2015-03-17

Secret of Mana / Seiken Densetsu 2 / Legend of the Holy Sword 2 in the works (originally posted on uCoz)

So I've finally gotten around to putting together full text dumps from the game, with everything decoded except control codes (movement, sound effects, music changes, and so on are all lumped in with the text, instead of the text being in its own block and called by reference like it is in most of the other games I've dealt with). Aided by a reasonably functional editor, and a gamefaqs message thread with a remarkable amount of technical data as reference, and BizHawk's RAM Watch functionality to test some things, the section is shaping up rather nicely.

It's not ready to be posted yet, but for now, here's a list of some findings I thought were interesting and possibly surprising.

General technical data:
  • Hit rate and evade rate are, by most standards, weird. Though misses are of course possible, especially against evasive enemies, the main thing high hit or evade does is to minimize both the chance and the severity of penalties to effective attack or defense power.
  • There are also hit and evade rates for magic. Though magic can't miss outright, both offense and defense can suffer from penalties, much like physical attacks.
  • Random variation is added to both attack and defense, but only when not penalized for failing a hit or evade check.
  • Mana energy (number of seeds activated) does more than just increase maximum spell level. The value squared is also added to the effectiveness of attacks and spells, except when they get the penalty for failing their hit check.
  • Critical hits appear to ignore defense in addition to doubling attack power. A hit that is not critical cannot inflict more than (999 - target's effective defense) damage. This is why attacks against the final boss most typically cause a bit under 550 damage (it has a game-highest 450 defense).
Armor data:
  • All headgear gives +23 evasion and +23 magic evade. All body armor gives +37 evasion and +15 magic evade. All accessories give +15 evasion and +37 magic evade. Get a full set of equipment as soon as you can!
  • Some armors give stat bonuses, raising Strength, Agility, Vitality, or Intelligence (or a combination of them) by 5. This is documented on any number of websites, but other than checking your status screen before and after each change, the game itself doesn't seem to give any indication of this. There's also a flag set for some of them that looks like it ought to raise Spirit, but it has no effect.
  • There are some other bits in the stat bonuses byte that have the same value for all armor/headgear/accessories, and might be meant to change the amount of the bonus.
  • Some armors prevent certain status effects. Again, this is documented here and there, but the game itself gives no indication of it. One useful one is that the Ochu Ring (Ivy Amulet), sold in the town by a mountain full of flowers that like to spam Sleep Flower, prevents characters from being put to sleep.
  • The armor data has two bytes that are set to 0x00 for everything. I expect one was meant to give an elemental affinity. The other I'm less sure of, but bonus defense against specific monster groups seems like a good possibility.
  • The Ruby Armet, a 46 defense / 46 magic defense helmet for The Girl, appears to be unobtainable.
Weapon data:
  • Most level 9 weapons have no special traits, not counting slightly higher attack power than the level 8 version. Other than the Sword of Mana (which is deliberately overpowered and only supposed to be obtained by special means in the very final battle), the only exception is the Fuuma Shuriken (Ninja's Trump), which keeps the Shuriken's critical rate bonus. However, that's not generally such a big deal, since it turns out that most of the special bonuses weapons get are largely useless anyway. More specifically...
  • Racial bonuses are glitched and do not work. All those "effective against"s do nothing.
  • Every weapon that has +2 to a stat in its description actually gives only +1. This is generally too little to make a meaningful difference.
  • However, critical rate bonuses and, of course, status effects work just fine, so there's that.
  • Speaking of critical hit rate, Bows, Boomerangs, and Javelins have half the critical hit rate of the other weapons.
  • The Diamond Flail (Nimbus Chain), Masamune, and Deathstrike Bow (Doom Bow) give a critical hit rate bonus of 10 per 256 hits. The Shuriken and Fuuma Shuriken (Ninja's Trump) give a bonus of 5 per 256 hits. (these numbers account for the halved rate for weapon type where applicable)
  • In addition to letting you charge attacks higher, gaining weapon skill levels also provides a small bonus to critical hit rate (3 per 256 hits for most weapons, or 3 per 512 hits for the halved ones).
  • Each character also has an innate bonus to critical hit rate, which is added after all the weapon modifiers (meaning it's not halved regardless of weapon). The Girl's bonus (10 per 256 hits) is twice as large as The Boy's and The Sprite's (5 per 256 hits).
Magic data:
  • At equal levels, Freeze is as powerful as Earthquake (Earth Slide), Thunderbolt, and Explode(r). So are Dark Force and Saint (Lucent) Beam.
  • Burst has the most raw power of any spell, but is hindered by being gained so late in the game and being unable to take advantage of a weakness. Conversely, over a dozen enemies, including several bosses, resist it.
  • Lava Wave is weaker than Fireball. Its only discernible advantage is that it's easier to spam.
  • All Saber spells give a damage bonus based on spell level and number of targets. They also change the weapon's element, but due to a glitch, this normally has no effect. As a result, Thunder Saber and Holy (Light) Saber are only good for the attack bonus.
  • Refresh (Remedy) doesn't remove status effects directly, instead reducing their remaining duration. Usually, though, the spell is strong enough, and durations short enough, that it works out to the same thing.
  • Antimagic (Dispel Magic) works similarly for most effects. When used against Countermagic (Wall), it instead removes a number of remaining uses, depending on spell level.
