Submission #5560: node1729's Genesis McDonald's Treasure Land Adventure "Beginner (NTSC)" in 19:59.49

(Link to video)
Sega Genesis
Beginner (NTSC)
BizHawk 1.11.6
71877
59.922751013550524
7485
Unknown
McDonald's Treasure Land Adventure (USA).md
Submitted by node1729 on 6/20/2017 11:28:16 PM
Submission Comments

McDonald's Treasure Land Adventure

Most people scoff at the idea of a video game based around McDonald's, however this game was produced by Treasure, the same company that made the games Dynamite Headdy, Wario World, and Mischief Makers. Ronald and his friends find a piece of a Treasure map, that leads them on a search, along the way they encounter a Giant Tomato, a Goblin-like alien, and a pirate named Captain Lips, each of these bosses give Ronald a piece of the magical treasure map, which then changes its form and leads them to the moon. They hop on a rocket and travel to the moon, the rocket crashes and they have no way home. Ronald is abducted into an alien spaceship, where he must then find and fight the final boss. The final boss guards the treasure, which was a chest that allows Ronald and his friends to travel home, which is all they wanted in the end.

Reasons for choosing Beginner difficulty over Normal or Expert

  • Beginner has the fewest amount of screens, cutting out on quite a few autoscrollers.
  • Beginner is the shortest category, so this is also effectively an any% category, but it does not obsolete Sonikkustar's Expert TAS.
  • This is a different category from Sonikkustar's TAS of this game
  • Beginner is the most run version of this game on the leaderboards.
  • People in the MTLA community wanted to see what an ideal run would look like without any mistakes, which are always present in WR runs, no matter how small.

Controls:

  • C - Jump
  • B - Attack
  • A - Scarf (allows Ronald to grab onto hooks)
These buttons can be changed in the options menu.

Basic Mechanics:

  • When on the ground, using the Scarf and Attacking force Ronald to stand still.
  • Jumping allows Ronald to use the Scarf and to Attack while moving.
  • Crouching and Attacking optimally has 16 frames between each attack.
  • When Ronald falls to the bottom of the screen, if he has a balloon, that will be used to save him. This can be controlled.
  • The amount of damage Ronald does is based on his Magic Level. The maximum level is 3.

The Run

Glitches/Exploits used in the run:

  • Damage Boosting - This is pretty straightforward, when Ronald takes damage, he is boosted upwards and backwards. Ronald must be moving into an enemy to get the damage boost. This is used twice on 1-2 (Tree) to get onto platforms that have an otherwise slow path to get to.
  • Scarf Glitch - The scarf can be used to avoid the boosting effects of damage, however Ronald still takes damage. This is used extensively in the run.
  • Animation Cancelling - Many different places in the game can be exploited to cancel the attacking animation. This allows attacks to happen quicker than the standard 16 frames. Most notable usages of this are during the first boss, and the final miniboss.

Monitored 68K RAM addresses:

Address Function Length
A640 Boss Health 2 Bytes
The address for Boss Health has some interesting properties:
  • None of the bosses had Health that used the upper byte, yet when modified it still gave the boss more Health.
  • The first, third, and final boss die when their health is less than 0x0064 (100), and the second boss dies when its health would be less than or equal to 0x0000 (0) after being attacked.
  • Modifying the boss's health below their kill value resulted in them dying, yet not triggering the cutscene for killing them.

The Secret Code:

  • There is a code that can be input during the main menu. This unlocks a level select by pressing START GAME, or going to OPTION allows the language to be set to Japanese. This saves over a second in the run.
  • (The code is Left Right A B C)
  • The run is done on the US ROM, yet in Japanese because the Japanese ROM has more levels, so is slower.

The Major Discovery:

On the second Ship level, after encountering Captain Lips, there was a new strat that was found by me, that skipped a cycle, it saves 4 seconds over the previous WR strat.
  • This is the biggest time save found in the game, and is what allowed sub 20:20 to be possible RTA, although it hasn't been done yet.

Time Saves over current RTA WR (20:20 by Boon_7)

This will go over the Time Saves that are not basic movement.
  • On 1-1, after using the second zipline, landing on the platform and jumping down saves 1 frame.
  • On Cave-1, there is a different strat used with the hooks in the waterfall section. This allows Ronald to constantly be moving. It is not RTA viable.
  • On the first boss fight, animation cancelling is done to save 8 frames.
  • On the second boss, animation cancelling saves a few frames.
  • On the Sea autoscrollers, lag reduction strats were found, these are used in WR, but weren't previously.
  • The third boss is animation cancelled extensively and performs the quick kill. If WR would've done this it would be sub 20:20.
  • In the third room, using a slope on the bottom of the screen saves at least 5 frames.
  • The final miniboss uses a lot of animation cancelling and saves a lot of time, I didn't bother timing it compared to WR strats, as optimizing it took multiple hours.
Before making this TAS WR was 20:26 by DeathJohnson. After discovering the fast strat on the second Ship stage, WR was tied, and then beaten three times by Boon_7. Before I could even get this up, Boon has managed to get the WR to be a 20:19.
The RTA timing of this run is a 19:57

Noxxa: Judging.
Noxxa: The technical quality of this run looks good. It looks better than the published Expert difficulty run. And as a comparison TAS for Beginner real-time runs, it certainly does its job.
However, the difficulty setting chosen is an issue here. Choosing Beginner difficulty rather than Expert difficulty skips certain stages/sections, and this affects more than just autoscrollers. Skipping out on stage content is generally a bad reason to pick a lower difficulty, and in this case I don't see it being justifiable enough to obsolete the Expert difficulty run (especially considering the mixed feedback), nor to publish it alongside that run (as the content overlaps too much). Without a viable way to publish the run, I'll have to reject it.
I do look forward to a new Expert difficulty TAS with the same level of optimization as this run.
Last Edited by adelikat on 10/23/2023 2:08 AM
Page History Latest diff List referrers