fix(curriculum): last test for step 1 RPS project to better check for final result

[jdwilkin4]

The last test for the Rock, Paper, Scissors game review project allows for incorrect answers like this to pass

function getRandomComputerResult() {
  const options = ["Rock", "Paper", "Scissors"];
  return "Paper"
}

this update should better test for the correct final result so hardcoded solutions like this don't pass

Checklist:

• I have read and followed the contribution guidelines.
• I have read and followed the how to open a pull request guide.
• My pull request targets the main branch of freeCodeCamp.
• I have tested these changes either locally on my machine, or GitPod.

Closes #54814

[JoyShaheb]

i approve the solution 👍

[lasjorg]

It seems like it is also possible to pass this with an incorrect index that gives back undefined.

return options[Math.floor(Math.random() * 3) + 1]

Obviously, getting it to pass on the first try is...well, random.

[lasjorg]

I'm not sure if we should specifically have a regex test for that. If anything, I think it should test for valid random generation and not invalid.

It seems fairly improbable that the Set wouldn't contain all three strings after calling the function 50 times. So how about we just check the Set contains what we expect it to?

I guess we can also remove falsy values and check the length, but I'm not sure if there is any benefit to that over just checking for the expected strings.

[jdwilkin4]

@lasjorg

So how about we just check the Set contains what we expect it to?

I am not sure I am understanding what you mean here.
Because we already have this test, to check to make sure their answer returns one of the valid random choices

Your `getRandomComputerResult` function should return one of the options in the `options` array.
```js
assert.include(["Rock", "Paper", "Scissors"], getRandomComputerResult());
```

So wouldn't it be redundant to do another assert.includes here for the last test?

[lasjorg]

That test checks that calling the function one time produces at least one of the 3 strings in the array.

The other test calls it 50 times to "guaranty" the function produces each of the three strings at least one time.

[lasjorg]

function getRandomComputerResult() {
  const options = ["Rock", "Paper", "Scissors"];
  return options[Math.floor(Math.random() * 3)]
}

const results = new Set();

for (let i = 0; i < 50; i++) {
  results.add(getRandomComputerResult());
}

console.log([...results].filter(Boolean).length === 3); // true

---

function getRandomComputerResult() {
  const options = ["Rock", "Paper", "Scissors"];
  return options[Math.floor(Math.random() * 3 + 1)]
}

const results = new Set();

for (let i = 0; i < 50; i++) {
  results.add(getRandomComputerResult());
}

console.log([...results].filter(Boolean).length === 3); // false

[jdwilkin4]

I had reached out to Naomi and went with her suggestion.
Tested it locally and it works on my end.
It looks like both your's and Naomi's suggestion are pretty similar

[lasjorg]

That works, I wasn't really suggesting anything, just showing what I meant in code.

But do we really need to loop that many times?

I'm no statistician, but even with 50 loops the probability of Math.floor(Math.random() * 3) not generating all three number in the range is very low.

Edit: I know it is quick on a desktop, but for low power devices it might be a little overkill. Just a thought.

[jdwilkin4]

Yeah, that makes sense.
I updated it to be run 50 times

[lasjorg]

This looks good to me.

We could use 100 or even 200 for the loop, but I think 50 is just fine.