TL;DR I have reason to believe that the ways used to speed up RTA are not consistent, but IGT is.
So, I was poking around, looking for ways to speed the game up.
Tonight I found a couple more:
All of which intending to reduce the amount the game renders, because I suppose AoM has no upper limit when in singleplayer. Even on the same computer, the speed-up seems kinda inconsistent, although maybe it's level-dependent due to stuff happening offscreen, I don't know.
However, I also found another 'strategy': letting the game run minimised. As a quick test, I did this and took level 18 (tamarisk tree) from 5:14 to 3:30. This is bad because you can't record a minimised AoM.
As minimising the game is a very easily-replicable thing, rather than aligning the camera perfectly in the same game state etc, I did some timing tests with the very long autoscroller. I got slightly different outcomes, and even more different when i fiddled with some computer settings, how i was recording, etc.
This leads me to believe that the methods used to speed up the game are inconsistent, and therefore not fair. By contrast, I timed Achilles running corner-to-corner in the Editor with and without these various speed-up methods, and he always took the same time IGT (51 seconds) despite wildly different RTA times.
Due to the odd specificity and granularity of RTA speed-ups, I do not think they could be practically banned or limited. Therefore, I think we should use IGT for IL's.
(as for the problem of "how much opening cutscene you watch affects where the IGT starts", one could just subtract that from the ending time)
|SarsartSarsart and Adam_akAdam_ak like this.|