Easier interactive problem, implement a partial tournament sort that identifies the top k participants out of a pool of n.

Interactions allow users to have two participants compete with each other.

Problem difficulty varies based on maximum number of allowed queries. The maximum number of queries will be limited to prevent solutions from prior groups.
1. O(n log n) - Fully sort the list and output the top k
2. O(n log k) - Use a sliding heap to keep track of the running top k
3. O(n) - Use quickselect to identify the kth partition point.

I would lean towards the second difficulty where a sliding heap can be used but the list cannot be fully sorted.

C++ nth_element or partial_sort or Pypy heapq.nlargest would provide fast implementations with operator overloading for interactions.