Two-set prediction of 1 Rep Max

In strength conditioning (weight-based) the one rep max (1RM) refers to the maximum weight you can do ONE and ONLY ONE repetition with.

In a previous article, I listed the many reasons why this number should be kept as an indicator for strength assessment. One shall avoid however pulling that weight because it is heavy and straining, while the same result can be achieved with less weight and few more repetitions.

A plethora of one rep max prediction algorithms can be found on the Internet, but people, including myself a while ago, have been using them with more of less satisfaction.

There is two facts to consider when using those algorithms. First, there is tons of them, using tons of different formulas yielding of course different results for the same set of data in input. A fairly exhaustive state of the art of those methods demon how pointless they are on a quantitative basis. Second, even when you would be lucky and use a “good and serious” calculator based on published statistical research, these are made out of AVERAGES on a large heterogeneous set of volunteers or an homogeneous but not representative set of volunteers. Thirdly, because the muscle groups do not function in the same ways, the studies yield different equations for different exercises. Typically, the formula for prediction of 1RM at the Leg Squat would be very different from that to predict the benchpress 1RM.

In an excellent article on Strength Assessment, Matt Brizcky, health coordinator at Princeton University, states that :

A number of prediction equations have been developed and used to estimate a 1-RM based on the relationship between strength and anaerobic endurance. While some of the equations have proven to be reasonably accurate, one problem with them is that they do not take into consideration individual differences.2,3 (…)

Because of these genetic influences, especially muscle fibers, some people perform either less than or more than 10 reps-to-fatigue with 75 percent of their maximal strength. Westcott reported data on 141 subjects who did a test of anaerobic endurance with 75 percent of their 1-RM.6 (Remember, it has been suggested that an individual could do 10 reps-to-fatigue with this workload.) According to the data, the subjects completed an average of 10.5 repetitions. However, only 16 of the 141 subjects (11.35 percent) did exactly 10 reps-to-fatigue with 75 percent of their 1-RM. Many of the subjects were within a few repetitions of 10. In fact, 66 of the subjects (46.81 percent) were able to do between eight and 13 reps-to-fatigue. On the other hand, 75 of the subjects (53.19 percent) did either less than eight reps-to-fatigue or more than 13. At the extremes, two subjects did only five reps-to-fatigue and one managed 24. (…)

Another approach to attain an individual-specific estimate of a 1-RM is to use a prediction equation. The most frequently used prediction equations are based on the reps-to-fatigue done in one set.2,3 . However, a test using one set does not account for individual differences in anaerobic endurance. A better way to assess muscular strength from anaerobic endurance is to use a prediction equation that is based on the reps-to-fatigue obtained in two sets.


2. Ware, J ., et al. Muscular endurance repetitions to predict bench press and squat strength in college football players. Journal of Strength and Conditioning Research 9: 99-103, 1995.

3. Mayhew, J., J. Prinster, J. Ware, et al. Muscular endurance repetitions to predict bench press strength in men of different training levels. Journal of Sports Medicine and Physical Fitness 35: 108-113, 1995.

Westcott’s research clearly emphasizes the risk taken when using calculators that work for an average. If you looked at the distribution of the statistical sample, you certainly realized how inaccurate such methods may be for you if you do not have the chance to fit the average. Average figures are indeed one of the very typical misuse of statistical studies in everyday life.

Unlike all one-set generalized methods (including those presented by Brizcky) , Brizcky’s two-set method ensures to yield a personalized accurate 1RM as well as being universal, meaning that one can use it fairly accurately regardless of the muscle group and exercise being considered.

I have computed below the two-set algorithm described by Brizcky and I show as well what his older (one-set) algorithm would have given you, just so that you get a feel of the difference.



Be Sociable, Share!

4 thoughts on “Two-set prediction of 1 Rep Max

Leave a Reply

Your email address will not be published. Required fields are marked *