Question: You are working on a difficult passage from a new piece you are learning on the piano.You wish to play it correctly 4 times before calling it a day. If you have a probability of 2/3 of playing it correctly on every attempt, and the attempts are independent (unfortunately!), what is the probability that you attempt to play it at least 8 times?
So I'm taking the compliment of in order to figure out what is the probability that you attempt to play it at least times by doing 1-prob of getting it right 4 times in under 8 attempts.
I went though this case by case. The first case is in 4 attempts I play the song correctly 4 times.
The second case is getting it right in 5 attempts. I know automatically the fifth attempt I have to play the song correctly or I would have stopped early (would fall into the previous case) thus that leaves 4 combinations to play.
The third case is getting it right in 6 attempts.Again the 6th attempt I must get the song correct or I would have stopped earlier which leaves 10 combinations... and so on.
What is wrong with this approach?
In a direct answer, I think you're on the right track. I'll share my method, which sounds similar to yours, but it treats the "history" as an all-at-once rather than try to combine runs of 4, 5, 6, etc. separately.
The question wants to know how likely is it that you have to go at least as many as 8 tries to succeed. In problems like this, it's useful to think of what has to happen to require an eighth try.
In your previous 7 tries, you must have succeeded precisely 3 times, no more, no less, and in no particular order. Think about how many combinations that is, and what fraction that is to the total number of outcomes in 7 tries.
That gives you a fraction of the likelihood of getting to the 8th try, and you have a 1/2 odds of getting it right on the 8th try.
Can you work it from there?