Benchmark for list range removal

ListRangeTests.java

import java.util.ArrayList;
import java.util.List;

public class ListRangeTests {

    public final static int COUNT = 100000;
    public final static int REPEAT = 20;
    
    public static void main(String[] args) {

        // Test all methods
        for (int i = 0; i < 3; i++) {
            
            long time = 0;
        
            for (int j = 0; j < REPEAT; j++) {
                List<Integer> source = constructList(COUNT);
                long start = System.nanoTime();
                
                testMethod(source, i);
                time += System.nanoTime() - start;
            }
            
            System.out
                    .println("Method # " + i + ": " + (getMilli(time) / (double)REPEAT) + " ms");
        }
    }

    private static void testMethod(List<Integer> source, int alternative) {

        switch (alternative) {
        case 0:
            // Slow method!
            int length = source.size() / 2;
            int index = source.size() / 2 - 1;
            
            for (int i = 0; i < length; i++)
                source.remove(index);

            break;

        case 1:
            int toIndex = source.size() / 2 - 1;
            
            // Alternative one
            for (int i = source.size() - 1; i > toIndex; i--)
                source.remove(i);
            
            break;

        case 2:
            // Alternative two
            source.subList(source.size() / 2, source.size()).clear();

            break;
           
        default:
            throw new IllegalArgumentException("Not an algorithm: " + alternative);
        }
        
        // Sanity test
        if (source.size() != COUNT / 2)
            throw new IllegalStateException("Algorithm failed to divide the list.");
    }

    private static double getMilli(long time) {
        return time / 1000000.0;
    }

    private static List<Integer> constructList(int count) {

        List<Integer> source = new ArrayList<Integer>();

        // Add 100 000 consecutive integers
        for (int i = 0; i < count; i++) {
            source.add(i);
        }

        return source;
    }
}