When the kyu number gets lower, the problems get harder. Compare your kyu rating with those of the people on the leader board.

Okay, I fixed that.

Agreed and fixed.

I add double sharps and double flats in the next kata. I've never in my life seen a triple sharp or a triple flat, so I wouldn't worry abou them.

I think you're talking about a key signature, which is different from a key. A key has to do with which notes are being used in the music and how they relate to each other. To say that a piece of music is in some key or another is actually a basic type of harmonic analysis. A key signature, on the other hand, is simply a notational convenience that allows the composer to say, "Any time you see a C, play a C♭ instead." The main reason for it is to keep the music on the page from getting too cluttered. It isn't necessary to use a key signature, and there is some music that doesn't use one at all.

Yeah, it's mostly about the harmonic or melodic function of the note. For an example, a C♭ major scale:

C♭ D♭ E♭ F♭ G♭ A♭ B♭ C♭

This can be rewritten, without changing what it sounds like, as a B major scale:

B C♯ D♯ E F♯ G♯ A♯ B

No, E♭ is the same as D♯. But F♭ is the same as E.

Added. Thanks!

@animatedgif Thanks! I was misreading the error message. By "return," I thought it meant to return the cards to the deck, not from the draw method. It turns out that I had another unrelated bug in my initialize method. Thanks for the help!

There are a bunch of tests (at least in the Ruby version) labelled, "returns the cards that were drawn." All of these are fail for me, because that feature is never mentioned in the description. I've tried setting up a method_missing to find out what method is being called that's supposed to allow somebody to return the cards, but there is no such method being called. When is this card-returning supposed to happen, and how have 44 people managed to pass this kata without that information?

Real programmers derive the equations themselves. ;)

It's strange to me that the expected range is of the player's possible rolls when the player has a higher initial score, but that it's of the opponent's possible rolls otherwise. If the opponent's initial score is, say, three points higher than the player's score, wouldn't it make more sense to have the expected result be (4..6) to reflect the player's rolls that could result in not losing, so that a returned range always refers to the same player?

Why is this still in beta?

The tests require cards to be somehow returned to the deck. This feature is unmentioned in the description, and there's no way to determine under what conditions the cards are supposed to be returned. Is there some method that's supposed to handle that?