The power of analytics – lessons from a world-class coach

Christoph Papenfuss
Cycling is a tough sport

Cycling Power Meter

A few months ago, I wrote about my personal use of analytics to improve cycling performance. This post continues to be quite popular. There are indeed a lot of parallels between the use of analytics in sports and in business. So, I thought why not learn a few things from one of the leading experts in this area? Why not learn from somebody who is driving the use of analytics in sports? Why not listen to the person who has developed one of the first software packages for analyzing cycling and workout data?

Please meet Hunter Allen. He is widely known as one of the top experts in the world in coaching endurance athletes using power meters in combination with powerful analytical software. Hunter and I met almost six years ago. He hosted a cycling camp in Solvang, California. During those seven days out on the road, he taught me a lot about using analytics to improve my cycling performance. And I was lucky to catch Hunter over the phone the other day.

Christoph: You started your career in professional cycling. Professional cycling, like other sports, has a lot of old-fashioned traditions. What led you to take a fresh look at the traditional training methods?
Hunter: Actually one of my coaching clients asked me to train them with a power meter and I figured what the heck, I can do that! He was a VERY early adopter of the technology and that forced me to really learn it and figure out how to understand what it meant from a coaching, sports science and racing perspective.

Cycling Power Meter
High Tech in Sports - A Cycling Power Meter

Christoph: You advocate the use of high-tech training devices like power meters for cycling, GPS and accelerometers for running. These devices record a lot of data about each workout: heart-rate, power output, pace and cadence to just name a few. Why record so much data? Shouldn’t an experienced athlete know how to listen and react to his/ her body?
Hunter: We have to listen to our body, but sometimes our bodies don’t tell the truth or can’t give us enough information to help us make the right decision. For example, this past summer, I was training for the Masters Nationals races and I was in the last two weeks of the hardest part of the training. Normally my Heart Rate would be in the 170-174 range when I am riding at threshold power (equivalent to your highest avg. power you can maintain for an hour), but my cardiovascular system was quite fatigued and I could barely get my heart rate to come over 158bpm. However, my watts were higher than ever and I was doing more work than I had in a long time. Had I been going on heart rate, I would have rested and denied myself optimal training. The more data the better, because it allows us to organize it into useful information from which we can make correct training decisions.

Christoph: Back in 2002 you started developing a popular and sophisticated software package called WKO+ (Workout Plus). It allows athletes to upload the data from their power meters and running watches. What happens with the data in the software?
Hunter: Our Software is no more than a fairly sophisticated program that can take multiple ‘channels’ of data, organize it into charts and graphs which are presented in an easy to understand fashion. One of the keys to this is the ability to chart the data over any time period you want to chart it. So, I can tell if an athlete has improved in their sprint power over the last year, last six months, last month or even week to week. Data is useless until you can intelligently turn it into information and we’ve done that through a visual software package that allows the user incredible flexibility.

Christoph: Can you give us an example of what athletes can learn from the huge amount of data? Can even highly experienced athletes that really know their body well use these insights to their advantage?
Hunter: One of the greatest things that an athlete can learn is their own personal relationship between a ‘training dose’ and their bodies ‘response’ to that training load. Once they know that two weeks of training in the mountains and one week of rest will give them the fitness needed for a race, then they can begin to plan and predict a peak performance. This is the biggest revolution that we have been able to create: Peak performance prediction. If I have enough data from an athlete, I can plan their training nearly perfectly so that on the exact day or week that they want to have the best ride of their life, they will. That’s “powerful”! 😉

Christoph: Athletes often have to make critical decisions about their training schedules, their race tactics and their general behavior out on the road? Does the information provided by your software help athletes make better decisions? If yes, can you provide an example?
Hunter: For sure, this is something that can help them make decisions. Some of the best data is their race data. We can learn if they are pedaling too much for example! It seems counterintuitive, but the best, most winningest road racers often pedal the least in races. Cycling is a sport of energy conservation and the best racers save the most energy and then use it when they need it. Another example could be in a long solo ride, as a rider might use their power meter for pacing their energy output so that they can finish strong.

Analytics in Cycling - Software provides critical insights

Christoph: Athletes often work as a team. The team includes coaches and other athletes. Does your software make a contribution towards better communication and collaboration between these different parties?
Hunter: One of the benefits of using these devices is the added sense of accountability that an athlete has, since they have been given a workout prescription and if they don’t do it, then when they upload their completed training file to me, then I’ll know exactly what they didn’t do. So, it helps the athlete with program adherence and the coach with communication. Without this information, I can only guess if the athlete is training correctly.

Christoph: We see more and more athletes train and race according to your methods. Would you say that training and racing with a power meter can lead to a competitive advantage?
Hunter: It’s a competitive advantage if you learn how to use it properly. Just sticking one of these on your bike isn’t going to make you a faster cyclist. You have to know some of the basic power training principles and then follow them. If you do that, then yes, absolutely it’s an advantage.

Christoph: The world of sports is often characterized by a culture of old traditions. Do you find it hard to promote these new training approaches?
Hunter: I don’t find it hard, because I am by nature a very flexible and open minded person. I am always open to new ideas, thoughts and approaches, but not everyone is. Any change takes time and here we are nine years after our first software version came out and still I am teaching new coaches, new athletes about it. I look forward to learning about the next tech tool that will help my athletes as well!

Thanks for your time, Hunter!

