hey! thought i'd give Rui a bit of a break and have a go aha
So this isn't too bad if you break it down a bit. So let's just put one of that pair in the left-most slot in the first rung (call them Player A). Let's look at the possibilities they verse Player B, depending on where Player B is.
If Player B is next to them in the same bracket (a 1 in 3 chance), the condition is already satisfied. So, 1/3 there. If Player B is in one of the other slots (a 2 in 3 chance), two things need to happen. Player A needs to win their game, and Player B does too. So multiply 2/3 by 1/2, and then 1/2 again, and you get 1/6.
Add these together!
1/3+1/6=1/2
(sorry I don't know how to use the fancy math text like Rui but hopefully this makes sense aha)
OI TRUUUUUUUU.
Now, hints for the second part:
(whereas for two we only had \(\binom42\binom22 \frac{1}{2!} = 3 \) )
Try splitting the cases this way before going into subcases:
Case 1 - The pairings are split something like (A,B) (C,D) || (E,F) (G,H), so A and C can meet by the second round. This will be the case that best reflects the computations in part (i) (although they won't be an
exact replica)
Case 2 - The pairings are split something like (A,B) (E,F) || (C,D) (G,H), so A and C are forced to meet in the third round.
That allows you to generalise it upwards