March 26th, 2009
I read a scientific article on a topic of which I have some experience, namely NES-era Mario Bros physics. It was “Acceleration Due to Gravity: Super Mario Brothers“. I was very disappointed in the quality of this article, as the entire basis of the article was made under more than one incorrect assumption. For shame Adam Lefky andÂ Artem Gindin.
When I created this demo my usual formula for velocity/acceleration used on Mario simply did not work. I had to look carefully at his movements until I realized his acceleration was not constant. After some experimentation I estimated his acceleration downward to be 0.6 in pixels per frame (1 frame = 1000/30 ms) before the peak of his jump, then it dramatically increases to 3 after the peak of his jump. This produces Mario’s hovering jump, which can be hard to notice unless you’re looking for it, but it’s there. This is not taking into account when a player releases the jump button mid-jump, which also changes the acceleration.
Additionally, there is also a maximum velocity in the Mario world, which probably came into play for some of their experiments. Not that they’d notice while wontedly ignoring the scientific method. I am very disappointed in you Adam Lefky andÂ Artem Gindin.