This is the most clever answer I've ever seen...

Yes, my first attempt relied on Math.Round(x, MidpointRounding.AwayFromZero), that is absolutely correct mathematically. I am always upset when I have to adjust the right solution to the wrong tests.

This solution returns 8 if n=0, which is wrong. n=0 isn't forbidden in the description, maybe the testcases should be updated.

whoa, that's brutal...
Apparently the Kata editor only validates the full tests - the example tests aren't validated/executed at all. That's why I missed that problem.
Moved the function to preloaded - should be working now with full tests and example tests.
Thank you guys, I learned something today ;)

In c#, example tests still no doTest ;-)
Put it to preload block?

The example tests in C# are full of problems. There is no method "doTest" and in the example tests the method problem is expeceted to have a capital P.
The real tests work without any problem.

Approved

You're right - anybody can rename these. I just meant it'd be nicer if the prepared names are 'meaningful'. It's good practice - and authors should lead by example, especially in a fundamental (beginner level) kata ;)

There is no 'correct' return value - there are many. To be more precise:

(9 * NoOfDigitsInResult) - 1

many. (-1 because the leftmost digit mustn't be 0)

Translated to C#.

Agreed, but you are free to use whatever variable names you want - just edit the initial solution :D