The most direct way to get a good model for fractional sets would likely be to compare hypertrophic response of ancillary movers in common movements to hypertrophic response in isolating exercises for those same muscles. However, at this point my impression is that we don’t have enough data to create this model. In lieu of direct hypertrophy comparisons EMG activation levels might provide valuable insight for calculating fractional sets.

I am still reviewing the literature in this space and it may occur to me that hypertrophy data is prevalent enough that we could use this more direct measurement to create a model for measuring fractional sets. If that ends up being the case I will write another article discussing how we could rely on hypertrophy data to do the same thing I describe in this article.

What Is EMG?

EMG (Electromyography) is a technique used to measure the electrical activity produced by skeletal muscles during contraction. It helps assess the degree of muscle activation during exercises, movement, or muscle function tests. EMG is usually given in percentages and represents the % of electrical activity in a muscle compared to it’s electrical activity during a MVC (Maximum Voluntary Contraction). EMG is frequently given as the average electrical activity during a particular time frame or the peak electrical activity during a particular session.

We Can’t Know Our Fractional Sets If We Don’t Know How to Qualify a “Hard Set”.

To measure fractional sets we first need to agree on what a “hard set” is. A hard set in reference to an exercise itself occurs when you cannot complete even one more repetition while maintaining your desired form.

Does this mean that all exercises which are taken to failure should qualify as a hard set for the primary mover?

If you’ve spent enough time lifting, you know that just because a movement purportedly has the primary mover you are trying to target, does not mean there aren’t movements that give you a better pump, burn, stretch, or even result in more hypertrophy. Because of this folk-fact, I don’t think it is wise to assume two exercises that have the same primary mover will generate the same hypertrophic response nor the same average or peak EMG. This isn’t the only reason to be dubious of classifying any exercise taken to failure as one hard set for the primary mover. For instance we also know that shortened partials stimulate less hypertrophy than lengthened partials even though both should generally have the same primary mover.

So what do we do? Well, I don’t think there’s a perfect answer here. But I think a very good answer that will lead our model in the right direction is this.

Let’s take standard popular movements that from a mechanical perspective should be close to optimal for the prime mover, assume that is a hard set (because, well, we already do) then measure hypertrophic response and EMG data.

From this, we will have a baseline to measure other movements against.

Is this perfect? No. But does anyone think a set of leg extensions when taken to failure with good form and tempo shouldn’t count as a hard set for the quads? Also no. A model that considers leg extensions as a hard set for quads will be wrong. I also think it will be useful.

Before looking at popular movements that are by-consensus considered to be “hard sets for the primary mover”. I think that we can narrow the range of hypertrophic response and or EMG data by looking at rep ranges.

Heavier vs. Lighter Sets

Heavier sets tend to result in higher average EMG activation compared to lighter sets (https://pubmed.ncbi.nlm.nih.gov/26270694/) (https://pubmed.ncbi.nlm.nih.gov/26159316/). For more information on EMG levels between heavy and light sets see this article by SBS: https://www.strongerbyscience.com/emg-amplitude-tell-us-muscle-hypertrophy/.

However, the exercise science community is fairly well convinced (and for good reason) that sets of around 30 (lighter sets) generate very similar hypertrophy on average to sets of 5-10 (heavier sets). So we know that EMG activation can vary independent of hypertrophic response. Does this mean EMG as a proxy for hypertrophy is useless? Not necessarily.

Using Average and Peak EMG Activation Floors as a Proxy For a ‘Hard Set’.

It is necessarily the case that everything we qualify as a hard set based on mechanics and gym-bro-folk-wisdom will have an average-average EMG activation level floor and an average-peak EMG activation floor. These peak and average EMG activation floors COULD be a useful proxy in determining minimum activation levels required to consider a set “hard”.

Unfortunately this does not NECESSARILY mean that sets that meet these EMG activation level floors should qualify as a hard set (but it might!). Just that below these levels they should NOT qualify as a hard set.

I will need to consult with people very familiar with the relevant literature or continue reviewing it myself to determine what these EMG activation floors are and whether exceeding them should qualify a set as “hard”.

This is where my investigation has led so far. It would be nice if I could conclude that exceeding these EMG activation floors would qualify a set as 1 hard set for the primary mover, but at this point I cannot. As I review the evidence further I can concurrently develop this model and improve it’s value even if this one flagged assumption is not entirely correct.

To refine the peak EMG floor and average EMG floor we should use EMG data from movements that we consider to be highly effective based on mechanics.

We now have two important numbers that we hope to be valuable: Peak EMG Floor of Relevant Movements Taken to Failure and Average EMG Floor of Relevant Movements Taken to Failure. Let’s call these PF (Peak Floor) and AF (Average Floor)

Where do we go from here?

Comparing Sub-PF and Sub-AF Hypertrophic Responses To Hypertrophy of Hard Sets

So we now “know” the PF and AF that qualifies a set as hard. But what is the relationship between hypertrophic response as we fall below PF and AF?

We have to stop theorizing and return to the data at this point. I will need to compare hypertrophic response of muscles that received stimulus as an ancillary mover and had sub PF and sub AF to the hypertrophic response in an isolation exercise targeting that same muscle where PF and AF are met or exceeded.

The more data points I have here the better. But with even just one data point showing hypertrophic response during sub AF and sub PF compared to hypertrophic response exceeding AF and PF, I can start to draw a continuously differentiable curve based on this data, and from there infer fractional sets based on peak and average EMG activation.

RationalFit is currently developing the world’s most accurate fractional sets calculator based on this reasoning! To be notified upon it’s release signup for our newsletter!