Mokaygee | Pikmin 1 TAS: All Parts Speedrun in 54:23 [Human-TAS] @Mokaygee | Uploaded August 2017 | Updated October 2024, 8 hours ago.
READ THE DESCRIPTION FOR AN EXPLANATION

Segment times:
IS1: 6:56
FOH: 14:36
FN: 13:31
IS2: 3:07
DS: 12:23
FT: 3:50

(Before we start, there is only commentary for Day 1, the rest has no commentary. This is because it was taken from my stream when I set that time, and I'm happy to leave it in this segmented-TAS. I'm starting to think I should have redone Day 1 anyway, but never mind, it's in there now.)

After looking at this TAS arrangement, you might be asking yourself, what was the point? If you're going to make a TAS, why not make 0 mistakes? If you're going to make a segmented run, well you'd do it in less segments. So what is this?

I would also quickly like to blame some of the sloppiness in this on the fact that I made this in just 1 day. I also used my old IS1 6:56 because I couldn't be bothered to TAS it. But hey, this isn't supposed to be perfect. That's the point! Give me a break, I only spent a few hours on this! Also I'm bad at seamlessly arranging segments, sorry about that.

This is what I'm calling the human-TAS of pikmin 1. It uses all modern methods and routes, but makes some small mistakes and does not get all cutscene skips. It is essentially among the lower bounds of what a human can expect to ever achieve with what we know at the moment. With some Human-TASes, they make absolutely no mistake, just execute all human possible strats perfectly. With this game, that didn't make much sense.

The only TAS tool I used was savestates. While technically this is a TAS, I in fact never used frame-advance techniques to play optimally (only to pause so I could save states). I always played at 100% game speed, so I wasn't able to throw any faster than normal.

I also left any reactionary timings, such as whistling pikmin from a bridge that was just completed, to be completely reactionary. I made sure that I was not gaining time from knowing when the pikmin would finish.

The whole way through was me playing, but using savestates. That said, I did manipulate RNG (by whistling and using the c-stick differently) in some cases to benefit me, such as getting optimal snitchbugs, and the red pellets on day 2.

I should like to point out, this arrangement is actually faster than the previous TAS by a short amount of time, despite not using optimal TAS methods. This is because the route used is much better than in the old TAS. I did not necessarily plan to beat the current TAS, but it just barely did anyway. That wasn't why I started this project (which was to look at theoretical optimal segments, I was considering making a segmented sub-56 speedrun [and I still am]).

I used various slight route changes that are unsafe for use in a real run, unless you fancy resetting all the time, but may be used in a segmented run. The biggest one I can point out is the growth stage of the red pikmin in day 2.

So there you have it. I deem this to be unbeatable by a human with the current strats we already have, unless you play at TAS-like levels of perfection and get perfect RNG.

Finally I would just like to apologise for the video quality at times. I'm really not good at combining clips, and for some reason my emulator goes buggy when recording the final 30 seconds of some TAS-movies. Luckily this never affects the actual gameplay, nor the time taken in any way.

What next? Well, I really enjoyed TASing this, even if it wasn't the technical and slow paced type of TASing you have to do for optimal TASes. I would like to produce a full TAS, with even more timesaving strats and better routes, with perfect execution. It would be cool to see how low I can take it. So, we will have access to a modern TAS and a theoretical Human-TAS by the time I am done.