Every position is a pair $(h,g)$, the coordinates of the hunter and goose. Let $P(h,g)$ the probability the goose wins starting from the position $P(h,g)$. Then $$ P(h,g)=\frac16\bigg(P(h,g+1)+P(h,g+2)+P(h,g+3)+P(h,g+4)+P(h+5,g)+P(h+6,g)\bigg) $$ This, combined with the base cases $$ P(h,g)=1\qquad\text{when }g\ge 12 $$ $$ P(h,g)=0\qquad\text{when }h\ge g $$ allows you to calculate $P(h,g)$ for all values of $(h,g)$, starting with the simpler cases (the ones closer to the end of the game), and working backward to the starting position $P(1,7)$. Otherwise, this is just brute force. Not that much brute force, though, since there are less than $20$ positions (the goose can only ever be on spaces $7$ through $11$, the hunter on $1,6,7,11$).