Hello everyone, and welcome to our TAS of Smurfs: Rescue in Gargamel's Castle. In this game, you play as a nameless Smurf, trying to rescue Smufette from Gargamel's castle. Your Smurf dies if he touches an enemy, falls into water, or even if he just bumps into a wall in front of him while walking or jumping! As such, the only way you can get past obstacles in this game is by jumping. The B button is pressed to start the game, but after that, it serves no purpose. This is because jumping is done by pressing up on the joy stick. Your goal in the game is to navigate through 7 screens, reach the room Smurfette is in, and then reach the platform she is on to free her. After this, you go back to the start of the game, and the game loops. The game has 4 loops in total. In each successive loop, you go through each room besides the first and last rooms one more time than in the previous loop (ex. in the third loop, you would go through the room with the fence three times in a row before exiting the next screen takes you to what was originally the room right after the fence room in the first loop). Additionally, starting from the second loop onwards, enemies are added to the game besides the one spider in the cave. With each passing loop, the enemies will get faster and more aggressive as well.

The way that movement works in this game is as follow.

Sounds complicated, right? Who would have thought that for a console with only one button on it, the game's manufacturers would decide to make their platformer without using that button at all in the game?

In this game, the fastest form of movement is doing long jumps forwards. The second fastest form of movement is doing short jumps forward, and the slowest form of movement of all is walking. As such, this TAS aims to maximize how much distance is covered using long jumps as opposed to short jumps or walking. Also of note is that if you press up during a screen transition, then you will spawn on the next screen already in mid jump moving forwards, which is faster than waiting for the screen to load before jumping. Whether this offscreen jump is a long jump or short jump appears to depend on which of the 10 frames of loading between rooms you press jump on, and also where in the air you are when you hit the load trigger for the next screen (since jumping is the fastest form of movement, all screen transitions in each loop of this TAS occur mid-jump).

One strategy that I found was the fact that it's possible to do a short jump in the room with the fence before you long jump over the fence. Only a long jump to the right is capable of clearing the fence, and you have to move pretty far from the start of the room before a long jump can get you over the fence (if you touch the fence, you lose a life and go back to the start of the room). Normally, if you start this room from a standstill, do a short jump forwards, and then do a long jump, you will be too close to the fence when you start your jump, which will cause you to hit it and die. Thus, the original strategy used in this room was to run forwards, do a short jump straight up in the air, and do a long jump forwards after you land.

What I found is that if you are in a specific position in mid air when you hit the room transition to go to the fence room, then if you jump during the screen transition, you will start off in the room in mid-fall from a long jump. If you had just the right position in the air when you hit the load trigger, then you can do a long jump as soon as you land and just barely clear the fence by a few pixels. Using this strategy saves about 30 frames compared to going through the room the old way. As such, since this room is visited 10 times in this TAS, this saves about 300 frames (or 5 seconds) over the course of the entire run!

As a result of this, you will sometimes see us do a short jump intentionally right before entering the fence room in order to have the right height upon entering the room.

ViGadeome's discovered while TASing this game that if you held nothing while landing and pressed up again within 30 frames, you could still long jump. This was something I never noticed, since I originally made my TAS of this game holding right the whole time regardless of what I was doing, since I figured there was no harm in doing that (if only I had known how bizarre the controls in this game were!). Because of how fast long jumping is in this game, there were several spots where to get over an enemy that was too close to long jump over, it was faster to wait until it moved and then long jump instead of doing a short jump forward immediately, followed by a long jump to get over the enemy. This was especially helpful in the room with the spider in the cave in the fourth loop, as this strategy saved a sizable amount of time each of the 4 times we went through the room.

While the initial TAS I made of this game was about 3:32 in length, the collaborative TAS that me and ViGadeomes made was able to finish in 3 minutes and 29 seconds. I hope you enjoy watching the product of our work!

ThunderAxe31: Judging.

ThunderAxe31: Optimization seems all right. I confirm that this games features exactly 4 different loops, so there are no ulterior challenges that this game has to offer beyond that.

I also have to note that the player has the ability to choose from which loop to start playing, known as "skill", and that each loop features an ending cutscene. This means that the game can be considered as beaten in a movie in which the player clears the last loop only, similarly to Super Arabian. Still, the other three loops feature slightly different challenges, so the movie for this submission can be considered as full-completion goal, and thus be accepted for Vault. And since full-completion goals always need a branch label, I'm placing the label "all skills", in accordance with feos.

With that said, I need to add that a movie that aims for fastest-completion for this game, that is starting from the last loop, would probably feature completely redundant contents with this submission, and thus be rejected due of superseding mechanics.

Audience response was mediocre, which happens for most Atari 2600 games. I too didn't find it enough compelling. Accepting for Vault.

feos: Pub.