Discussing Diminishing Returns: A New Scoring System for Powerlifting
Jean C. Peyen, Ruheyan Nuermaimaiti, Joshua Cunningham · 2025
Read Discussing Diminishing Returns: A New Scoring System for Powerlifting by Jean C. Peyen, Ruheyan Nuermaimaiti, Joshua Cunningham (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-13040.txt.
Introduction
Jean Peyen and co-authors propose a new scoring system for powerlifting in a 2025 paper. Scores must scale lifts fairly across body weights. The current federation system rests on a growth curve tuned to elite lifters. It assumes gains shrink as body weight rises, with no early phase of rising returns.
Data from a vast open lifting database show that early rising phase is real below a weight threshold. Their fix swaps in a logistic curve that bends both ways. The aim shifts too: fair standards for the general public and casual athletes, not just elites near world records. The paper reopens how strength sports judge pound-for-pound worth.
Catalog record
| CREATOR | Jean C. Peyen, Ruheyan Nuermaimaiti, Joshua Cunningham |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-13040
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Discussing Diminishing Returns: A New Scoring System for Powerlifting Jean C. Peyen, Ruheyan Nuermaimaiti, Joshua Cunningham Abstract Maximal strength increases with body weight, this is why scoring methods have been developed in order to fairly scale powerlifting performances based on athletes’ body weight. The International Powerlifting Federation (IPF) Good Lift (GL) system, introduced in 2020, is a scaling method based on a Von Bertalanffy function. It is specifically tailored to elite powerlifters—defined as those achieving at least 84 % \% of the current world record in their weight class, according to the IPF’s “Golden Standard”. A key assumption of the GL system is a principle of diminishing returns, stating that performance gains decrease as body weight increases. However, data from the Open Powerlifting database reveals the presence of a phase of increasing returns below a certain body weight threshold. To capture this initial phase, we propose a new scoring system based on a logistic function. This approach is designed to establish strength standards for the general population and for athletes not specialised in powerlifting. Keywords: powerlifting, growth models, diminishing returns, Von Bertalanffy function, logistic function, resampling, kernel density estimate Powerlifting is a strength sport consisting of three movements: the squat, the bench press, and the deadlift. In a powerlifting competition, an athlete must perform three attempts in each movement. Only the best attempt in each movement is kept, and the total of these three attempts determines the winner of the competition. This paper is using the Open Powerlifting dataset [ 2 ] . This is a continually updated dataset that compiles competition results from multiple powerlifting federations since as early as 1964. We only consider unequipped (raw) athletes in the open age division. The present analysis focuses on the relationship between body weight and total lifted (thus, we only consider athletes who performed at least one valid attempt per movement). In competitions, the heavier athletes are expected to lift heavier weights, so federations have established weight classes to ensure fair competitions. For example, Table 1 shows the weight classes on World Games and International Powerlifting Federation (IPF) open division. In order to take advantage of this system, athletes can adopt drastic measures, such as a restrictive diet, or dehydration to enter a weight class which is below their natural body weight while maintaining their muscle mass [ 9 ] . The effect of this practice can be observed in Figure 1 where we can see an accumulation of the top performances near the upper value of the weight classes. Interestingly, this phenomenon is more prominent in males, indicating that gender-related factors significantly influence the ability to lose weight for an upcoming competition. This is consistent with the fact that testosterone is known to facilitate fat reduction [ 4 ] , and hormonal fluctuations during the menstrual cycle contribute to variations in fluid retention [ 8 ] . Category World Games IPF Open World Games IPF Open Females (kg) Females (kg) Males (kg) Males (kg) Lightweight Up to 57.0 47.0, 52.0, 57.0 Up to 74.0 59.0, 66.0, 74.0 Middleweight 57.01 - 63.0 63.0 74.01 - 93.0 83.0, 93.0 Heavyweight 63.01 - 72.0 69.0, 76.0 93.01 - 120.0 105.0, 120.0 Super Heavyweight Over 72.0 Over 84.0 Over 120.0 Over 120.0 Table 1: World Games and IPF Open Weight Classes for Females Figure 1: Total lifted by Females (left panel) and Males (right panel) in IPF competitions and IPF weight classes (red dotted lines) To compare the athletes from different weight classes, performances can be scaled according the athletes’ body weight. An ideal scoring formula should take the following form \text{Score }\propto\frac{y}{f(x)}, (1) where x and y are respectively the body weight and the total of the scored athletes, f(x)={\mathsf{E}}(y|x) is the expected total for a random athlete of the same body weight. Implementing a scoring system can have multiple benefits aside from determining the strongest athlete “pound for pound”. Federations may consider using a scoring system for international qualification events to allow athletes to transition smoothly between weight classes instead of undergoing unhealthy diets. This could also help comparing overall performances of different teams. In practice, f is a nonlinear function, and various models have been proposed to estimate it. These models can be based on a sufficiently flexible family, such as an ad hoc polynomial family, which fits the data well but has limited explanatory power and discards any notion of parsimony. This is what is done with the Wilks score, which rescales the total by a 5 -th order polynomial [ 10 ] \text{Wilks Score}=\frac{C\cdot y}{a+bx+cx^{2}+dx^{3}+ex^{4}+fx^{5}}, (2) where x is the athlete’s body weight (in kg), y is the athlete’s total (in kg) and f_{\text{Wilks}}(x)=a+bx+cx^{2}+dx^{3}+ex^{4}+fx^{5} . The coefficients a,b,c,d,e and f are specific to the sex of the lifter. The constant C is set to 500 in the original version of the Wilks formula and to 600 in the 2020 version known as Wilks-2. The coefficients can be found, for instance, in the IPF models evaluation report [ 1 ] . Alternatively, f can be chosen from a family of functions that conforms to some postulates, providing a more interpretable and theoretically grounded model. In 2020, the International Powerlifting Federation (IPF) has implemented the Good Lift (GL) score to replace the Wilks score [ 1 ] . The IPF GL score is defined by \text{IPF Score}=\frac{100\cdot y}{A-B\cdot\exp{}(-Cx)}, (3) where the coefficients A,B and C are all positive and f_{\text{IPF GL}}(x)=A-B\cdot\exp{}(-Cx) . The IPF GL score is justified by the following set of assumptions: 1. With increasing body weight, the absolute lifted weight increases, 2. the relative performance (defined as the total lifted divided by the body weight) decreases, 3. there is a potential physical limit for the result both for each value of body weight and the absolute weight limit for the entire population. The second assumption, which can be interpreted as a principle of diminishing returns —see figure 2 —may be challenged as below a certain body weight, an individual may lack the necessary mass to maintain optimal physiological functions (e.g. of vital organs and passive structures). This suggests that we should consider a model that incorporates an inflection point below which the relative strength of an athlete would be an increasing function of the body weight. Figure 2: Average total lifted by males in each body weight myriad for the Open Powerlifting dataset 1 Methods This section is structured as follows: First, we review the definition of the exponential family used to design the IPF GL score. Next, we review the definition of the logistic function. Finally, we propose to resample the dataset in order to obtain fits that are suitable for athletes with extreme body weight. Resampling is performed using a Kernel Density Estimate (KDE) of the body weight distribution. 1.1 Von Bertalanffy function (IPF GL score) We replicate the type of formula used for the IPF GL score. It is a variant of the Von Bertalanffy function, which can be presented in the following way: f(x)=L\left[1-\exp{}(-k(x-x_{0}))\right] (4) It exhibits an initial phase of linear growth, followed by a slowing of the growth rate. It is commonly employed to model the growth of animals as they age [ 3 ] . This behaviour is similar economic growth