Mayhew made separate formulas for bench, squat and deadlift using advanced math and data from college athletes.
https://wikimedia.org/api/rest_v1/media/math/render/svg/8506fbc4b91488a00a05c90b9736b0ff96299fad
The Mayhew et al. 1RM prediction formulas are a set of equations derived from a research study published in 1992 in the Journal of Strength and Conditioning Research.
Unlike some of the other generalized prediction models, Mayhew developed separate equations for different lifts:
Bench Press 1RM = (100 x Weight) / (48.8 + 53.8e-0.075xReps)
Squat 1RM = (100 x Weight) / (52.2 + 41.9e-0.055xReps)
Deadlift 1RM = (100 x Weight) / (49.3 + 49.9e-0.073xReps)
Where: Weight = Load lifted (kg) Reps = Number of reps completed
To derive these lift-specific formulas, Mayhew analyzed data from 37 male and 24 female college athletes performing maximal bench, squat, and deadlift tests.
Using non-linear regression analysis on this dataset, they found cycling exponential models like the ones above provided the best fit for predicting 1RM from sub-maximal rep-load data for each lift.
The different coefficient values (e.g. 48.8, 53.8 etc.) in each formula represent the best-fit parameters from the regression modeling for that particular lift.
A key advantage of Mayhew's approach was deriving separate, lift-specific prediction models rather than a single generalized equation.
This accounted for the different strength curves and fatigue characteristics between lifts like the bench, squat and deadlift.
While slightly more complex formulas, Mayhew's models demonstrated increased 1RM prediction accuracy compared to other generalized models when validated on their college athlete sample.
This highlights using advance regression techniques and lift-specific modeling as an approach to improving 1RM prediction accuracy.