HiMCM preparation resources
PRACTICE THE
DECISION CHAIN.
Use resources to strengthen a modeling process, not to collect formulas. The aim is to connect a real question to evidence, a mathematical structure, tests, and an honest interpretation.
Problem definition
Operational definitions, objectives, constraints, stakeholders, units, and measurable outputs.
Evidence handling
Source quality, cleaning, missing values, uncertainty, provenance, and appropriate citation.
Method selection
Regression, optimization, simulation, networks, differential equations, scoring models, and hybrids chosen for a reason.
Evaluation
Back-testing, cross-validation, scenario analysis, dimensional checks, stress tests, and sensitivity.
Visual explanation
Figures and tables with a purpose, readable labels, appropriate precision, and interpretation in the text.
Technical narrative
A traceable paper in which notation, claims, results, references, and limitations agree.
How should a team prepare for HiMCM?
Start preparing about eight weeks before the November contest window. In weeks 1-2, read the official rules, the AI-use policy, and two or three past problems from mathmodels.org to set a baseline. In weeks 3-4, practice individual modules: data cleaning, model selection, sensitivity analysis, figure design, and technical writing. In weeks 5-6, integrate those skills on one full past problem as a team, with each member owning a role such as modeling, computation, research, or writing. In weeks 7-8, run at least one timed simulation of the full workflow, then debrief what broke. Throughout, study awarded papers to learn how strong teams state assumptions, defend model choices, and structure a 25-page report.
COMAP’s official preparation resources
START WITH THE
ORGANIZER’S LIST.
COMAP explicitly encourages advisors and faculty to coach and prepare students before the contest, and names its own resources for doing so. Build any preparation plan on these first:
- The official rules and prior problems on the HiMCM/MidMCM contest website – the controlling source for every requirement.
- Mathmodels.org – COMAP’s archive of previous years’ problems and solutions.
- The Consortium HiMCM and MidMCM articles – the Director’s Article and Judges’ Commentary describe what judges look for in each section of a solution, with exemplars for each section.
- The official “Procedures and Tips” article – COMAP’s guidance for students and advisors before, during, and after the contest.
- The AI-use policy – COMAP’s policy on large language models and generative AI, which applies to research, model development (including code), and the written report, and requires a Report on Use of AI appended to the solution.
COMAP also specifies the working toolkit: students should have access to computers for word processing, spreadsheets, and dynamic modeling of graphs and geometric relationships; graphing calculators may also help. Practice with the same tools the team will use during the 14-day window.
EIGHT-WEEK
STARTER PLAN.
Baseline
Complete one short problem, identify tool gaps, and establish shared folders, citations, and version naming.
Modules
Practice data cleaning, model comparison, sensitivity, figure design, and concise technical English separately.
Integration
Solve a past problem over several sessions with explicit milestones and an internal reviewer.
Simulation
Run a timed rehearsal, inspect the final PDF, then convert the debrief into a contest checklist.
USE A RESOURCE
ETHICALLY.
- Read past work to study structure, not to reuse its wording, model, code, figures, or references.
- Verify data licenses and cite the source actually used.
- Keep a source log while researching instead of reconstructing it at the end.
- Disclose software, external help, and AI tools as required by the current official rules.
- Retain reproducible code and notes so results can be checked.