I have a reputation, perhaps deservingly so, a scientific coach. The irony is that while I use principles and blend knowledge and science into my coaching practices, it’s not like we’re sitting here measuring VO2max, or even caring about it honestly. As a person, I love crunching the numbers and the data that comes with it, but as a runner and coach, I kind of despise wearing a GPS watch and wouldn’t be caught dead wearing a HR monitor.  These two inner self’s battle each other a lot, but the practical one generally wins out.

Which brings me to my point. Although I love all of the gadgets, there are only a few that make it into my mainstay, like RunScribe described previously. The ones that make it give me actionable data that is easy to track and measure.

The problem with most data is that it’s not actionable or its a pain to get people to use it. My goal is always to find ways to better my own coaching practice and understand things that might influence my athlete’s performance. Instead of hooking my athletes up to machines and making them lab rats (okay…we do that too…but only for research purposes!), I like finding ways of answering questions with data that is easy to take and understand.

This year, I wanted to answer the question of what actually matters and how does it change?

I preach recovery and living the lifestyle of a runner to my team, and we know it makes a difference. But what I wanted to get at is how do those markers change over the course of a collegiate season. With college athletes, you are dealing with a wide range of stressors, so it’s not simply go do the workout and then wait until they come the next day to practice.

What I wanted to get at is how does that change and can it give me a better understanding of how to modulate the workouts.

Understanding Data:
Over the next few paragraphs, I’m going to give you a sneak peak of the data that we collected. Why? Because I think it’s really cool. And secondly, I like sharing things. Coaches who don’t share and think they have a secret are ignorant of the process of coaching. I haven’t had a chance yet to go back and do a comparison with workload, workout type, races and so forth, which is the next step.

So what we did was simply create a nice little app where athletes could rate a variety of measures on a 1-5 Likert scale. Yes, it was all self-reported data, but again we are talking about something that college kids will fill out. Secondly, my college team is awesome. I don’t have to worry about them filling out fake data or worrying about their coach seeing their sleeping habits. There’s a level of trust. I had kids put 3hrs of sleep down and although that might not have been optimal, it was refreshing to see that they were honest about it and then tried to get better at it.

We tracked a range of things including:

  • RPE (Physical exertion of workout)
  • MPE (Mental exertion of workout)
  • Stress level
  • Energy level
  • Soreness
  • Pop (springiness/how the legs felt)
  • Sleep quality
  • sleep hours
  • Overall performance (1-3 simple scale…below avg. average, and above average)

Stress and Sleep:


This is one of my favorite graphs because it almost looks like a mirror image of each other. There is a clear trend to see reduced sleep correlating with increased stress levels. Now we can’t make inferences based on causes, but since the stress levels increased the same day as the sleep decrease, we can assume that it was the sleep most likely playing a role in impacting the stress instead of vice versa. Why? Because the kids were reporting the previous night’s sleep and tracking their current day’s stress. If the stress was causing reduced sleep, we’d see a one day delay, which we tend not to.

While it’s not profound, it’s still fascinating to see the relationship between sleep and stress.

Pop and Performance

I like talking about this idea of muscle tension that Marius Bakken first introduced me to. To quantify a subjective feeling of what that is like, we tracked the idea of “pop”. We see pretty clearly that pop relates to how they graded their overall performance that day. It makes intuitive sense, but what this indicates to us is that at least how they feel about their “pop” in their legs seems to impact how the workout goes.

It’s good reinforcement for understanding that we need to get the athlete’s legs feeling right on race day.

The ultimate goal is to go back and look at what workouts did to “pop” to see if we can notice a trend and see if we can further refine our idea of getting athletes to feel good on the right day.

Physical versus Mentally Demanding Workouts


Here we have a tracking of their physical exertion versus their mental exertion for runs and workouts.

The first thing to notice is the modulation, you can clearly see the delineation between hard workouts and easy runs. Additionally, you can see variation in how demanding different workouts were. It’s not simply hard and easy. There’s a spectrum of physical and mental efforts.

The other thing to notice is that while the mental an physical effort of the workouts are synced, for the most part, there are differences. There are workouts that are entirely more physically demanding then mentally demanding. It will be interesting to see which ones these are and whether athletes tend to think a particular workout is hard or easy.

