SDA logo

Fresh off the smash success of Tactics Ogre for the Super Famicom, top developers at Quest were hired by Square to create a similar game with a Final Fantasy theme. In January 1998, the result of this project was released as Final Fantasy Tactics for the PlayStation. The game introduced most westerners to tactical RPGs and remains a fan favorite to this day. Ramza Beoulve, a young nobleman, is caught up in the machinations of civil war in the kingdom of Ivalice. While others around him attempt to manipulate the situation for their own benefit, Ramza's motivation is simply to rescue his dear sister Alma, who has been captured by the Church for mysterious reasons.

 

Best time: 3:33:15 by John 'chessjerk' Kearsley IV on 2014-12-14, done in 30 segments.

Get Flash to see this player.

Author's comments:

Final Fantasy Tactics 3:33:15 Segmented Speed Run (NA version, any%, using Math Skill)

Run finished: December 14, 2014

Author's comments:

Welcome to the latest update for my Final Fantasy Tactics segmented speed run! This is my fourth segmented speed run of FFT using Math Skill, and this run saves about 44 minutes over the previous version of this run (and an additional 5 minutes due to a new timing method, see below). The following comments detail the many significant changes for this run compared to the previous run.

Due to changing timing standards on SDA, this run was timed differently than my previous segmented runs, which used in-game time as of the load of the final segment plus real time up through Altima's death. For this run, every segment was timed manually from gain of control at the beginning of the segment to the beginning of saving at the end of the segment. If this run had been timed using the older method, the time would have been 3:38:25. The "time saved by Chapter" indicated below is based on comparing the times from this run using the older method to the previous 4:22:16 run.

Changes for this run are:

Running FFT on a PS2 with Fast Disk Speed set
Ramza's Zodiac sign (Capricorn)
Use of generics (they are used for only two battles)
Use of Orlandu (he makes many Chapter 4 battles faster)
New battle strategies (detailed below)

Chapter 1: Segments 1 through 10 (total time saved in Chapter 1: 16:42)

Segment 1: Orbonne Monastery

As usual, the opening battle is mostly uncontrollable, so my goal is to do the best I can and hope the AI doesn't waste time by healing, casting spells, missing attacks, or wandering around doing nothing.

Segment 2: Gariland Magic City

A different set of qualities are needed for this run than were needed in previous runs. The basic qualities I am looking for include:

Ramza needs high starting Squire and Chemist JP
Either Chemist needs high starting Chemist JP and Faith

One of the most important considerations at this point is starting Job Points. After the initial Orbonne battle concludes, the party is set with a random amount of JP between 100 and 199 in all of the jobs in the game. Therefore, even if I get a decent Orbonne battle, if Ramza's or either Chemist's stats aren't good enough, I need to start over with another Orbonne fight.

This battle, like the Orbonne battle, is largely random, since I am unable to change anyone's jobs or equipment before the battle. Ramza needs to take a few actions in battle to get to 200 Squire JP (in order to learn Gained JP UP) and the chemists acting will get him to 200 Chemist JP via spillover JP (to unlock the Priest and Wizard classes). The male Chemist must also get to 200 total Chemist JP to unlock the Wizard class.

This segment also demonstrates the "JP Scroll Glitch." Quickly recapping the procedure for the glitch:

1. On the "Learn Ability" screen, select a skill that can be learned.
2. Before confirmation, hold square and press down; the skill list will shift downward to a skill that cannot be learned.
3. Select "Yes" to confirm. In some cases, all abilities are instantly learnable, otherwise go to step 4.
4. Cancel back to the job list screen; it should say that 0000 JP are available.
5. Cancel back to the unit screen and re-enter the job list screen, it should now say that 9999 JP are available.

Naturally, this glitch cannot be used on classes where the skill list is not longer than one page. This means that, unfortunately, Calculator cannot be glitched.

Segment 3: Mandalia Plains

The next six segments are Ramza's grinding segments. In Segment 3, Ramza needs 500 Squire JP for Accumulate and Yell (two important self-targeting abilities for grinding). At the beginning of the battle, the option to not save Algus is chosen. The main reason why is because he will be killed off to avoid watching him waste time. A minor secondary reason is choosing this option gives Ramza's team a small Brave boost.