  • Evil Gate works sort of like a Final Fantasy-style Gravity spell except that it averages the target's current and maximum hit points instead of going off just the current value. Other spells normally kill faster, though, and don't cost as much to cast. Evil Gate does have the advantage of ignoring magic defense, at least.
  • Twinkle (Lucid) Barrier functions similarly to the Golem summon in some Final Fantasy games, taking a set amount of damage in place of the target. The amount increases with level, but is divided by the number of targets, and also capped at the target's max HP. Furthermore, the shield takes damage based on its own defense, which is always less than the defense of the character it's protecting, and lower still when the spell is multitargeted.
  • Spell animations use new data for every two levels, though for many spells some, or even all, of the data just duplicates the previous levels. The interesting part is that not only do the definitions go up to level 8 (and 9), but each spell also has an entry for level 10 and 11. The data for levels 10 and 11 is identical to the data for levels 8 and 9 in all cases, but that fact that it exists at all is intriguing.
  • On that note, most spells continue to work normally and increase in strength if spell level is artificially set above level 8. The Saber spell attack bonus and Evil Gate's damage divide by zero when single-targeted at level 10, giving no bonus for Sabers and 999 damage for Evil Gate. At level 11, Sabers continue to provide no bonus, while Evil Gate seems to have no effect at all.
  • Speaking of spell levels... many players know that spells raised to level 8 plus some additional experience have a chance of having a "super" effect that (usually) has a special animation and also stops time while the animation is playing. What's not so obvious is that this is the only time the spell acts as level 8. When a level 8 spell does not use the "super" effect, the game treats it as a level 7 spell.
Monster data:
  • Most monsters have no defense or magic defense. Those with no defense also have no evasion, and those with no magic defense also have no magic evade. However, every monster with any defense also has 90+ evasion, and every monster with any magic defense also has 99 magic evade. This is presumably due to the odd side effects of low evade mentioned above under general technical data.
  • Nearly every monster has a category (race/type/whatever), but most don't matter due to weapon bonuses against them not working. However undead/demon monsters reverse HP draining effects used against them, and incorporeal monsters reverse MP draining effects used against them.
  • Monsters with the Moon or Wood element seem to resist their own element (which is normal), but unlike the other elements, have no special vulnerability to the opposite. As noted above, this hinders the usefulness of Burst, since it can never take advantage of a weakness.
Status effects:
  • Most status effects have an associated timer and will wear off when the timer expires. Being shrunken is one of the few exceptions.
  • For whatever reason, shrinking someone with the Munchkin Hammer (Midge Mallet) also sets the timer for Moogle to its maximum value (which equates to 85 seconds).
  • Many status effects will remove Saber spells as a side effect. This is reportedly due to them force-equipping special internal "weapons" to trigger their graphical effects.
  • There's an unused status effect, からみつき, which translates to "Tangled" and is called "TANGLED UP" in the list of status effects in the English version. Based on what appears to be the message for when it's inflicted, this status would have prevented its victim from moving for the duration. This is distinct from the status that the Speed Down spell and several weapons cause (that's おもり, which translates to "heavy" or "weighed down" and is called "SLOWED DOWN" in the English version).
Story and events:
  • It's probably common knowledge that Gemma (Jema) will revive The Boy as needed if his fight with the Mantis Ant goes badly. It turns out that it's also not possible to get a Game Over before the fight. If you manage to die before then, The Boy regains consciousness where he found the sword, and continues from there.
  • Weapon Orbs aren't a thing in the Japanese text. The weapons gain Weapon Power instead. This does make it hard to explain how NPCs can give it to you, though.
  • Goblins won't ambush The Boy if he never talked to the soldiers outside the Water Temple who are preparing to go after the witch. It's unclear whether this is deliberate or just the result of reusing event flags without being careful enough about it.
  • Avoiding the Goblin ambush means The Boy doesn't meet The Girl there, so she won't immediately join him if they encounter each other in Pandora, which also means he can go talk to her father and suitor in the meeting room. If he then goes to her house, things get a bit weird, with The Girl vanishing when talked to instead of walking out (apparently the event scripts were never completed), and her father trying to say her name even though it was never set.
  • If The Girl is not in the party when The Boy heads to the witch's forest, he finds her under attack by Werewolves. This can lead to a rarely-seen event. Finish the events in Dwarf Village so that The Sprite joins the party. Then, make sure The Sprite finishes off the Werewolves when The Boy is dead. Having The Sprite talk to The Girl always gets a special reaction, but since only The Boy talking to her can advance the story, if he's dead she'll also revive and heal the party, saying something about an item she bought from Nikita (Neko). Unfortunately, this is the only time she'll provide any healing at all in this event.
Other than playing around with spells, I've only gotten up to saving Undine so far, so who knows what else might lay in store!

If you have any requests for NPC dialog or other text you'd like to see that I might be inclined to skip otherwise, or anything to try doing along the way that might have interesting results, feel free to share in the comments.

No comments:

Post a Comment