Bubble Gum Rossie is a game on the famous "Action 52" NES cartridge.
The story follows a blonde-haired girl in her adventure to reach the end of the game using only her bubblegum powers.
This game has a famous glitch where if you die in the second level twice, creating a game over scenario, the game will then attempt to load the first level using the second level's mechanics, creating a glitch scenario.
This level instantly kills you... or so was thought. Spending a fair amount of time navigating the level in TAStudio, it is possible to beat.
Not only this, I purposely trigger the load trigger to allow me to continue with the game normally.
Lastly, this game is known for ending at level 3 and looping back to level 1. It is known to have a level 4, however this is not accessible during normal gameplay and requires cheats/memory hacking to access.
Due to this glitch I discovered, it directly caused another glitch, wherein you can access level 4 through normal gameplay and complete it.
nymx: Claiming for judging.
nymx: Changing branch name to something that describes the run correctly. At this point, Glitched is the only category that I can see will fit this type of execution...whereas 110% was arbitrary and not describing the run as it truly is.
nymx: Updating submission with an improved movie file.
nymx: I'm setting this to "Needs More Info", because it is going to take me some time to figure out this glitch problem. I've talked with dwangoAC to figure out some details on if it can be console verified. Also, I'm going to attempt the manual steps to confirm.
nymx: I've thought on this a bit and I've taken in a lot of angles on what to do here. I think the most convincing argument goes back to the accuracy of NESHawk. If it is truly an emulation issue, then we'll take arguments in the future and look at another submission. Regardless of the frame count (within reason though), it will obviously obsolete this submission...mainly because we want true emulation of the NES console.
We all understand that Action 52 is a horrible package of rushed developed games. Many of them don't have a true ending, which leaves us to determining an ending point that meshes with our rules. In this case, it does appear that the game is so poorly designed and can be tricked into showing content that wasn't intended by the developers. This is sorta like the SMB Minus World runs, where broken coding generates wild and bizarre layouts to explore. Bubble Gum Rossie appears to be a similar case, although achieved much differently. Just note that because of broken code, there are always situations that can be generated that don't make sense. Even I have discovered things like this among recent TAS I've done. So it is not uncommon that breaks happen in this fashion.
I want to state that I have actually enjoyed seeing TASes of these games. As an RTA run, they don't provide much for analysis; however, TASing is certainly exposing mechanics and glitches that aren't necessarily capable human strats. So, its been a fun and interesting ride. I look forward to more.
Thanks to the following for your help over the past few weeks:
{I may have forgotten some that contributed to the talks. If I did, let me know and I will add you to this list}
Also wanted to point out GJTASer2018 for bringing up the possibility of emulation issues. Because this was a concern, we certainly didn't want to dismiss it. Our staff did think it was worth investigating.
Accepting.
feos: Since we're bypassing a dead end in a game without a proper ending, we support the farthest ending point to obsolete all the earlier ones, and that still counts as Fastestcompletion, be it a kill screen or looping back to level one. For that reason I don't think this movie needs a branch label.
despoa: Processing...