I think the real basic secret of TASing is your mindset going into it. Thoughts are what ultimately lead to actions, so thinking the right things will eventually lead to solid publication-worthy TASes.
The way I've been looking at TASing recently has been something like this:
1. My first attempt is never good enough.
2. There's always a faster way, I just need to find it.
3. My knowledge is just as important as anyone else's.
4. Each screen is its own game.
1. No matter what it is, if it's a brand new movie for a game not on the site or an improvement to an existing movie, even if it's an improvement to your own existing movie, you'll always be able to improve your first attempt. You'll honestly be surprised at how much you can improve your first attempt when you attempt to improve it.
The key is that with each attempt, you learn more and more. It's like learning any other skill: You never start off being the best, but each time you try you do it a little better. You run through a stage for the first time, it's not perfect. You try it again and you start thinking
"Okay, I think there's a jump coming up? I should prepare for that." You try it a third time and start thinking
"Maybe I could land earlier after that jump and set up for a second jump to bypass that slope?"
Eventually, you get so familiar with the section that you can start planning long in advance. That's when you know it's safer to move on.
2. This is about testing. Lots and lots of testing. Test every route, every decision, every frame, and when you think you've found the best one, keep testing to see if there's something better. This also applies to constantly watching and re-watching your run every step of the way. Ever since DTC5 I've gotten into the habit of watching my progress a run I'm working on every single time I start working on it again, and I'll often think
"I can probably do that faster" and most of the time I'm right.
3. I tend to go into my first run of a game "blind". I don't look up speed tricks or information until I'm finished with a test run and start working on something I may want to submit. There are a couple reasons for this, one being that I just like discovering things on my own. It's a really nice motivation boost, something that keeps me going in those moments where I think I'm making no progress.
"Hey, you found this glitch and the world has to see it!" or
"Hey, this looks awesome, you gotta keep going!" are thoughts that go through my head a lot.
Second reason is because when I've looked up information before starting a run, I overlooked things that I should have tested, assuming that the information laid out before me is all I needed to know. Moon Crystal comes to mind almost immediately: My initial improvement was just optimization and lag reduction, and Hotarubi's currently published run included a lot of strategy and route changes I hadn't even considered.
To put it more simply: The less I know about a game, the more I end up discovering by the time I've finished. I only look up information when I don't think there's any more information left for me to gather on my own.
4. This is sort of related to point 1, but my method has always been about setting benchmarks and trying to beat them. This is more or less how I approach each screen:
"I did this section in 1000 frames. Let's see if I can do it faster."
"Okay, now I did it in 995 frames. I'll try a different route this time."
"Okay, that route was 996 frames. Maybe a second look would be beneficial...?"
"Yep, it's down to 992 frames now, but looking back I think I did this a little slow..."
"Interesting. I did the first half of this section in 465 frames in this last attempt, but 460 frames in the first attempt. Where did I go wrong?"
"Ah, okay. So maybe the first route is faster? I'll try it again..."
"Aha! It is
faster!"
TASing is like any "real" long-term goal, almost like losing weight. You have to treat it as a series of small goals instead of a single large goal. You shouldn't go into a TAS thinking
"Okay, I just started this, I have to get through 30 minutes of game." You should sit down and think
"Okay, I just started this, I have to get through 1 minute of Stage 1." Small victories are motivating, and you'll always get to the large victory at the end regardless, so why not keep yourself motivated along the way?
I apologize for the lengthy post. I probably
think about TASing more than I actually sit down and
work on one. Note that this is just how I think about things and that these points aren't exactly guaranteed to help, but I'd recommend at least keeping them in mind if you're new to the hobby and looking to improve.
EDIT2:
Link to video
It's a really interesting level!
Maze of Galious works fine on BizHawk, that 260-in-1 thing doesn't. I think you might be out of luck on that one. Shame, really.
