Introducing Effort with TRIMP
Modeling fitness, fatigue and performance with your workout data
While I've always enjoyed trail running, I was recently lassoed into running on a track at 6am. I was entirely unprepared for the exacting nature of structured intervals: hitting splits and timing individuals laps as opposed to a single continuous run.
One of the most disturbing things was seeing a wide variance in my pace. I was always looking for a straight or descending pace throughout the workout, and track workouts produced a graph I didn't understand. Plus, the map is an oval, which is simply unhelpful.
Still, I wanted to know how hard I was working; how difficult the workout was; how much effort it took to complete. Pace alone didn't tell me this. Between the elevation change, the terrain, my physical health and even the weather, the number of variables were too great to be able to compare workouts easily.
So we're excited to point out Effort, a new feature that appeared last week.
SportTracks users have a wide variety of data quality. Some use a GPS watch, some have power meters, and some use nothing beyond a stopwatch. We work hard to make sure that each level is served properly.
So we use the available data to calculate the most accurate Effort possible. If there is only time and distance, we'll use simple algorithms to calculate an Effort for that workout. If there is a GPS or distance track, we can better estimate an effort over the time-series data. Add elevation to the mix and we use level pace rather than flat pace.
If heart rate data is present, then we can use more complex and accurate algorithms. We will incorporate zones and heart rate reserve as valid data is available.
(While pace is an measure of the indirect result of effort, heart rate is particularly interesting because it's a measure of the physiological response to effort.)
There are a few great reasons to look at Effort.
First is to make sure the workouts are achieving the necessary levels in your training program. Both under- and over-training are serious concerns in the battle to reach our goals. If you under-train, you're may not hit your goal. If you over-train, then you're at risk of injury.
Another great reason to look at Effort is to compare previous and current workouts. For example, I ran the same loop twice in about the same time, but my Garmin 310XT logged a distance of 2.80 miles, while the second recorded a distance of 2.95 miles. This discrepancy could be misunderstood by looking at the paces, one of which is underrepresented.
However, if I look at the Effort of 22, it makes sense. A similar run within a few seconds of each other should be about the same amount of effort. This is because the Effort is calculated by using heart rate instead of pace.
Digging deeper into TRIMP
As the saying goes, we are dwarfs standing on the shoulders of giants. There are smart people before us who tried applied scientific method to the physiological response to effort. One of the first models was developed by Eric Bannister in the 1970's, who named it Training Impulse. It was quickly abbreviated TRIMP, and remains a reliable metric.
Calculation of TRIMP has seen a number of developments that coincide with progression and availability of biometric sensors. With the dropping price of GPS watches, heart rate sensors and other dedicated devices, more people became interested in using and understanding the technology.
There is an excellent reference maintained by Jon Savage at http://fellrnr.com. If you have the interest, there is a wealth of detailed information on using data-driven science for long-distance training for runners, much of which applies to the general endurance athlete.
In addition, DCRainMaker did a fantastic blog post on his experiment looking at the accuracy of GPS watches, one of which is the Garmin 310XT.