I have a lot of history with this game, and was also wondering how well I would be able to make a TAS. So I did the only thing I know how, jump right in and learn as I went. I am a rather big fan of speed runs and TAS's, I hope my TAS will be welcomed.
Game Objectives: Beat the game with out killing any enemy unless it hinders forward progression (bosses and such).
- Emulator used: FCEUX 2.1.5
- Pacifist
- Contains speed/entertainment tradeoffs
- One player in a multiplayer game
- Heavy luck manipulation
The TAS took just under 2 months to make. That being between real life stuff and the game desyncing SO often. But I feel it was worth it. Please note that this was all done on TAS Edit. TAS Edit likes to crash often, reset tweak counts every time it is reloaded, and just over all is not the best thing to use. But for what ever reason my computer would not allow me to use a controller with an emulator. So I was really only given one choice.
My tweak count almost always loaded up at ~30k tweaks. I ended a bunch of sessions with that count well over ~200k tweaks, but reload and it is back to 28-30k. I don't get it but but it is not a big deal. I did stream a bit of me making this and was given live feedback or ideas when I was unable to make forward progression. But I NEVER resorted to killing anything. I did almost restart to hit and pick up weapons from weapon pods for extra points giving me extra lives for stage 7. But then it just dawned on me how to make it to the end without having to restart.
Stage 1: More or less strait forward. I chose to not manipulate most spawns as I felt it gave more entertainment over the Contra TAS. That was a general rule I used unless a section was either really tight or the spawns make it impossible to pass. First up is the first cannon. This took me a while to get past and I am glad I was able to in the fashion I did. I also went with "M" and not "F" because I was testing each for best time, but I was able to get the last turret on the boss to not spawn, making it the best choice for what my goals are.
Stage 2: I was able to put the tank skip I discovered (but AnS was able to explain to me, and also thanks to scaryice for helping out with it as well) into use and saved 1 kill. The final bridge was a real pain as I needed to manipulate more than I wanted to. But I did have to despawn the red guy standing in front of the door to prevent yet another unnecessary kill.
Stage 3: This is the stage I was most worried about. But it turned out to be rather fun and a lot less stressful than I thought it was going to be. You may notice that just before the cannon, a red guy tries to shoot at me, but nothing comes out. Forcing this to happen was really the only thing that allowed me to pass the cannon smoothly without much manipulation of the spawns. The water section up to the mini boss was all rather easy. I did have to manipulate the enemies jump height as they entered water. Besides that, it was just me trying to troll as much as I could.
Stage 4: The opening bubbles were a real issue. I ended up having to pause for a few frames right at the start and was able to get past the first set a bit slow, but the second set was very fast. Most of my efforts ended up with either both being rather slow, or one taking a very very long time. So I kept this. At the first set of falling bombs, due to so many sprites, some of the bombs despawned, allowing a faster ascend. After that I was just trying to make this stage as entertaining as I could. Hope yall enjoy it.
Stage 5: This stage is evil... I had so many desyncs on this stage that my stream viewers had a fun time laughing at me getting angry. But I was able to push on. The section with the falling rocks had special attention as I wanted to either skip the 1 falling rock (cant be done from what I can tell) or make it spawn early so i would not have to wait. But the jetpack guys had plans to make that as hard as they could. I ended up having to slow down just a bit before the rocks to allow the jetpack guys to be a bit further up, allowing me to pass smoothly, or at all.
Stage 6: Figuring out where in the stage to abuse the manipulation of the spawns was rather hard. I needed to end up on the right side to pass the third area with the red things. I also spent a lot of time on getting the position for the boss right as i was getting a lot of desyncs just at the boss. And that is why I have dead time... sorry for that.
Stage 7: I got rather test happy with this stage. Trying to get the purple pillar things to shoot at the correct angle and be in the specific location to be hit by them was really time consuming once the idea of how to properly pass everything formed in my head. After respawning, I shoot the pillars to make them move forward so i would have enough I frames to pass threw them and be able to change my direction a bit sooner as well (Thanks to TheReflexWonder for reminding me to say that). I was getting stuck at the pods before I figured it out though. By the way, the hit boxes of the pillars and the pods are WAY TOO BIG!!!!!! I had a lot of fun with the stage 7 boss as I was far away and did not have an insta kill.
Stage 8: Most of my time in this stage was spent on the start and the first set of gaps. But I was just trying to make it as entertaining as I could. Please note that the hit boxes for the homing red things are huge too. I tried to break the most boring part of this stage by getting clipped into the floor a few times. Just to give some viewers something to do a double take at.
I would like to give a huge shout out to Mayor K, as he found all of the RAM values that interested me at any times notice.
Something fun to do is watch this with the .lua script that shows the hit boxes. Pasky13 was a pal and posted it on page 4 of the Super C thread.
Improvements can be made just by simply going for pure speed. But I think that would take all the fun out of this run.
I can only think of 1 frame that would be nice, hope others can find better ones.
Frame 29424
Hope you enjoy!
Nach: There were many points to consider in judging this run. What exactly is pacifist? Is it optimized enough? Is it entertaining?
The exact meaning of pacifist is debatable. Some would consider pacifist to be a run where you don't kill anything unless it prevents you from progressing, or you have no way out and would die unless you kill something (self defense). Others would argue that last point and would require pacifist to include self sacrifice if too many enemies fill the screen at points. Another point, is a pacifist run allowed to use more powerful weapons when fighting enemies that prevent progression? Or is one required to always stick with the most peaceful firepower available.
Since there is no clear definition, I don't find this run to violate its own rules of being labeled "pacifist". If a future run wants to obsolete this with a different meaning of "pacifist", we'll discuss it then.
Regarding the optimization, it was pretty good in most places. In the few places where it wasn't, more enemies swarmed the screen which led to trickier situations. These tricky situations lend well to the goal of TASing in general, act like nothing fazes you, and run between bullets and so on. So while they can be done faster, the non-optimal segments did increase entertainment.
Overall, audience response about entertainment was good. Therefore, I am accepting as first run, and welcoming a new branch which showcases what TASing is all about.
Ilari: Processing