About BeastScore

BeastScore® 101

BeastScore is a measure of fitness in terms of work capacity and is calculated using an athlete's scores for eight scientifically selected exercises. Using work capacity as a measure of fitness has been discussed by fitness experts for decades and in 2009 CrossFit® founder and CEO Greg Glassman said fitness is work capacity.

Work capacity for an individual is calculated by plotting the average power produced during various exercises (see figure below). Each data point (an X on the graph) represents a different exercise. High power / short duration exercises like back squat and clean & jerk (hundreds of ft*lb/sec but only lasting 3 to 5 seconds) are on the left and low power / longer duration efforts like the 20 minute workout Cindy are on the right.

The eight components of BeastScore (represented by Xs on figure for an average male athlete) are clean & jerk, back squat, 500 m row, 1 mile run, and four popular 'WODs': Fran, Grace, Annie, and Cindy. The average power exerted during these eight exercises for an average male athlete is shown in the graph below.

power-vs-time
Average Power Vs. Duration of Effort for the average male athlete. The Xs represent the eight BeastScore components: clean & jerk, back squat, 500m row, Grace, Fran, 1 mile run, Annie, and Cindy.

The idea behind measuring fitness this way is that you are as fit as you are able to sustain high power output during many types of exercises for many different time domains from short (3 to 5 second lifts) to long (4+ hour marathons) efforts. Fitter athletes can sustain higher power for longer periods of time, so the Xs on their graphs are positioned higher on the chart. This means that when a trend line is fitted to their scores/Xs, the area under the trend line (the blue shaded area) is greater. The area under the curve is total work capacity, which we call BeastScore.

Calculating power output for the eight BeastScore components involves identifying the masses/weights moved around (including portions of bodyweight), multiplying by the distance the weight was moved, and then dividing by the time it took to move them. However, each BeastScore component presents challenges when it comes to accurately identifying weight, distance, and time. Plus, in order to keep the process of receiving a BeastScore as simple as possible, we must make certain assumptions and generalizations regarding athletes' body measurements as well as performance data such as how long it takes to do a 1-RM back squat.

To calculate the average power during Grace, for example, we calculate the work performed during a single clean & jerk, multiply by 30, then divide by the athlete's Grace time to get average power output during that time. Work performed during a single clean & jerk is equal to the weight of the barbell multiplied by the distance it was moved, plus any portions of the athlete's bodyweight that are moved during the movement multiplied by the distances those portions of bodyweight were moved. Power during Grace is then the sum of those two work quantities divided by the Grace time. The distance the barbell was moved and the entire bodyweight component of work will vary from athlete to athlete and is based on individual anthropometric data such as femur and tibia lengths. However, we use the same height and weight (and other body measurements like tibia length for example) for all athletes when calculating power for a score (specifically 5'8"/165 lbs., the average height/weight of the 2015 CrossFit regionals athletes). This way, if two athletes have the same PRs, then they have the same BeastScore. Otherwise, it is possible for taller and heavier athletes to have BeastScores higher than athletes who actually squat more and run faster.

By using the same height and weight (the actual height and weight chosen is arbitrary) for all athletes, BeastScore is able to show who can get the job done best – who squats the most, who runs the fastest, and who gets up and down the pull-up bar faster. That's what is really important. True individual work capacity is not what's rewarded in competition – what's rewarded is lifting the most weight and being the first to finish the task. So by removing individual traits from the calculation, BeastScore is a true measurement of what an athlete can accomplish, not just how much work they are doing during the process.

We considered adding a new fitness score called Work Capacity, which includes individual traits such as height and weight, but as far as we can tell, this value has no practical application or value. Just in case anyone is interested, the value is reported on athlete profile pages. But it is not tracked over time or ranked on leaderboards like the other fitness scores.

In addition to making certain assumptions regarding total weight and the distances they are moved, we must also make some assumptions regarding duration of effort in some cases in order to calculate power. For example, when recording a new clean & jerk 1 RM, athletes typically record only the weight, not the time it took to perform. After reviewing hundreds of PR videos of athletes performing these exercises, we came up with average times for performing lifts such as clean & jerk and back squat and use those for calculating power.

To produce a BeastScore value, we take the power calculated from the eight components and plot them as shown in the graph above. Then we fit a trendline to the data and integrate the curve to get the area under it. That area, which has units of ft*lbs, is work capacity, which we report as an athlete's BeastScore.

BeastScore values range from about 50,000 ft*lbs (any person who is physically capable of doing the movements has at least this score) to 240,000+ ft*lbs (we estimate Rich Froning at around 245,000 ft*lbs, the highest we've seen, which is also good validation of our model). To make BeastScore values more user-friendly we divide them by 1000, so BeastScores are reported on the site as ranging from 0 to 250.

In addition to our particular selection of exercises to use for the power vs. time graphs, the mathematical models used to perform the work capacity calculations ("area under the curve") are also unique to BeastScore. Our selection of BeastScore components along with the regression and integration methods utilized are the secret sauce of BeastScore and what make it such a revolutionary new tool.

BeastScore Values

BeastScore Examples
245 Rich Froning
220+ Men's Games Champs
180-190 Women's Games Champs
130-150 High School Varsity Football Players
100-120 Athletic Teen 13-15
70-90 Average Kid 9-12

 

* Note Regarding Power During Running

The most confusing calculation for many is power output during running. If one were to simply research "power during running," one would find information ranging from calories burned to force plate measurements. However, metrics such as calories burned and oxygen consumed are related to energy consumed, not useful work performed. After losses due to inefficiencies of the human body, energy out is much less than energy going in. To maintain apples-to-apples comparisons of the power outputs of the various BeastScore components, we must calculate the useful work performed against the external environment during running, and ignore all the other useless work like flailing arms. Some might even argue that the net work during running on a level surface is actually zero since the direction of gravity is normal to the surface while the change in height (distance moved) is zero. Although the net work may in fact be zero, the useful work performed by the body is not. The human feet and legs do useful work against gravity in lifting the body a few inches off the ground with each step. Then gravity does negative work to bring the body back to earth. The small amounts of work done by the feet and legs during each step add up to a significant amount and is on the order of power exerted during Grace or Fran for an average athlete.