Killing Delita on the first turn is also very important, since he also just wastes time if he is alive. A thrown stone from an enemy squire and an attack from my squire is enough to kill him, and this also brings the enemy squire to a spot where my Wizard can kill him as well. Algus's second turn kills another enemy and causes the rest of the enemies to line up to be killed by the Wizard's second spell.

This leaves just the enemy Red Panther, who is the slowest of the enemy units, and as such, will get the fewest turns during the grinding. Using Algus's dead body as a blockade, Ramza is free to alternately attack and heal my Wizard until the required JP is earned. I am fortunate that Ramza and the Wizard have bad Zodiac compatibility, since that allows Ramza two attacks for every Potion needed to heal, thus allowing me to skip buying Potions before the battle.

Segments 4 through 8: Mandalia Plains random battles

In the rest of the grinding segments, I am looking for a particular random battle, namely one with a solo Goblin. Ramza continually Yells and Accumulates for experience and JP to get to Level 4 Wizard, Level 4 Priest, Level 3 Time Mage, and Level 3 Oracle in turn. This unlocks the Calculator class. Finally, the Calculator grind is less about JP and more about experience, since a particular level target must be hit to make the rest of the run faster.

I need this particular random battle since Ramza can win it by himself (even as a very slow Calculator) and it saves time by minimizing the number of enemy actions. However, the odds of getting this particular battle are quite low. Random battles have a certain number of slots where there can be up to three types of potential enemies, but these slots can also have "Nothing" in them. A random battle entering Mandalia from the east can have up to six enemies. The first slot is a guaranteed Goblin, the other five slots can have an assortment of Goblins, Red Panthers, and Yellow Chocobos. Three of the enemy locations have "Nothing" as the third slot, and the other two have "Nothing" in both the second and the third slot. The odds for each slot are: 50% chance for the first slot and 25% each for the second and third slots. Therefore, the odds of getting a solo Goblin and no other enemies is (1/4)*(1/4)*(1/4)*(1/2)*(1/2) = 1/256. I was able to get this battle for 4 of the 5 random battles (after many many tries, I settled on a two Goblin battle for one battle, which is still winnable but about 1 minute slower).

In addition, there is a place where the Goblin can be tricked into just waiting and doing nothing on its turns, allowing Ramza to do whatever he wants. In general, normal physical attacks have a vertical tolerance of 3 when attacking downwards, but only 2 when attacking upwards. So, if Ramza stands on a rock with a height of 4 and the Goblin is below him on a panel with a height of 1, Ramza can hit the Goblin, but the Goblin cannot hit Ramza. Also, if the Goblin cannot reach a panel to hit Ramza in one turn, it will just stand there doing nothing. So, even though Ramza can be hit on one side, since the Goblin cannot reach it in one turn, it won't try.

Now Ramza knows all the abilities he will need for the rest of the run. Here is a list of what is learned and why:

Action abilities:
1. Squire - Accumulate: This increases Ramza's Physical Attack by 1 and has a very fast animation.
2. Squire - Yell: This increases Ramza's Speed by 1, but has a much slower animation than Accumulate.
3. Priest - Holy: This is the most powerful calculable attack spell, it cannot be evaded, and it can be strengthened by equipment that increases damage done with the Holy element (i.e. Excalibur). However, the animation takes significantly longer than the other attack spells at my disposal. It is only used when Flare or Bolt 3 don't hit hard enough or when I need to bypass significant enemy magic evasion.
4. Wizard - Bolt 3: This spell has one of the fastest spell animations in the game. It is used early in the game when story battles enemies are still weak as well as later in the game against low HP enemies.
5. Wizard - Death: This is only used in one battle: the Zirekile Falls battle. At Zirekile, Bolt 3 cannot be used since Ovelia has good Zodiac compatibility with Ramza and will die from the spell. However, she is immune to instant death, so Death won't hit her while it can hit all of the enemy Knights.
6. Wizard - Flare: This is the main damage dealer after Ramza changes to a Priest in Chapter 2. It is the second-most powerful calculable spell (behind Holy) and it has a very short animation, but it can be evaded and it cannot be strengthened except by raising Magic Attack through equipment or by equipping Magic Attack UP.
7. Time Mage - Quick: This is only used once at the Bethla Sluice battle in Chapter 4, and it is used to instantly gain a turn, hit both switches, and open the floodgates before the enemies can do anything important.
8. Time Mage - Demi 2: This deals damage equal to 50% of the target's maximum HP. It is used mainly against Zodiac bosses who cannot be killed by one spell, since two uses of Demi 2 kill anything in the game (except the final boss, who has over 3000 HP and Demi 2 damage caps at 999). However, it doesn't have a 100% hit rate.
9. Calculator - All Math Skill abilities (except Exp): The parameters used for Math Skill include: CT, Level, and Height. The factors used are: Prime Numbers, 5, 4, and 3. The parameter Exp isn't learned because while enemy levels are generally fixed, experience within those levels is not, so it is not a dependable factor to use. The general strategy for using Math Skill in battle depends on the objective of a particular battle (detailed below).

