Why Relying on Standard Models Stalls Math Modeling Competitions

In math modeling contests, a common problem is the over‑reliance on "plug‑in" existing models, which hampers innovation and makes it difficult for entries to stand out.

Problem Root: Understanding vs. Applying

Contestants often have limited time to solve complex real‑world problems, leading some to directly apply familiar classic models to save time. However, not every problem fits these models, and each real‑world issue has unique background and constraints.

Model Selection: Convenience vs. Suitability

Many participants choose models they have seen in textbooks or papers, assuming they are reliable. This blind selection can be inappropriate; for example, using linear regression for a complex socio‑economic issue may ignore non‑linear relationships, and blindly applying advanced machine‑learning models such as SVM or neural networks can cause over‑ or under‑fitting.

Lack of Innovation: Safety vs. Creativity

The most fatal issue is the lack of innovative spirit. To play it safe, contestants stick to known models and avoid trying new methods or hypotheses, preventing their work from showcasing unique insights.

Case Study: Illegal Wildlife Trade

This 2024 MCM paper (ID 2413565) addresses illegal wildlife trade. The authors first performed thorough data analysis and background investigation to deeply understand the problem, recognizing that simple single‑model approaches cannot capture its multi‑dimensional impact.

The study combined classic methods such as AHP and EWM with innovative approaches like a win‑win model and a supplemental vector model, ensuring comprehensive understanding. Extensive real data and statistical analysis, including multivariate linear regression, were used to validate model effectiveness.

The key lesson is how to innovate on top of classic models—by introducing supplemental vector and win‑win models (the authors’ own concepts) to overcome the limitations of single models.

How to Avoid Model Pitfalls

First, spend time deeply understanding the problem’s background, constraints, and requirements instead of rushing to find a ready‑made model.

Second, select models flexibly based on the specific characteristics of the problem, combining multiple methods to leverage their strengths.

Finally, be bold in innovation—try new methods and ideas even if they fail, as each attempt provides valuable experience.

Mathematical modeling contests are not just about model competition; they are about innovation and thinking. By moving beyond blind "plug‑in" and embracing deep understanding, flexible selection, and creative experimentation, participants can stand out and contribute smarter solutions to real problems.

References:

https://www.contest.comap.com/undergraduate/contests/

http://www.mathmodels.org/mathmodels/2024/2413565.pdf