You can also check out my fork with added comments : https://www.codewars.com/kumite/61443ee026b2c50033a0499a?sel=61b1dce4b2b845002806a738

spoiler

More answer than I could ever give. :P

( I just run the random tests a couple of times and take an approximate average. Accurate to about an order of magnitude .. )

not directly. You can use the fork panels, but you have to build some code on your side to actually be able to compare the different solutions on the same inputs. Usual (proper) way:

1. build a set of random inputs with their expected result up front
2. paste in the test the code of all the solutions you want to test (or in the solution area, but then you'll have to import each function in the tests)
3. build a loop on the functions you wanna compare. Inside that block, build one single it block where you test all input on that function only
4. run the tests and look at the timing of each it block in the output panel

that's rather cumbersome, and there are a lot of things to considere to be sure what you do is correct to estimate/compare the performances, but here is the general idea.

.

You'll need to post your code before any conclusions can be made.

Thanks!