Velocity-Based Training: 1RM Prediction

In this snippet from our new Ultimate VBT Guide, explore the intricacies of predicting 1RM using Velocity-Based Training.
Instagram LinkOutput Sports Twitter LinkOutput Sports Linked In LinkOutput Sports Facebook LinkOutput Sports Youtube Link

1RM prediction is probably the most researched area of VBT, and often generates the most interest, questions, and challenges for S&C coaches. In reality, is 1RM prediction really necessary? One could argue that the previous method, load manipulation, could be enough to fully autoregulate intensity utilising velocity. Nevertheless, this section will explore how to predict 1RM most accurately, and the best ways to implement this approach.

1RM prediction works in a very similar way as previously outlined when manipulating load. The only real difference comes in the data being predicted. Setting velocity targets from full profiles uses a method called interpolation, where values are predicted inside the measured dataset, whereas 1RM prediction extrapolates to the predicted value, which sits outside of the measured dataset. As a result, the interpolation method is more likely to render more accurate predictions placing challenges on accurate 1RM predictions. But it is possible…

The aim of 1RM prediction is to accurately predict one’s daily strength levels to ensure loading is reflective as it can be, and stimuli matches the desired adaptation as closely as possible. To do so, an athlete is required to perform a submaximal profile at the start of each session, and then that data is extrapolated to a reference velocity for 1RM (e.g., the velocity previously recorded for 1RM in a pre-intervention testing session) (Figure 1).

Figure 1: A graph demonstrating how to predict 1RM from submaximal loads using a reference point of velocity at 1RM previously recorded and linear regression.

Understandably, it is often important to try and limit the number of loads (which are typically warm-up loads) to ensure a time efficient practice, however, this can impact the validity of the predictions. For example, there is an approach called the “2-point method”, whereby an athlete would perform two loads, use a reference velocity (e.g., velocity at 1RM), and predict their daily 1RM off just two loads. As you can see from Figure 1, it is possible for 1RM prediction to be quite inaccurate. The reason for this could be due to the statistical model being applied. The straight line in this example renders a 1RM prediction approximately 20 kg higher than their actual 1RM. This can be an issue if the profile is not truly linear. Whilst most profiles are, and thus software such as the Output Hub can be very accurate, there is always the chance that a more sophisticated model could account for these more complex profiles.

Figure 2: A graph to show 1RM prediction using the second-order polynomial (curved model)

As you can see, the line now hits the reference velocity for 1RM, meaning that the prediction will be much more accurate, within about 2 kg to be precise. Therefore, showing the importance of choosing the correct mathematical modelling when trying to implement arguably the most complex applications of VBT. 

If you enjoyed this snippet, keep reading more from our new ‘Ultimate User Guide to Velocity-Based Training' with insights from esteemed thought-leaders in the VBT space, including Dr. Steve Thompson, Chris Tombs, Dan Baker, and Nic Gill: Download now!

SPeak to a performance specialist

Book a demo to see how Output how Output can support you and your goals.