I had some trouble with run time errors initially, I was making it way more complex than it had to be. I find it helpful to note that the number of subsets starting with a letter is: the length of the string - the index of the letter. I.e. if the string is 'banana', the number of subsets starting with 'b' is len('banana')-index('b') = 6-0 = 6 : [b, ba, ban, bana, banan, banana]

Using this method you can add the number of subsets starting with each vowel, and the same with each non-vowel.

for i in range(len(string)):
    if string[i] in v:
        c1 = c1 + len(string) - i
    else:
        c2 = c2 + len(string) - i

You could also make two seperate lists - one with the indicies of each vowel, and one with the indicies of each non-vowel. Then get the scores by calculating the length of the string * length of the list of indicies - sum of the list of indicies.

for i in range(len(string)):
    if string[i] in v:
        a.append(i)
    else:
        b.append(i)
c1 = len(string)*len(a) - sum(a)  
c2 = len(string)*len(b) - sum(b)

Then you can determine the winner by comparing the scores.