ACM-ICPC Problem Setting Best Practices
Overview
Comprehensive guidance for creating high-quality algorithm competition problems, covering the full workflow from idea conception to publication.
When to Use
- Creating problems for ACM-ICPC, CCPC, Codeforces, or similar contests
- Writing problem statements with LaTeX
- Generating test data using testlib
- Setting up validators and checkers
- Organizing algorithm competitions
Quick Reference
Core Principles
Problem Quality
- Original idea - No duplicates or trivial enhancements
- Clear statement - Every concept defined, no ambiguity
- Complete constraints - All variable ranges specified
- Strong samples - Catch wrong interpretations
Data Quality
- Edge cases - Min/max values, boundary conditions
- Diverse constructions - Random + handcrafted
- Format compliance - Linux line endings, no trailing whitespace
Platform-Specific
- Polygon - Integrated workflow for teams
- Codeforces - Requires 5-25 rated contests
- Luogu - Requires competition awards
Red Flags - STOP
| Anti-pattern | Fix |
|---|
| Undefined terms in statement | Add definitions in problem description |
| Inconsistent terminology | Use same word for same concept |
| Weak samples | Include edge cases and wrong interpretations |
| Incomplete data ranges | Specify ALL variables' ranges |
| Wrong time limit | Test std × 2 minimum |
| Poor subtask design | Use clear structure, avoid percentages |
Essential Code
testlib Generator
#include "testlib.h"
int main(int argc, char* argv[]) {
registerGen(argc, argv, 1);
int n = opt<int>("n");
vector<int> a(n);
for (int i = 0; i < n; i++)
a[i] = rnd.next(1, 1000000);
println(n);
println(a);
}
Basic Checker
#include "testlib.h"
int main(int argc, char* argv[]) {
registerTestlibCmd(argc, argv);
int jans = ans.readInt();
int pans = ouf.readInt();
if (pans == jans) quitf(_ok, "%d", pans);
else quitf(_wa, "expected %d, found %d", jans, pans);
}
References
