next time it comes to race day. Now, I'll leave it there for now on the training side of things because the next steps for you is to take this information and go try it. And don't be afraid to fail and then try it again. If you do, tell me about it and try an NP workout and drop it in the comments of how it went. Okay,

NP For Pacing In Races

SPEAKER_00 16:58

so now quickly, let's talk about how to use normalized power in racing. Some of this is similar to the training concepts that I was talking about before. But first, when you're going for max efforts, it's important to know your top three performances for your peak power durations. And I'm talking about normalized power durations. You can see in the chart that I made here, starting really at 20 minutes, I'm tracking my peak normalized power of the past year. So I know my max limits of what I've done. If you track normalized power like I do here, you get familiar with these numbers and you know where your 30 minute, 60 minute, 2 hour, 3 hour, 5 hour maxes are. As I already said, this is super helpful to know when you're trying to pace for longer distances like gravel racing, marathon mountain bike races, hilly road races, and the effort racks and stacks over time, fatigue adds up. Heart rate can drift up or down, but knowing where your NP limits are may help as an additional guide, especially on the longer days like a BWR, unbound, or something like that. You can also use it in crits and circuit races where a peak 60-minute full gas is hard, especially on the techie punchy courses. I'm not saying to use your power meter to make every decision about racing, quite the opposite here, just race your damn bike. But if you're feeling good, pacing like a mad person, check in with that NP display after about 30 minutes and make sure that you're not like 30 watts over the peak normalized power all time. You might need to stay off the front for a bit if you're feeling a little fatigued or get off the gas for a while. Or maybe you're really on one and you just keep sending it. But you can use normalized power as well as your internal rate of perceived effort to guide you in the short crits as well as long gravel races. The key takeaway is to use normalized power as one of your guides, not just as the only thing to make decisions about pacing throughout. In the end, you make the decisions. You're just using information to help guide that decision. And hopefully you land on the podium if you're doing it right. You can also use normalized power in historical data tracking. I tell my athletes that performance can be measured in three ways. The first one is increased average power durations. Second, increase your speed. And then third, you get on the podium. If you got that top step, that's a good performance. The concept is to increase your mean max power or average power, control your body weight, be arrow, and increase your speed. Learn race tactics, pacing, and fueling strategies that will make you faster over time. Then you put it all together on game day to get yourself on the podium. Mean max or peak average power should be the number one thing that you track for your performance. But tracking peak normalized power for key durations is helpful so that you can see the context of bigger days. Again, because most bike racing has surgery components to it, using peak average power to set your training zones and peak normalized power to guide you in some specific training sessions is the ideal way to do it. In the end, whoever crosses the line first wins. Sometimes it takes everything you got, sometimes not. Knowing how you got there and what it took is really helpful when you're looking back on it historically. NP plays a big role with that, especially in the on-off efforts. Finally,

NP Versus Weighted Average Power

SPEAKER_00 20:25

let's talk about how normalized power differs from weighted average power or adjusted power. Main difference are just slight tweaks in math, at least from the looks of it. All weighted averages do is they weight higher power values versus the lower power values. Recall that this is pretty much what normalized power does. Basically, if you upload the same file and compared normalized power to weighted average power, you'll see that whatever platform you're using, or two of them, some some of them may be the same, similar, or totally different. So which is better? When they're estimating and using similar numbers, and the normalized power and the WAP is like 1%, I really don't care. My thing is this normalized power is a trademarked algorithm that is public and well established across tons of platforms. And I'd rather know the math that I'm using to analyze and model my data versus guessing or being in the dark about some other weighted average power or some super specific secret proprietary thing that they're claiming is better. So to

Final Rules And Next Steps

SPEAKER_00 21:33

wrap this thing up, I hope you learn more of how and how not to use normalized power when training, racing, and pacing, and how to analyze your data post-ride. In the end, it's a very important and valid metric that's likely nested deeper into the system that you're already using more than you realize. So it's helpful to know how it works, what its limitations are, and some practical guidelines around it. And finally, it's never just one metric or one number or one thing that you should look at when getting information or when making decisions about your training. And I know that's confusing. So if you need more advice beyond what this video provides, that's where our CTS coaches can help out. Head over to trainwrite.com and click on free consult, and that'll get you started down the right path with a great coach to help answer your questions and get you set up to train right.