Wait I think I figured it out...

Another example of why I'm confused:

test case where input = 451054599
correct result is 21
my result 112

451054599 % 13 = 8
21 % 13 = 8
112 % 13 = 8

I just meant it's funny that if I input my incorrect result as input, the new result is the correct answer.

My code is working on everything besides 33 of the random tests. I logged n to see what the input is and it doesn't seem any different than the other test inputs other than it being a different number. Can't figure out what is causing the majority of inputs to work, but not for a few.

Anyone know the difference here between the random tests and basic tests?

I'm passing all basic tests and 168 / 201 random tests.

failed random test example: expected 113 to equal 22.

Locally if I input 113 as n the result is 22.

What am I doing wrong?

Going by the instructions given, the result is incorrect.

I'm sure that is part of the reason this is only rated at 85%.

The instructions don't say either a-b or b-a, they say very clearly a-b. They don't mention the difference having to be a positive number.

in my example:

a-b = (-1272 - 7169) = -8441. Does it not?
-8441 is less than the given margin (1339). Is it not?

According to the instructions given, these params would return 0.

.

.

This isn't a problem about distance, it's a problem about integers and negative integers are very real.

The tests are setup wrong for this problem. For example:
a = -1272
b = 7169
margin = 1339

Difference between a & b = -8441

1339 > -8441 therefore according to the instructions (if the margin is higher or = to the difference, return 0), this should return 0 but the passing result is -1.

Simply genius

This is the only 100% correct solution.

Simple enough until the random test case hits you with the year 3845682734695.
I would call this one harder than 8kyu.