User Files from MESHUGGAH

History:

Also note that all unnecessary jellybean switching is removed.

Ends in 4902 frames. The end.

86 frames faster than Blobv2 (last updated movie file in his workbench submission)

Some things you maybe interested in:

I'm going to restart the TAS alone with new tricks like OoB without jellybean throwing.

Revamped random hack and slashes. Trojan now usually sync/act/move to sound when he can, or just shows how crazy is he.

Stage 1-1 to Stage 3-2 is same regarding frames.
Currently 104 frames faster @ Stage 4-2 start than adelikat thanks to manipulation enemy spawn side.

Comparison sheet: http://docs.google.com/spreadsheets/d/1qW7_UFKPhVVKdtoNhUTDTOdic0MqA9Zo0AlquZMZfH8/edit?usp=sharing
LUA script (not all object groups are implemented yet): http://pastebin.com/Lq5DDLCc

4-1 dog doesn't hits red block so he follows us even with 1.5px sacrifice (or delayed spawn, etc).

Shoudl need to change slot 5 pos or slot 7 or time.

Scumtron, god of ninja gaiden just popped in (twice) and solved our problem.

238 frames faster than feos pacifist test run. Level comparison/notes: http://docs.google.com/spreadsheet/ccc?key=0AjzBXzk5TnLYdE8ybHpRZTFVNWI4bEZwWTFqZ0I0Y1E&usp=sharing

36 frames faster than adelikat up to 3-1 end (there's 7*2 stages (2 part each stage).

Spammed various patterns throughout to movie to decrease the monotonity of jumping and slashing. No complaints allowed.

Delay route is same and fastest currently (possibilities is 16 subpos * time value (maximum delay which could save time is something like 5 frames, this mostly depends on level architecture and the possibilities of spawning sides for enemies) * number of enemies need to be manipulated (usually 1, sometimes 2 edit: and don't forget to add yourself in the calculation, as your position also manipulates and carries over stages as well as from the map screens).

Level comparison and notes: http://docs.google.com/spreadsheets/d/1qW7_UFKPhVVKdtoNhUTDTOdic0MqA9Zo0AlquZMZfH8/edit?usp=sharing

Added scumtron's 0.5px to stage 2-2 second screen final part to hit the end 1 frame faster.

Movie resynced to 3-3 boss. Currently we kill the boss at timer value 1. If I'm correct, we need timer value 60. If scumtron's correct (his logic is more sanic, check NG1 thread last posts), then we are much farther away the target timer value.

edit: comparison table http://docs.google.com/spreadsheet/ccc?key=0AjzBXzk5TnLYdE8ybHpRZTFVNWI4bEZwWTFqZ0I0Y1E&usp=sharing

28 frames faster than master adelikat, up to the armadillo boss death/end (Stage 2-2 perhaps?).

Randomness depends on time (not in-game time) and player X subposition. Therefore I
- Delay 1 frame starting the game to spawn first enemy on left side
- Delay 1 frame at the end to manipulate next level first enemy position
- Delay 2 frame after 2-1 boss and before 2-2 boss.

Player subpos ranges from 0,16,32,48...240 (16 values)
Timer I believe is only a byte, I should write down more values.

Level by level comparison: http://docs.google.com/spreadsheets/d/1qW7_UFKPhVVKdtoNhUTDTOdic0MqA9Zo0AlquZMZfH8/edit?usp=sharing

Have no idea if I posted my lua script already, here's a totally not newbie friendly one: http://pastebin.com/hrJDu5VV

Removed the 1 frame delay on 1-2 boss which was required to manipulate 2-1 enemies.

This changed a few things, especially 3-2, and now I'm 1 subtimer (in-game time) behind on 3-3, which makes me lose 3 frames (timer rule).

Regarding previous "required 1 frame delay":
The last part of 1-1 consists important slots in this TAS.

I've tested all the possible sacrifices BEFORE the 1-1 ending part (last tyson recovery walljump) but either slot 7 (rwj) or slot 4 (last tyson hitting) became worse because of the improvement beforehand, however only losing 0.5px on RWJ deemed to be the best and fewest sacrifice to manipulate everything miraculously.

So, I need to remove 1 interrupt again from the route, not sure if it's possible.