A University of Houston researcher and his collaborators have developed a mathematical model that helps identify whether a competitive environment is healthy, stagnant or skewed.
Published March 16 in the journal npj Complexity , the study led by UH Computer Science Professor Ioannis Pavlidis presents a general, falsifiable framework for assessing competition quality and fairness. The model works by analyzing the statistical pattern of repeated success and reverse-engineering the kind of competitive system that produced it.
"My hypothesis was that there is a universal pattern across human endeavors," said Pavlidis, senior author of the study and Eckhard-Pfeiffer Distinguished Professor. "We tested that idea by analyzing competitive activities across a broad range of human achievement."
Supported in part by the National Science Foundation, the study examined data from three distinct domains: Olympic athletes, scientists competing for federal research grants and World War II fighter pilots.
Optimizing Fair Competition
The team found that high-performing competitive systems strike a delicate balance: they must be demanding enough to push people to improve and allow exceptional performers to emerge, but not so extreme that success becomes effectively out of reach. This "sweet spot" fosters both excellence and opportunity.
"In demanding but fair competitive systems, competition itself becomes a learning mechanism," Pavlidis said. "Everyone is pushed to improve, but some improve more than others, and over time those accumulated gains can produce striking differences in success."
The research uses mathematical modeling to classify competitive environments into three distinct shapes:
- Tough but fair (optimal): A small group of high performers emerges, but positions are not locked in. New competitors still have a real chance to rise, helping maximize learning, innovation and performance.
- Winner-take-all (overly concentrated): One or two individuals dominate persistently. This points to structural imbalance, discourages participation and can ultimately lead to stagnation.
- Broad award (too evenly distributed): Success is spread so widely that differences between participants remain small. While this may appear "fair," it often reflects weak competitive pressure and lower overall achievement.
Building Better Systems
The applications of this model extend far and wide. Organizations can use this methodology to assess competitive environments such as research funding systems, career progression systems and military training programs, with the goal of fostering excellence.
In future studies, Pavlidis plans to extend the methodology from individual-based competitions to team-based competitions across domains and over time, testing whether they too follow universal, falsifiable laws.
"The sky's the limit," he said. "This research offers a new way of thinking about competition, which is a fundamental part of civilization."
