[Question] Benchmarking Collection operations

[darthkurak]

Hi. I try to benchmark some collection operation like Add on List. For example, I want to benchmark List<T>.Add on different size of pre-populated collection to show difference in Capacity growing. Something like this:

public class CollectionTests
{
    private static List<Student> _initialList = AutoFaker.Generate<Student>(131071);
    private List<Student> _listA; //adding to this still fit Capacity
    private List<Student> _listB; //adding to this should grow Capacity
    
    public CollectionTests()
    {
        _listA = new List<Student>(_initialList);
        _listB = new List<Student>(_initialList);
        _listB.Add(AutoFaker.Generate<Student>());
    }

    [Benchmark]
    public void List_A()
    {
        _listA.Add(AutoFaker.Generate<Student>());
    }

    [Benchmark]
    public void List_B()
    {
        _listA.Add(AutoFaker.Generate<Student>());
    }
}

what is the best approach to this? To have pre-defined collections and reuse them between iterations, but do not count collection preparation to execution time? This static approach and using GlobalSetup as well, do not give me correct results.