Reaction abilities (these activate a percentage of the time equal to Ramza's Brave):
1. Wizard - Counter Magic: This is used at Bed Desert to return a spell gun shot (which counts as casting a spell) back to Balk so only one Holy is needed to finish him off.
2. Time Mage - MP Switch: This converts any HP damage to MP damage. It is useful when Ramza must target himself with attack spells (ensuring that he doesn't kill himself).

Support abilities:
1. Squire - Gained JP UP: This increases JP gain by 50%. It is useful to speed up the Chapter 1 JP grinding.
2. Wizard - Magic Attack UP: In general, this increases magic damage by 33%. It also slightly increases the hit rate of Ramza's percentage-based spells (Quick and Demi 2).
3. Calculator - Gained Exp UP: This doubles experience gained on every successful action. It allows Ramza to grind to a lower level in Chapter 1 and still get to Level 23 by Golgorand (see below).

Movement ability:
1. Time Mage - Teleport: This increases Ramza's potential movement range while ignoring terrain and enemy units, but with an increasing risk of failure the farther away Ramza tries to move. It is only useful at Bethla Sluice for the same reason as Quick.

Here is a list of general battle strategies with Math Skill:

In boss battles (or "Defeat X!" battles): Find a Math Skill combination that hits only the specified enemy to end the battle. If no combination exists, find the combination that hits the least number of other units.

In escort battles (or "Save/Protect X!" battles): Find a combination that hits all relevant enemies and not the unit to be saved/protected. If no combination exists, some specific battle strategies may be needed to quickly end the battle (see the Zirekile battle and the Barius Valley battle in Chapter 2).

In normal battles (or "Defeat all enemies!" battles): An explanation of how a battle progresses is helpful to explain what happens here. The unit of time in battle is called a "clocktick". For every clocktick that passes, all units gain CT equal to their speed. When a unit's CT reaches 100, it is their turn to act. Knowing that CT can be used as a parameter for Math Skill, I can manipulate CT in order to hit everyone on the battlefield at once.

There are two possibilities here. In the first scenario, I try to gain a speed advantage over all of the enemies and adjust Ramza's effective speed to 12, 9, or 7. With these speed values, Ramza acts first on the 9th, 12th, or 15th clocktick, respectively. Since all of these clockticks are divisible by 3, every unit's CT will be divisible by 3, and can thus be hit with the combination CT 3. Also, when a unit reaches 100 CT, that unit is treated as having exactly 100 CT for the purposes of Math Skill (instead of that unit's actual CT), so Ramza will not hit himself in this scenario, but all enemies will get hit with a spell and die.

If Ramza cannot gain a speed advantage over the enemies, then the second scenario comes into play. Ramza goes with a 7 speed setup (acting on the 15th clocktick) and uses CT 5, which is guaranteed to hit everyone on the field (including Ramza). Hopefully, MP Switch or magic evasion (if any) prevents Ramza from killing himself. Luckily, this strategy is not needed as much compared to previous versions of this run due to the leveling strategy in Chapter 1.

Over the course of the game, after Ramza attains the use of Math Skill, most battles will be over in one action (sometimes two for Zodiac boss battles). Therefore, I am able to predict what Ramza's experience level will be throughout the run. The key level in this run is level 23, as this is the level when Ramza as a Priest gains a speed point. Ramza MUST be level 23 before reaching the Golgorand fight in Chapter 2, as this is the first battle with an enemy with 8 Speed. However, Priest Ramza with a Green Beret has 9 Speed, allowing for Ramza to take advantage of the 9 Speed strategy listed above. Also, if I plan to use "Level" as a parameter in a particular battle, I try to have Ramza's level not equal to a level where he is targeted (although this isn't possible in all scenarios). Therefore, Ramza's level after I finish grinding Calculator JP must be at least level 20, meaning that having too high of JP in various classes just means that more grinding is necessary as a Calculator.

