Kata

Python: Random tests are vulnerable to input modification

1. Confusing description. For warriors who want to tackle this problem, please refer to https://nashpy.readthedocs.io/en/stable/discussion/strategies.html for a real example.
2. Some very annoying rounding issues.

The kata needs to be rewritten with assertApproxEquals instead of rounding 2 decimal places, because the tests keep hitting the rounding boundary with the random tests, which means floating point error is significant here.

Ruby 3.0 should be enabled. KaTeX should be used.

Encountered rounding issues in the JS version. I had to try submitting several times before passing the tests (1 or 2 fails each time). For example :

0.44499999999999995 Expected 0.45
0.09499999999999993 Expected 0.09
0.26499999999999996 Expected 0.26

-0.325 Expected -0.32
-0.375 Expected -0.38

The description is awfully bad: please considere that every warrior didn't study such a field, so make it understandable.

... where each dimension has rows/columns corresponding to a single player's actions, each table entry contains a vector of utilities (a.k.a, payoffs or rewards) for each player. 
The payoff table for Rock-Paper-Scissors is as follows:

 | R   |P   |S
R| 0,0 |-1,1|1,-1 
P|1,-1 | 0,0|-1,1
S|-1,1 |1,-1| 0,0

First, you don't say which one of the players is in row/column. Secondly, neither do you say how we are supposed to read those tuples.

Which can be expressed in python as follows:

utilities = [[0,-1,1],[1,0,-1],[-1,1,0]]

...there I was like... "WTF!?" => strictly speaking, you're telling that a list of lists of integers is the same as a list of lists of tuples of integers. That clearly needs to be explained, how you go from the first to the second (not that hard to figure out, but hell, that's your job to explain this, not the warroir).

Edit: btw -> "can be expressed in python as..." -> make the description as much "language free" as possible.

• It's not explained anywhere what p0 and p1 are!
• about that, the sample tests aren't helpfull at all since they provide only integers. Would need to add one or two examples with floating point values This way, it could be more evident that they are probabilities...

Description: ...

-> that, we can already guess. What would be more useful is that you split the description in "Context" and "Task"

Cool kata with some economy-flavor (I studied that), brilliant explanation and created by the most restless quoran I had the fortune to met? I'm game ;)

If you don't mind, I will translate this nice work of yours into both JS and Ruby soon :)

[I must admit I was almost certain that you would have gone for a list comprehension solution, must be my competitive glands firing all they have got :D]