Python rolling mean starting on the next row

Question

I have results from a horse race and want to have the rolling win count for the start of each race. This is what I currently have:

Horse	Position
A	4
A	1
A	3

This is what I'm after

Horse	Position	Wins
A	4	0
A	1	0
A	3	1

So when horse A started the second race it had 0 wins but when it started the third race is had 1 win

Answer 1

Use the shift operator and cumsum

df["Wins"] = (df["Position"].shift(1) == 1).cumsum()

output:

  Horse  Position  Wins
0     A         4     0
1     A         1     0
2     A         3     1