Data Sets for Athlete Monitoring by Carl Valle



The core problem with most monitoring strategies is that they are reactive as opposed to preventive and focus on poor strategic approaches for interventions. Most coaches can benefit greatly by combining periodization (planning) and monitoring (measuring) into a detailed performance model (success strategy). A simple way to do this is to approach modelling as a composite of pushing biological limits and good risk analysis, which means gathering details or inputs.

Though no matter how dedicated they may be, athletes have a finite bandwidth and attention span for giving data to coaches. That’s why it’s best to focus on sustainability and the essentials when collecting data. By doing more with less, the athlete is less likely to fall off or get bored and disenfranchised with the process. Here is a rubric on data sets to look into and what to do with the information once collected as I find the synergy of data priceless.

Heart Rate and Heart Rate Variability (HRV) Monitoring

Comparing resting and training heart rate with HRV is a very powerful way to observe patterns of fatigue every day, and apps like ithlete Pro and HRV4Training are growing in popularity. Most users of heart rate data who care about strength and power should look at weekly and monthly trends, since neuromuscular fatigue can lag or rear its ugly head at different daily courses than automatic nervous system (ANS) responses. Users of Omegawave may see data that overlays with DC Potential which is a good indicator of fatigue, but direct measurement of power has no filter which makes VBT priceless.

The trick in analyzing heart rate is to use general condition sessions to clarify how repeated work is responding physiologically. I use five field tests that can be done nearly anywhere as a way to see basic trends, like if an athlete is still needing time to catch up to baseline.

The trick in analyzing heart rate is to use general condition sessions to clarify how repeated work is responding physiologically. I use five field tests that can be done nearly anywhere as a way to see basic trends, like if an athlete is still needing time to catch up to baseline.

Monitoring the heart rate data of training is a lost art. Before body load metrics became all the rage Banister was modeling performance with heart rate for decades. While the variables are extremely limited (minutes and HR), the elegant inception of modeling is still useful for getting the training impulse (TRIMP) for those wanting to compare conditioning data to Velocity Based Training. If one is not collecting all training load data, it’s hard to adjust weight training or tone down workload to maintain leg power.

Three key points are listed here to ensure coaches have the right balance of aerobic training and power development. Managing the strength training data with VBT indices, as well as simple conditioning heart rate metrics, will enable coaches to fine tune workouts more effectively.

  • Low HRV can stem from too much volume, or is the response from riding the horse of both intensity and density a little too long. Decreasing the frequency and total volume of intense maximal work is a great way to improve without resting too cautiously.
  • A high training heart rate, or lower than expected recovery rate, with conventional heart rate training and familiar workouts usually spells overreaching or overtraining if the problem is chronic. Sometimes dead legs from too much speed and power training can turn recovery workouts into training sessions. It’s important to come into each training phase knowing what is acceptable and normal, and what is potentially digging a deeper hole. Nothing’s wrong with training hard and going through heavy training periods for deep adaptations. However, only when athletes are prepared and show consistent improvements in training can the aggressive workload create a benefit in the long run.
  • An unstable resting HR is common with speed and power athletes, but too much variance could indicate that one has a low work capacity. Variability is necessary, but the ranges should not be excessive. If one is seeing a high variance of HR and low HRV score: reduce the work slightly and see if things trend up. If one is seeing a very static set of data points and the software is warning of something wrong, even if the data is trending up there could be something going on, or there could be cause for concern. A good resource to explain why sometimes paradoxical data can arise in training can be found here courtesy of Andrew Flatt.


Wellness Questionnaires

Subjective data is not going anywhere soon. The primary drawback of this method is that asking repeated questions is boring and athletes may give halfhearted responses. The solution is not rocket science; give back more information than they give to reinforce value, and actually use the data to drive programming. Don’t just use this an alarm system for problems.

Timeline view in the ithlete Pro web application

Timeline view in the ithlete Pro web application

The absence of data is actually information. Lack of compliance is often a symptom of poor athlete wellness as seen with the ithlete dashboard. The highest wellness readings matched both compliance and positive performance in training output and physiological recovery.

  • Body soreness scores need to be in flux. The most likely periods of accuracy are when athletes are training heavily or during recovery periods. Not seeing trends of soreness with bar velocity scores is normal, but tapered or rest periods should have an inverse relationship with low soreness and high outputs.
  • Energy levels and RPE (perceived exertion) isn’t perfect, but sometimes they can show an athlete who is digging themselves into a hole and isn’t responding to the work given to them. Cross validating perception and actual work done objectively is a great way to calibrate who is tough, and who might be a little bit lazy.
  • Mood Scores are the most eye-catching to sports psychologists because they can relate to the attitude towards people or training. Usually the body becomes stale or bored before the athlete is, so look for training monotony and the need for a change of scenery.
  • Sleep Quality and Quantity is still valuable to do. Most sleep sensors are not tracking REM periods because most consumer or enterprise devices are not using brain waves to get the necessary data. Athletes who are sleeping longer may not always be a good thing as illness like mononucleosis can improve “sleep scores” but are definitely not enhancing performance!
  • Diet Indicators are also important for athletes who are training multiple times a day and can’t create a natural rhythm of “rest and digest” recovery periods after “fight or flight” workouts. Athletes who lose appetite and lose weight, or athletes who are hungry all the time might be low on glycogen stores, so watch diet scores carefully.

Wellness questionnaires are still relevant even in this day where technology is prevalent and growing. Another instrumental area that needs more monitored attention is short logging what athletes input daily. A few sentences really can really capture the nuances to give more meaning to the numerical questionnaires.


Due to the infrequent rhythm of saliva or blood testing it’s a smarter idea to get deeper with testing rather than more frequent because one can use non-invasive options from physiological monitoring. The soul of blood testing is to screen out the known and focus on new, actionable interventions, such as solving for a deficiency or identifying a glaring problem. Power grows in a fertile environment; having the perfect training program for a bad body never works.

Using direct measurements like training with daily physiological monitoring, can give context to why blood and saliva tests might support trends, or reveal limitations to the timing of tests. Biomarkers may not make any sense if taken out of context. Careful interpretation is needed to see why periodic audits of the body make sense, or why data might appear misleading.

Free testosterone is a simple way to see if fatigue is a culprit to poor bar speed. Free Testosterone is a composite metric using total testosterone and Sex Hormone Binding Globulin, adding more resolution to recovery patterns. One test isn’t enough to make a conclusion, so testing no less than three times is suggested.

  • Use the fT:Cortisol ratio and Creatine Kinase to help differentiate local muscular breakdown from outside training influences like mental or emotional stressors. It’s not perfect and requires interpretation, but it’s one of the most revealing processes coaches and athletes can reveal when paired with subjective indicators.
  • In addition to hormones, repeatedly test vitamin and mineral biomarkers in blood to screen out deficiencies that interfere with both aerobic and muscular performance. What is so simple and obvious often is neglected because it’s underestimated.
  • Systemic inflammation markers like hs-CRP, body enzymes like ALT or AST, and chronically low WBC is an indication that training is going beyond the normal limits. Any training that causes slow recovery and has out of range blood values, coupled with poor training data, is a sign of stagnating overload.

As you can see from the above suggestions, teasing out information still requires human experience and is not just data. One problem of face time with athletes is that it’s hard to scale. Technology enhances the opportunity with better talking points, or self-reflection, if one is training on their own. Due to the fact blood testing is so invasive, the conversations between the coach and athlete, or just seeing one’s own data, can be a powerful wakeup call or a pleasant reward for successful planning.