You can also get an idea of periodizing efforts throughout. The two major increases correspond to two of our bigger races

Ultimately, I’m going to track our main workout types and see how they fluctuate across the season.
Sleep Hours, Quality, and Energy Levels


This time we get to look at sleep hours in addition to quality and compare that to perceived energy levels. While not as robust of a trend, we can see that energy levels do tend to mimic sleep quality and hours. Once again, telling us that perception of an athletes energy levels is tied to how they slept the night before!
Here we have a simple graph tracking our soreness throughout the season. What we hope to do here is track it with workout type and see what causes or reduces soreness. Another thing, I’d like to highlight is that there was a downward trend across the season, which makes sense. With the exception of one anomaly towards the end (which then dropped big time the following day), it’s a nice negative trend. Showing that, while we were doing workouts that were still bringing some soreness, we were eliminating that pretty quickly and the athletes were adapting to the workload.
Although I’ve highlighted this in another picture, it deserved one on its own. While it will be interesting to come back and correlate it to practice times, travel, races, and so forth, it’s worth a look on its own. This is the reality of sleep for college athletes. And overall, I’m pretty pleased with it. They’re consistently trying to get around 8hrs a night it looks like, and our average is right around 7.5hs. Additionally, there are some pre-emptive higher quantity sleep periods leading in to big races or workouts which is nice to see. Additionally, it looks like there’s an attempt to sleep a bit more as the season goes along. At least, you lose some of the lower sleep nights.
Some of those low sleep nights are actually following races because of the night race component of some our meets, or the travel or long run the next morning. So that is something to look at and try to figure out.


Lastly, I’ve shown this one before, but on its own you can see that there’s almost a slight U-shape quality to sleep quality. You don’t see this in the amount of hours slept, so it’s simply a perceived quality of sleep effect. What’s interesting is that in the middle of the season and semester, you see a slight drop in sleep quality before a slight trend to increase before the end of the season. While I haven’t looked into why it occurs, it makes sense that with mid-terms, being in the middle of the season, training hard, and many other factors likely play a role. It’s just something to look out for with collegiate athletes.
So what?
As I said, my hope is to delve deeper into these numbers and look at how the trends relate to our workout volumes and intensities.
It’s nothing revolutionary, but seeing the importance of sleep, recovery, and feeling of pop reinforces and refines some of the practices that we do.  If nothing else, it’s fascinating to peer into what training hard at the collegiate level really looks like in terms of physical, mental, and emotional response.


Get My New Guide on: The Science of Creating Workouts


    1. James Marshall on December 5, 2014 at 9:42 am

      Nice use of the data Steve, and you are right about the pain about collecting it.
      I did a similar thing with rugby players a few years ago (before the t'internet) using the REST_Q (http://www.pponline.co.uk/encyc/underperformance-syndrome.html) looking at the balance between stress and recovery.
      Intereseting finding of yours that stress (or perception of stress) follows poor sleep, rather than the other way round. So our ability to deal with life's stressors is reduced with lack of sleep.

    2. Ali husain on December 5, 2014 at 2:06 pm

      Have you used any form of heart rate variability to check recovery? Examples are the Rosco test or First Beats technologies software.

    3. Unknown on December 5, 2014 at 2:31 pm

      Very interesting information. You may have mentioned this in a previous post but is the app you used available to the public?

    4. Vynnadoc on December 6, 2014 at 4:34 pm

      Great information, Steve! Collegiate athletes have unique challenges. One important missing piece is nutrition, especially fueling and recovery. Much harder to collect than some of the other data, but can be done. Would love to discuss with you. Dr. Sue Kleiner drsue@vynna.com

    5. Enrique Cafuentes on January 7, 2015 at 7:22 am

      Does anyome have experience with VERTIMAX training product? Does it help with training how effective it might be? Does it help with speed?

    6. How to Crush Crits and Be a Better Racer on July 29, 2017 at 9:44 am

      […] Tracking collegiate runners- sleep, stress, soreness, recovery and performance […]

    Leave a Reply