Why only 10 items?

You're going to be shown every possible combination and asked to pick a winner. The formula for counting the number of pairwise combinations is:

n(n-1)/2

This means that 10 items produce 45 pair combinations. Any longer of a list would get pretty tiresome to rank!

How is this calculated?

Ranking algorithms tend to be complicated and imperfect — most assume transitive preference (i.e. if I like A more than B and B more than C, then I must like A more than C), but humans aren't always so rational.

As this app is just for fun, I devised a simplistic algorithm where points are awarded according to each choice you make and small decimal values are added at random to minimise ties.