About Hunter Allen:

Hunter AllenHunter Allen’s goal has always been to teach athletes how to maximize their training and racing potential through professional analysis of their power data. Hunter’s power training method has built success at all levels of cycling and endurance sports, training such well known professional & Olympic athletes such as Jeremiah Bishop (Volkswagen-Trek), 2008 US National Champion Mountain Biker, Daniel Lloyd (CerveloTest Team), 2008 Vuelta de Extremadura, Sue Haywood, 2007 World Mountain Bike 24 Champion, Dan Fleeman (CerveloTest Team), 2008 Winner of Tour of Pyrnees and with the 2008 USA Olympic BMX Team. Hunter is himself a former professional cyclist for Team Navigators and has raced for over 17 years in Europe, South America, U.S. and Canada and has over 40 career victories to his credit. Considered a great all-rounder, he was able to learn a wide variety of race tactics and skill necessary to succeed at the professional level.

Analytics Sticky

The analytics of cycling

Where is my auto-pilot? The Dolomites are stunning.

A few months ago, I posted an article about analytics in cycling. This post still remains one of the most popular posts on my blog. Thanks for all the great comments, feedback and questions. Many of you wanted to find out more. Here is a quick report of how technology helps improve cycling performance. On Saturday, I had the opportunity to ride the famous Alpe di Siusi climb in the Dolomites. It is a super nice and friendly climb: breathtaking landscape and not too long. It is fairly steep (average of 8.5%) but it is steady. In other words: a perfect opportunity to fit a short but challenging ride into a constrained schedule.


High Tech in the hub: My Saris Cycleops Powermeter

A few years ago, powermeters became an affordable training tool for amateur riders like me. Smart technology is packed into either the bike cranks or the hub of a cycle and it measures a lot of data: power output, cadence, speed, torque, etc.. The data is sent to my Garmin Edge 705 cycling computer via ANT+ (similar to bluetooth) where it is combined with further GPS and altitude data. Following each ride, I download the data into a smart software called WKO+. This is where I can analyze each ride and learn a ton of stuff about myself.


Blue line: fitness, pink line current training load, yellow line: freshness

Before leaving the house, I checked my performance chart in WKO+. This chart literally calculates and predicts my fitness using smart algorithms. The blue line for example shows my fitness ramp up. You can see that I had a rough spring: I was injured and could do not much until March. The yellow line indicates my ‘freshness’: a value below zero indicates that my legs are likely to feel heavy. A positive value indicates that I should have fresh legs. These algorithms work amazingly well and it really helps me put a solid training schedule together. For Saturday, I can see that I should have somewhat fresh legs (value is around +8). Perfect timing!


An amazing piece of technology: The Garmin Edge 705

The climb went pretty well. Recent tests in training sessions have shown me that I am starting to get back in shape: I can easily sustain around 260w for about an hour. Wattage higher than 280 is likely to cause me some trouble (I experienced light cramps in a training session a few weeks ago). That information helps me with pacing. The climbs in Italy are pretty steep and it is easy to get carried away and push too hard. But after a few minutes I realized that the recent training efforts have indeed helped me improve and so I settled into a higher than normal pace. And this strategy worked out pretty well: I felt great and strong throughout the entire climb. I did have to hold back a little as it was getting extremely hot and I still had another climb across Passo Pinei.


Back home in Munich, I downloaded the ride into the software. How did I do? First, I take a look at my power distribution. A quick report shows me how much time I spent in different calculated training zones. The two zones on the right are the ones that really, really hurt. This is where your legs & lungs are screaming in pain. Hmm…Looks like I had a good day based on that. It seems that I have made some good progress. Time to increase the overall training intensity?

Time spent in different training zones

Let’s take a look at the detailed ride. The orange line shows the altitude profile. You can see that there were two different climbs. The other curves show the important stuff: power output (yellow), pedal cadence (green), speed (blue). Looks wild, doesn’t it?

Orange linge: altitude, yellow line: power output, green line: pedal cadence

But I am interested in my climbing performance. So let me zoom in on the Alpe di Siusi climb. To make it easier to analyze the data, I am drawing a few lines in here: the yellow dotted line shows my current threshold power (the power that I can produce for about one hour without falling off my bike in exhaustion) along with my favorite cadence zone (anything below 60 rpms does not feel good…). Aha. Looks like I had a good ride. Training is paying off indeed. I consistently rode above my estimated threshold and even had lot’s of energy to spare at the end of the climb (look at the yellow peak). Awesome.

The Alpe Suisi climb

The software also calculates several important and helpful KPIs. I won’t go into the details, but these KPIs help me further manage my training schedule. One easy example is Kilojoules:  I created energy worth 732 KJs (the equivalent of two Snickers!). That information is really helpful on longer training rides: it helps you manage food intake.


Perfect. Despite not having much time, I managed to get a solid ride in. The powermeter helped me pace and push myself. The post-ride analysis showed that I need to adjust my training schedule: time for higher intensity workouts. Also, I will have more time to ride in the Alps over the next few weeks and will use this knowledge to ride some of the longer and harder passes.

It’s fun to ride this way: The analytics have allowed me to learn a lot about myself. I ride more consistently now. And I perform much better on climbs. It feels awesome when you have some energy left in the tank at the end of a climb: this time I raced a local delivery truck. Italians love cycling and the two guys in the car cheered me on while pushing the gas pedal a bit more. The sun was shining, the clouds opened up on amazing view and I felt like I could fly right in that moment.