Using PowerTool/GymAware: Short Video and Explanation

We have been using GymAware for the past few years, but recently we acquired the Pro version of the software that allows automatic data collection, data-basing, analysis, reports and what else on the cloud. This makes life much easier and also allows for tracking much more than a single parameter.

What we have been using it for the past year is to track mean external power output in the countermovement squat jump with 20 kg. This allowed to see trends with the players and gain some ideas on their freshness and readiness.

Statistical methods used to analyze the trends were rolling averages of the last 6-10 measurement used to get Z-Scores of the each individual. This way we can see trend, but also how much is one away from his normal variability (Z-Scores). The yellow flag was usually set to -1 to -2 and red flag for everything under -2 for Z-Scores. This is used besides visually checking for the trends over time (one could use rolling average to smooth the curve).

One thing with this method is that one needs to accumulate enough data to get more reliable estimates. But that’s always true with anything related to statistic. To make more confident claims, one need better and bigger data.

Random data to show the calculus and visual representation

What we plan doing this year is doing the same, but with Pro software we can track more variables to gain insight which one is more sensitive to readiness changes. Things like mean power, peak velocity, dip, jump height. It is easy because everything goes directly to the cloud. No need to write things on paper.

Besides this simple use, we use it to track improvements in squat with estimating 1RMs on load-velocity profile. I wrote about it here. You can check the short reportage from one gym session in Boson Olympic Center in Stockholm, Sweden.

The beauty of this approach is that it is submaximal (except for the speed of movements which should be intended to be as fast as possible within technical limits of the exercise). There are two ways to assess 1RM – actually finding 1RM or doing reps-to-technical failure and estimating 1RM from the tables (makes sure that the reps are around 3-6). Using velocity approach is novel method and I am still figuring out what is the best way to do it.

Basically one could do 3 reps at 55%, 65%, 75% and 85% as fast as possible. Research states that it is more reliable to track mean velocity instead of peak velocity of the rep. Also, one could use best rep or set average. I would use set average since it is less prone to errors I guess. The velocity difference (fastest – slowest) should be more than 0,5 m/s (talking about mean velocity method) – you might need to change % a bit to make this possible. This is important to get more valid and reliable regression coefficients. One could also play with standard error and get confidence
intervals and thus more magnitude based statistics to assess (real) change over time.

Load-velocity profile

Next comes the selection of the speed for 1RM estimate. It is usually around 0,2 – 0,4 for mean velocity (not peak velocity). This depends on the lifter, exercise, depth of movements, etc. One could also use LD0 (resistance at velocity = 0) but not as real/training 1RM, but rather as an indicator of change happening.

What is nice about this approach is that it could be done ANYTIME. It doesn’t need to detract you from your normal workouts. Could be done with the warm-up sets and some working sets. And using this, athletes get the idea why it is important to be strong to lift fast and be fast/explosive overall. It is also easy to see trends (with strength athletes this might be their monitoring tool to see the effectiveness of training/block and judge when to switch or modify the program – more on this in some future articles). Also, it removes the idea of grinding the weight, especially with the team sport athletes. Only technical and fast reps.

The negative side is of course errors in calculus. I am searching to find the best method (valid, reliable and sensitive) of doing it to assess real changes and correlate it with real 1RM. Athletes should also strive to lift as fast as possible and by lifting light weights slower one could easily “cheat” to get higher estimate, since that would flatten the curve. Anyway, until then I advise not to use estimated 1RM as training 1RM, but rather as an indicator to increase training 1RM. For example, if an athlete uses 140kg as his training 1RM (one that is used to calculate all the weights based on percentage – see more here) and over a training cycle his estimated 1RM at 0,4 m/s improves for 5-10kg, it might be a good idea to increase his training max for 5 kg. This brings me back to developing vs. expressing concept I alluded in numerous blog posts.

Having Pro account of GymAware it is now easy for me to collect data and do my own research to find the best method of load-velocity based estimates of 1RMs.

Stay in touch because more is on the way…