DocsPractice
Assess
Graded problems with hidden generated tests, measured complexity, hints that never give the answer, and timed sessions scored out of 100.
How code is graded
In Assess, Run checks the visible tests; Submit runs a hidden suite of at least 50 generated cases (baseline, edge, random and stress, up to 100,000 elements), with expected answers from a reference solution. Your code's complexity is measured by timing it over growing inputs, and its structure is checked for patterns, antipatterns and originality.
Hints
Four steps, each revealing a little more: the invariant, the pattern, the edge cases, then pseudocode. The next step depends on how your attempt fails. Hints never contain the solution.
Timed sessions
Pick Easy, Medium, Hard or Mixed. The clock and hint use are kept by the server, and the report records the session as it happened.
The score
Out of 100: correctness 55 (by test group), performance 20 (measured complexity against the target), quality 10 and time 15, minus hints used (invariant 4, pattern 6, edge cases 4, pseudocode 10). Once 20 sessions exist, your score also shows where it sits among them.
Need help?