[Original thread title: "Discussion: Thoughts on changing timing for 100-man to eliminate load time discrepencies?". I changed it to reflect where the discussion landed.]

Hello!

Load times have been a headache for this game for a while. The system, game cart type (physical/digital), and even SD card all have an effect on load times. This is a problem for every category in the game, but it's especially noticeable in the shorter 100-man categories where even a few seconds can separate leaderboard spots. As a result, we'd like to consider changing the timing for 100-man categories such that load time differences are no longer a factor.

There are a number of ways we can tackle this, but the general proposal is to change the official timing to be the total time spent in levels + some set time value per level that represents a standardized load time. We would keep the RTA time on the leaderboards but add a column for this "standardized time" and sort the leaderboard on that value. This would even out the game and allow everyone to be competitive, regardless of their set-up.

Basic poll on this idea (please participate - poll closes Wednesday): https://strawpoll.com/hy6wyucqr

Some more detailed points that are up for discussion:

1. What should the per-level standardized load time be? The reason I think we should add a per-level load time rather than removing them entirely is so that there's still a penalty for skipping levels similar to the penalty that would exist in RTA timing. My thought here would be to take an average of load times on the fastest known system/cart/sd combination so everyone gets that advantage, but would love to hear other thoughts on this.

2. How should individual levels be timed? We can standardize on RTA timing for the level (for example, start time on the last full-black frame before the level starts to appear, and end time on the first frame where a score shows up on the flagpole/axe/goal tape), which would be the closest to today's timing but which would be the most technical to retime and could be intimidating for newer runners. Alternatively, we could use in-game time (final timer - starting timer) which is much easier to use but reduces some of the sub-second skill expression. For the in game time we could also consider adding standard loads for screen transitions (3 seconds for a pipe, 3 seconds for a keydoor, and 1.5 seconds for a door, for example) so that starts that are faster RTA are still faster IGT. Thoughts?

Please let us know by responding to the poll (https://strawpoll.com/hy6wyucqr) and leaving your comments either here in this thread or in our discord channel - this affects our most popular category so we'd love to hear from everyone who's willing to share before making any changes :) Thanks!

Update: The first poll results are in, and the decision is unanimous! :D We'll be working on a rules update for timings that will remove the load time discrepancies caused by system/version/SD card differences :)

That brings us to our next poll! We've been working on drafting the new set of rules and we're running into a lot of situations where there's a pretty clear tradeoff between precision and simplicity. Some examples:

Please let us know what you think at this poll here by Friday and leave your comments below! POLL: https://strawpoll.com/pf32o4ka4

Update: Poll results were an even split :P Thanks to everyone who voted and gave feedback - the next step from the mod side is to put together a draft of what the new timing rules will look like taking all of this into account and then share it back once it's ready :) Thanks for your patience! :D

I know I've been quiet here for a bit, but I promise it's because we've been hard at work making this happen :) We're getting close, and now that we are I'm hoping for some help :)

*** COMMUNITY REQUEST *** If you're playing 100-man and encounter a weird level ending (anything out of the ordinary happens after you lose control of Mario) - please save the footage and send it to me if you can :) We're in the process of testing out the way-more-complicated-than-I-wish-it-was flow chart to account for end of level load times, and could use everyone's help finding the edge cases :)

UPDATE: We have a draft doc explaining the rule change. It goes into a lot more detail than I think most people want, so just stop reading once it makes sense to you or once you get bored :P As long as you feel like by the time you're done reading you understand the change we're good to go. I've made this link available for commenting so if you have questions or feedback feel free to comment in the doc directly or here in this thread. Please comment by SUNDAY so we can hopefully make the rules updates official :) Thanks! Doc: https://docs.google.com/document/d/1iJ4F4uNq6MEDQ-O_1cfsahyGQBSepVdMIjFC3RABHpI/edit?usp=sharing

I'm sorry for the lack of updates here, but a lot has been going on behind the scenes. We've got a ruleset now, and we've started retiming runs, but as that's happening we're finding new things to investigate. Plus it's going to take quite a while to get through them all :( Thanks again for the patience - I promise we're still working on this!

We're currently figuring out how to deal with internet lag when in the map and when you skip a level. Since we're trying to make everything as fair as possible, people who have a bad or a poor internet connection can when transitioning from level to level have a slower time getting into the next level because of the quality of their internet. Mitigating this has become quite the challenge as the variance is quite wide from one video to another.

Good news - WE DID IT!!! We’ve managed to find a way to get rid of the load times and gone through and retimed every single 100-man easy run. We’re going to update the leaderboard and rules soon, but just to be clear on what’s changing I’ve included both versions of the rules below (it’s just a new section titled “For Easy only”). Depending on how this goes we can look into other 100-man categories next, but I think we’ll take a break before doing more investigation here - as you can imagine this was a huge effort. Some fun stats:

• Time taken for retimes and new rules: Over 2 months
• Total runs retimed: 30
• Total load time removed: 41 minutes
• Total load zones timed: 1,193
• Total level endings timed to calculate level freeze load zones: 325+
• Total level ending combinations that affect level freeze: 64
• Total spinners timed to calculate spinner load zones: 269+

One line in the rule changes that I want to clarify here is this one: “Moderators can at their discretion choose not to retime a non-competitive run if it has excessive load zones, in which case the RTA time will be used.” That’s really only there so that if someone submits a troll run with 100 skips in it we don’t have to sit here and actually retime it - if you’re going for a good time in good faith we have no problem retiming a run even if it’s not at the top of the leaderboard :)

Finally, congrats to N00bie371 who is the new WR holder in this category! Thanks again everyone for the patience, and best of luck with the new timings :) I’m hoping to see some new WRs here soon :D

Current rules

Easy, Normal, Expert or Super Expert.

• Time starts on activating the "New Game" or "Yes" button and stops when you hit the Axe in the final level.
• Skipping is allowed.
• Video is required.
• Playing off of a cia file is permitted.

As of June 12th 2021 runs that are done using an emulator are permitted for the 100 Mario Challenge category.

• If possible, please show the Emulator border when capturing your gameplay.

New rules

Easy, Normal, Expert or Super Expert.

• Time starts on activating the "New Game" or "Yes" button and stops when you hit the Axe in the final level.
• Skipping is allowed.
• Video is required.
• Playing off of a cia file is permitted.

For Normal, Expert or Super Expert.

• RTA timing is used for this category and loads are NOT removed

For Easy only.

• The leaderboard is ordered by a loadless time instead of RTA. Moderators will retime each run after submission. For accurate retiming, try to keep at least the top screen of the 3DS on camera at all times and ensure that the in-game timer is legible after finishing a level.
• Moderators can at their discretion choose not to retime a non-competitive run if it has excessive load zones, in which case the RTA time will be used.
• You can view the full loadless retime rules and details on retimed runs here: https://docs.google.com/spreadsheets/d/10-yHDtCA4Jih2Z0Q4RuID3E2A4NjUe3lvclrPXOZzAk/edit?usp=sharing

As of June 12th 2021 runs that are done using an emulator are permitted for the 100 Mario Challenge category.

• If possible, please show the Emulator border when capturing your gameplay.