The only battle where defeating enemies is optional is the Bethla Sluice battle in Chapter 4. The objective here is to have Ramza stop on two specific panels during the course of battle. Here, Math Skill isn't needed. Ramza teleports to the first square, casts Quick on himself, then teleports to the second square to end the battle.

Segments 9 and 10: Sweegy through Zeakden (The rest of Chapter 1)

The rest of Chapter 1 is quite easy compared to the grinding segments. These battles show the power of Math Skill through CT 3 for enemy mobs and specific calculations for boss battles.

Sweegy: CT 3 Flare. This battle is actually much tougher than normal since the monsters share Ramza's level. Flare is needed over Bolt 3 because of the (relatively) high enemy levels.

Dorter: CT 3 Bolt 3. Very rarely, the Knight will block Bolt 3 with his shield at a 3% chance.

Sand Rat: CT 3 Bolt 3. No magic evasion here, just need decent Faith and no bad Zodiac compatibility.

I buy a Thunder Rod for Ramza here, which increases Bolt 3 damage.

Thieves Fort: Height 3 Bolt 3. This is the worst "Defeat X!" battle, since so many of the enemies share factors with Miluda. Height 3 is the best option, despite hitting two other enemies.

Lenalia: Level 4 Bolt 3. This also hits Ramza since he is still level 20, luckily MP Switch saves the day. Also, Miluda has 10% magic evasion from her Small Mantle.

Windmill: Height 5 Bolt 3. Wiegraf also has a Small Mantle for 10% magic evasion.

Zeakden: Height 4 Bolt 3. Algus has no hope to survive. Too late I found out that Level 5 Bolt 3 is very slightly quicker due to one less cursor movement.

Chapter 2: Segments 11 through 16 (total time saved in Chapter 2: 8:14)

This chapter has some key differences from the previous speed run. These include:

Ramza using Bolt 3 instead of Flare for most of the Chapter
Delita not knowing Crush Punch at Zirekile Falls (this causes him to kill Gafgarion)
Ramza using Death to kill everyone at Zirekile Falls
Ramza getting to Level 23 by Golgorand (allows a first turn kill at Golgorand and Lionel)
Agrias still blocks a fatal spell at Barius Valley with her Mythril Shield at a 5% chance, this time Bolt 3 instead of Flare

Segment 11: Dorter and Araguay

Gafgarion is turned into a Chemist to remove his equipment and reduce his speed to 6, since otherwise he would get turns and waste time in the next three battles.

Dorter: CT 5 Bolt 3. The 7 Speed Thieves force the use of CT 5 and hope for MP Switch.

Araguay: CT 3 Bolt 3. Sorry Boco, you will only slow us down. :(

Segment 12: Zirekile Falls

This is a tough battle to win quickly, since Ramza has good Zodiac compatibility with Ovelia and she must survive the battle. She also has no magic evasion and shares the same speed as all of the enemy Knights. As previously mentioned, she is immune to instant death, so CT 3 Death is the way to go.

However, it is not as simple as that. Traitor Gafgarion is also immune to instant death, so I need to find a way to take him out as well. Delita is back for this battle and he goes first, so I need him to take out Gafgarion so CT 3 Death can finish the battle. Delita's first action depends on if he knows Crush Punch or not. If he does, he will move to the right and try to kill a Knight with it. If he doesn't, he will move left and target Gafgarion with either Stasis Sword or Lightning Stab.

To make things worse, Death has a mediocre (at best) hit rate, and five Knights need to be hit with it. I figure that the odds of Death hitting all the Knights is about 1% on average, and Delita not knowing Crush Punch is about 30%. Now you know why this battle is its own segment!

Ovelia also has great equipment for Ramza to use at this point in the game (equipping Triangle Hat and Wizard Robe adds 3 to Ramza's Magic Attack)

Segment 13: Zaland to Lionel

Zaland: CT 3 Bolt 3. Enemy Knights are getting tougher to kill with Bolt 3.

Mustadio joins the party as a Guest. I change him to a Chemist to reduce his speed to 6.

Barius Hill: CT 3 Bolt 3. Same here.

Segment 14: Zigolas to Warjilis

Zigolas: CT 3 Bolt 3. The non-undead enemy can be a Morbol, a Flotiball, or a Uribo. The Uribo is very fast, so I need it to be one of the other two.

