Power Rangers Dino Thunder is the third and penultimate game Natsume made for the franchise on the Game Boy Advance. It seemed they had set out to make a bigger adventure than their previous titles, shifting from arcade-style beat-em-ups to a mission-based campaign, complete with various cutscenes (that we skip through whenever possible).
It's also incredibly basic in terms of platforming and combat (and somehow, they managed to screw that up), so speedrunning this game is all about routing and exploiting boss AI.
The game is played on Easy, because not only do bosses have less health, but more importantly, the number of objects needed for each level is reduced the easier the difficulty and finding them is a pretty big contributor to length.
You have a classic 3-hit combo, a jump kick, two heavy attacks (up, which lunges you forward, and down, which keeps you in place), and a screen-nuke, which costs 10 "dino gems" to use. This (and amazingly enough, the basic attack, is SO time consuming to use that it is completely worthless in a speedrun, especially in a TAS.) While I mess around with the basic combo at some point, you'll really only see jump kicks and heavy attacks used practically in this TAS.
Most bosses are generally designed to punish players that mindlessly charge right at them. However, many can thus be baited into attacking, and then you just nudge out of the way and hit them with your weapon. However, some bosses have specific weaknesses that must be utilized to actually hit them, since they otherwise just block. For example, Mesogog (the final boss) has a first phase where you must hit the energy ball back to hurt him. Also, his final phase was impossible to figure out the actual developer-intended strategy for, especially once he gets into his rage mode (at half health). I mainly just baited him like normally, then preformed a set of jump kicks in which requires VERY precise inputs to actually hit him. With just a frame off, he'll either teleport or land the hit on you.
Throughout the game, you can guide the black ranger to a zord to unlock it. These puzzles are somewhat time consuming, but I had to do them anyways. Not only are some of the unlockable zords good, but for every failed minigame, you're set up for an extra megazord fight against the White Ranger. This is the main reason why I went and unlocked some of these (that last zord is basically unusable.)
Megazord fights mainly consist of dashing forward the required distance to connect attacks, and alternating betweeen A and B in a sort of rhythm. Before the fight starts, you can choose two zords to use as arms. Some of the differences between them include range, speed, strength, and the number of inputs required for a successful combo might even depend on that. These combos, when complete, executes a "finisher" attack which deals about double the combo's damage, so you can see why that's so important to do. The problem is that the boss can just c-c-c-combo break you just by blocking, and some zords have such an insane wind-up that it's pathetically easy for bosses to do just that. As a result, faster and longer zords are preferable, although that cyan sawblade one had become the default once unlocked, due to only needing 4 inputs to combo.
CasualPokePlayer: Going through the first level of this movie, there immediately appear to be a few issues:
1. Jumping and attacking slow the player down (this is somewhat appearent by the judder they have, and quick testing or a speed display would show this is the case). Various points show jumping and attacking are done without regard to this, leading to some timeloss.
2. There appears to be some movement tech with ladders not abused fully in this movie: attempting to climb on a ladder will "pull" the player to the center of the ladder, which in turn is effectively a minor dash forward (normally walking is 2 units per frame, the pull at maximum range is 18 units in a frame). This can be used as a speedup even if you dont intend to use the ladder (start climbing it then get off it immediately, that is faster than not interacting with the ladder), it can also mean taking slightly different paths in order to gain these ladder boosts.
3. Throughout the first level a jump kick attack then a Down+A then a Up+A attack is used to take out some boxed enemy, but this ends up being slower by a frame compared to a jump kick attack then 2 Down+A attacks if you intend to go the other direction after the attack (for most cases, this applies).
4. There is 1 case in the first level where the attack on the boxed enemy is done on a floor where you are able to drop down to the previous floor. In this case, it's much faster to do multiple jump kick attacks, as you can jump down to the lower floor in order to do a jump kick attack quickly, compared to the jump kick attack then Down+A/Up+A attacks.
Using these, I was able to beat your first level by
59 frames, and presumably many more frames could be shaved off in other levels.
Now, the last optimization I don't really expect someone to find (that one was just me trying a lot of different things in that fight and stumbling upon that rather large optimization), although the first 2 optimizations at least were very obvious finds by simply observing a
speed display lua script, and cover over half of the saved frames, and the third one would have just been obvious if all possible combos were tried. Considering this, I don't think this movie is acceptable as is.
This movie is a very good attempt, and I would suggest trying again using my findings and looking deeper into this game. Simple mechanics are often deceptive with their simplicity, like my findings on movement and ladders show. There's possibly other things in parts of the game that might be obvious with just a little bit of
investigation.
Rejecting, better luck next time!
