First I guess you mean 10% and 5% instead of 10 and 5, am I right?
If this is the case, we have , and
By definition, two events and are said to be independent if .
But here we have and , hence they are dependent.
First part is easy, just wins 4 straight ;
The second part deserves a little analysis, there are 4 cases a team can win the cup: (1) the team wins the 4th game in the 4th game in the series (this is the case of 4 straight wins as we found in part one); (2) the team wins the 4th game in the 5th game in the series; (3) the team wins the 4th game in the 6th game in the series;(4) the team wins the 4th game in the 7th game in the series. We need to sum all the four probabilities for these 4 cases, the second, third and fourth cases are follow exactly the negative binomial probability distribution (Check a reference for detail), so the probability for T.O. Leaf to win the series can be written as:
where means the probability for case 2 above etc.
Thus the probability for Montreal to win will be .