This isn't at issue. TG has tracks for emulation and arcade. Submitting a game to the wrong track is against the rules. TG alleges that the plaintiff passed off emulated gameplay as live arcade play.

Arcade and emulated play have different verification standards. Generally you need to reset the cabinet, which shows the machine's uninitialized memory on screen, post game you need to show the board. For emulated play you typically need to provide a full recording of the game, which can be re-run in the emulator.

One difference between arcade and emulation comes down to inputs. The arcade joystick has a 4-way restrictor. When Jumpman is climbing a ladder the player can't provide left/right inputs. At a high level this impacts their ability to steer and group barrels. This is totally irrelevant apart from the highest levels of play.

With a emulator it's easier to use save states and splice different parts of the run together, or to otherwise cheat. Given that Billy Mitchell has stated that he never used a emulator but the footage contradicts that, a lot of people believe he may be lying about using an emulator to hide the fact that he was cheating using an emulator.

This is a deep and complicated question, but the top-line answer is it's the best way to prevent cheating (note I didn't say perfect). A big part of these classic games is how the random events are generated. It would be fairly easy to modify the game in an emulator to more often generate ideal random events, giving you an edge over people playing on the real hardware. So the community agreed that for the purposes of the leaderboards in question here, the game must be played on real hardware.

It's not only about obvious cheating like slowing down the game.

It might be easier to do certain things with different hardware setups.

An illustrating example: If you play quake 3 at 125 fps you can jump farther than you could if the game was running at 60fps.

While there are still some people who treat the arcade pcbs as inherently superior to mame, the general consensus nowadays (at least for DK) is that mame plays arcade perfect.

In fact TG has a separate leader board to track the mame high scores. Which is one of the most frustrating things about this whole affair... if Billy really wanted to use mame he could have just submitted his score under the mame category instead of lying and submitting it under the arcade hardware category!

There are two main reasons:

1) It's much easier to cheat (in many ways, adjust random generators to always be in your favour for example).

2) There can be differences (the differences are how he has been caught) -- it's common that games in emulators are faster as they don't bother implementing waits of various kinds correctly, as most games still work fine (often better, as they'll skip less frames), and it would just make your emulator slower.

Timing differences mainly. Plus emulators can be played at a slower speed and then the video sped up later.

The real reason is that Twin Galaxies were quite old fashioned (maybe still are?). Didn't allow emulators, glitches, various other things. It's partly why most speedrun leaderboards are kept on speedrun.com now, as the games communities get to dictate the rules, rather than the owners of Twin Galaxies.

Low level differences between emulators and the physical hardware could make a big difference to the high score/speedrun in theory, especially when developers take advantage of esoteric or undocumented features.