Some RTA (Real Time Attack; time with loads) categories are additionally timed with LRT (Load-Removed Time/ing; time without loads) on the leaderboards. This is because there is a large amount of variance in load times between consoles; for these leaderboards to be competitive at a high level, the playing field needs to be equalised.

In general, digital versions of the game (Wii U eShop AKA VC, and USB loaders) load more quickly than physical versions (Wii or Wii U disc, and all Emulator runs); this is because digital versions very consistently have no extra delay when you select a level (that is, the time between selecting a level and being able to select 'Play') or defeat a boss (that is, the subsequent animation on the map). This is particularly important in categories where you use a completed file (Mirror Mode, 200%, File Completion and all Full Worlds). Digital versions also save time on the results screens after levels (there's some inconsistency with how quickly you can exit out of them) and on the loading screens when exiting levels, though interestingly, physical versions tend to save time when entering levels.

Even then, though, there is a lot of variance between different players' digital copies and physical copies. As of writing this post (March 2025), the difference between the fastest and slowest digital (not including USB) loads in an Any% run is 1:49, which is an enormous difference; physical versions can also have a lot of variance. This is why LRT is necessary.

LRT spreadsheet + Current usage on the boards

Unfortunately, many types of loads in this game are not detectable by an autosplitter, and even though some are, using an autosplitter of any kind also creates problems with varying video quality (there are plenty of players who don't have a capture card). Therefore, the only way that LRT can be calculated is manually, which I've set up a spreadsheet for.

As of writing (October 2025), only Any% is currently retimed without loads, however I would like to also get this set up for All Stages, Full Worlds and possibly All Puzzles in the next year. Because it takes a considerable amount of time to retime a full run, only runs below a certain time barrier are retimed (similarly to the NSMBW boards); this barrier for Any% is sub-1:35 RTA. When submitting a run that needs to be retimed, please be aware that it will take a bit longer to verify than other runs (especially if the video is from a website other than YouTube), however it should take no longer than a week to get the LRT time.