To ensure that runs are comparable in a competitive way, the Version annotation has been changed to a subcategory. This means that the leaderboards are now split based on version. This only applies to All Stages as Any% is currently only done on HTML5.

tldr of main things that changed:

• Timing starts after skip instructions is pressed instead of tiles forming on stage 1
• Permanent video proof of sufficient quality is required
• Video must show starting new game from titlescreen and clearly show skip instructions button press
• In general, game rules, category rules, and verification guidelines have been written out more clearly
• Original 10/20/33 Level categories have been grouped into All Stages category to match Any% organization

As of 2022-05-27, the game and category rules have been updated with a change to timing being the most significant change. The aim of this update is to address some issues with timing, make the rules more clear and robust, and introduce some verification guidelines. The motivation behind updating the rules was to address some issues that have existed within timing and to improve the overall quality of the boards.

The old timer starting point of tiles being fully formed in stage 1 was somewhat arbitrary and this didn't get changed when Any% was added to the boards for retroactive timing compatibility. Players typically used an offset of around -3.3s from pressing skip instructions to start livesplit around when tiles would be fully formed. Timing now starts upon pressing skip instructions because it's a very clear starting point, it lines up with how players start livesplit, and it future-proofs timing for potential autosplitters since the button press is easily detectable as opposed to tiles forming.

The biggest issue with timing that is addressed but not solved, is that load times and movement of the block vary per machine, browser, and game version. For game version, a variable can likely be added to label whether runs are Flash or HTML5 without splitting boards. Many possible solutions were considered to normalize timing discrepancy, but most options introduced more problems than it solved, and runs would still have to be retimed for all solutions anyways. Only the timer start point has changed, so the way timing works now still does not normalize load times. Runners seeking competitive times are advised to determine optimal running conditions including factors like browser, game website, and number of processes running in the background.

The last thing is that some runs on the board were not permanent videos, not sufficient quality, not the correct game, or were timed inconsistently with respect to other runs. Verification guidelines have been introduced to maintain a standard of quality for the runs on the board and to make verification decisions easier. All runs on the boards have been retimed using a consistent start and end frame, and will continue to be consistent across runs moving forward.

Aside from rule changes, the categories have been reorganized to be consistent format and align better to the definition of a category. The original 3 categories of 10/20/33 Levels have been placed as subcategories in the new All Stages category. Around the introduction of the Any% category, these original categories were referred to as Glitchless. All Stages does not allow glitches and Any% allows glitches. All Stages was chosen over Glitchless for the category name since it more accurately describes the category in a way that properly makes a distinction from Any%.