Brandon Boyer

5 Replies

This is what we like to see: something to rival Dan Bruno’s Mother 3 musical malarkey and Kevin R. Grazier, Ph.D.’s fantastic 2007 Halo Science 101 (key finding: “For a Halo with a radius of 5,000 kilometers to simulate one Earth gravity, it would have to spin with a tangential speed of slightly over seven kilometers per second. That implies that the Halo would rotate once every hour and fifteen minutes, or 19 ΒΌ times a day.”) for sheer theoretical madness.

Brooklyn physics teacher Glenn Elert and students have meticulously measured Mario’s rate of descent in each game of the franchise from Super Mario Bros. to Super Paper Mario (the study having been done, presumably, before Super Mario Galaxy — or perhaps its distinct gravitational lunacy instantly set their computational units smoking).

Their conclusion:

Generally speaking, the gravity in each Mario game, as game hardware has increased, is getting closer to the true value of gravity on earth of 9.8 m/s2. However, gravity, even on the newest consoles, is still extreme. According to Wikipedia, a typical person can withstand 5 g before losing consciousness, and all but the very latest of Mario games have gravity greater than this. Also, with gravity that great, it is a wonder Mario can perform such feats as leaping almost 5 times his own body height!

Acceleration Due to Gravity: Super Mario Brothers

See more posts about:


  1. That’s the frame rate and number of frames from the recording, not the game itself. If the recording was sampled at 30 frames per second and an action covered 45 frames, then the amount of time it took would be 30/45 seconds or 1.5 seconds, assuming a constant framerate. Isn’t this essentially how time is measured using a high-speed camera for things like crash tests? Admittedly whatever they used to record probably wasn’t that granular or accurate, but it seems like the numbers are being handled correctly.

  2. 30/45 != 1.5. 45/30 == 1.5. Frame rate == frames/seconds. Frame rate / # of frames == 1/seconds. Other than that, the data looks relatively sound, at least until they start grasping at straws trying to find a trend vs. the bit rate.

  3. Wow, I failed at basic arithmetic. So someone accidentally transposed two factors in a formula. It’s a mistake, to be sure, but all the data matches the correct formula of time = frames/framerate. I just don’t get a comment like “The science only gets worse from there.” It seems to miss the point of someone doing a paper on gravity in video games.

Leave a Reply