Could you aim it a load bearing wall and undermine a whole building?

Never too late for more features 😈

If it ain't broke, keep fixing it until it is

Foucault was right

Lastly, action games are highly-concurrent, formally an intractable series of interlocking differential equations without closed-form solutions. So we use numeric approximations, which in variable timesteps introduce intrinsic RNG . So high-level players cannot reliably "reach the ceiling."

Adding on to this:

It's so much easier to add a feature to systems that are already working reliably. Aim to always have your code in a working state (even if it's ugly/placeholder), then you can see how it acts in context.

After the credits roll, the camera pulls back to reveal that you, the player, have been puppy truck all along

What are you talking about, that's the only piece of furniture I own...

I've had pretty good success with just "who, when, where, velocity" for enemy knowledge.

Just record it whenever you can and make decisions about it later, etc.