MonitoLol proposed to exclude loads for Any% Unrestricted runs. I would therefore like to gauge the interest and receive feedback in regards to changing the timing method for Any% Unrestricted to exclude the time spent in dimension loading sequences.

First let me outline how runs have been timed so far.

Time starts on first input and ends the first frame blocks are visible after exiting the Nether portal at the end of the map.

I got the inspiration for that timing method from this pdf made by the Minecraft Java Edition Mod Team: https://drive.google.com/file/d/1EkbgSrFERH7udaPi94_RN0Mq-Xnfq1uA/view?usp=sharing

I imagine that the new load removed time would run from first input to the loading screen on the first Nether portal. Then run again from when blocks first appear after loading in until entering the Nether portal at the end, on which timing would officially end.

I would like to know if runners are interested in this change and hear suggestions. Please respond to this thread if you have something to say.

I think it would be a good change, for a long time this game was never pushed this hard, now runners will be fighting for seconds of improvement at most. I personally think it would be sad to see a near perfect run miss it’s chance to be wr because it was performed on a weaker machine.

Also this change could be applied to the regular any%. I don’t really see why not?

Yea I think it would make sense that all categories follow the same timing method. I want to try it out for Any% Unrestricted first though, because there is a world where it makes no difference and it would just kind of be a waste of time and overly complicated.

I think there is much time left to save, so maybe we should start removing load times if it is needed at any point because there isn't much to improve. Otherwise if a new player jumps into the action and has a significant disadvantage due to his PC, we can still remove load times. As Holm said, it just might be a waste of time, especially at this point of the speedrunning journey.

Here's an update on things: I retimed a few runs to see what effect it would have. The time that would be removed from a run is around 0.7-1.5s (obviously depending on the load time). Don't take these numbers as any sort of accusation or conclusion, this is very preliminary testing.

MoustacheBleue and I have the longest loading times at around 1.5s in the runs I checked (I will release a document with all the timings once I get a bit more measurements of older runs for better accuracy) MonitoLol has the shortest loads with around 0.7s in total and 18ologic is in the middle with 1.1s. Again not robust numbers.

The difference in loading time is large enough that it plays a role for a world record level, but I think that it is not totally hardware related. I think that there are some in-game factors that could potentially effect loads and that we should test different settings and game versions, before we consider changing timing method, it's possible there's an ingame solution. First of all, because we don't re-time atm, fabricMC Sodium needs to stay banned, because it lets you move earlier when exiting portals. I think optifine would be better than vanilla purely because it's that bit more optimized and I also think that a lower render distance is likely better. In my run I ran on vanilla 1.7.10 with maxed out render distance. I have the game installed on an NVME drive, which is fast enough that I would not expect anyone to have a large advantage over me if the main constraint was PC loading speed. EDIT: Another thing is texture packs. I completely forgot. It seems that using the map's texture pack might be correlated with worse loads, which is kinda sad.

Actually, my load times were longer at the beginning. That's why I now throw an item before entering portals to shorten loading times.