Goug: CT 5 Bolt 3. 7 Speed Thieves again cause the need for CT 5. In addition, they can waste a lot of time if their reaction abilities activate (or if Mustadio has Counter Tackle).

Warjilis: A Green Beret is purchased for a nice +1 bonus to Speed. I don't need it quite yet, however...

Segment 15: Barius Valley

This is another very low percentage battle and deserves its own segment. Similar to the Zirekile Falls battle, there are no Math Skill combinations that hit all of the enemies and not Agrias. Even worse, Ramza has best Zodiac compatibility to her. Fortunately, she does have magic evasion (5% from her Mythril Shield), so I need her to block a very fatal Bolt 3 and hope the enemies all die.

Ramza gets to level 23 right on schedule, so the switch to Priest with the Green Beret is made. Magic Attack UP is also (finally!) equipped to increase magic damage.

Segment 16: Golgorand to Lionel (End of Chapter 2)

Golgorand: CT 3 Flare. 9 Speed Ramza outspeeds 8 Speed Gafgarion, so he nukes the whole field with Flare. Bolt 3 isn't used because Ramza can't equip the Thunder Rod as a Priest, so his Bolt 3 damage is lacking.

Lionel (outside): CT 3 Flare. More of the same.

Lionel (inside): Level 5 Flare/Level 5 Bolt 3. Queklain is the first Zodiac beast Ramza faces. He has 9 Speed and around 380 HP, so he gets the first turn, and Ramza can't kill him with one spell. His first action is always Nightmare, which has a 50/50 chance of inflicting either Sleep or Death Sentence. Sleep is very bad, but Death Sentence, not so much. Also, Ramza gets two turns in a row because Queklain moved and acted on his first turn, while Ramza acts and stays still. This conserves CT and allows for an extra turn.

After the long cutscene, Chapter 3 begins. New equipment is available to boost Ramza's Magic Attack (Wizard Staff and Magic Gauntlet for an additional +3 to Magic Attack).

Chapter 3: Segments 17 to 20 (total time saved in Chapter 3: 7:33)

Chapter 3 has some small optimizations as well over the previous run. 9 Speed makes the Save Olan battle much easier, as well as allowing first turn kills for the first half of the Chapter. Also, Flare is used instead of Demi 2 against Velius, since Flare kills a demon and saves a spell animation. Also, I bought a Reflect Ring at Yardow, but it turned out to be unneeded (see below).

Segment 17: Goland to Lesalia

Goland: CT 3 Flare. Olan has 9 Speed, so he goes first. In a perfect world, he would attack the enemy Mediator, get a critical hit, and knock him off the roof for a kill. The odds of that happening are quite small for a very small time savings, so I didn't worry about that. Olan getting his turn first also means his CT is not divisible by 3, so CT 3 doesn't target him.

Lesalia: Level 4 Flare. No combinations hit only Zalmo, Level 4 hits an enemy Knight as well.

Segment 18: Orbonne (Underground Book Storage)

2nd Floor: CT 3 Flare. The enemy Lancers are tough to kill here since they can have high HP and Aegis Shields (which have 50% magic evasion).

3rd Floor: Height 3 Flare. Some manipulation is needed here. With the 9 Speed setup, there are no combinations to only hit Izlude. However, changing Ramza's helmet allows Izlude to go first and move. Also, changing Ramza's starting position affects where Izlude moves on his first turn. Starting in this position draws Izlude to a 9 height square, and Height 3 doesn't hit anyone else. Izlude doesn't die from the spell, but like some other "Defeat X!" battles, you don't actually need to kill them, you only need to get them into critical HP.

1st Floor: Height 5 Flare. Wiegraf is back with more magic evasion (25% from his Elf Mantle).

Segment 19: Grog to Riovanes

Grog: CT 3 Flare. The front Squire has an Elf Mantle, the rest generally don't have magic evasion.

Yardow: Height Prime Flare. There is a very small chance of losing this battle before Ramza gets a turn. Rafa will be targeted by two throws before Ramza moves. The first Ninja will always have bad Zodiac compatibility to her, but the second Ninja may have best Zodiac compatibility, and if he throws a high strength weapon at her (i.e. a Partisan spear), and she doesn't dodge, she may die. After that, all the enemies are on height 2, so Height Prime does the trick.

Ramza is tired of being a Priest, so he changes to Wizard. New equipment available at Yardow includes a Wizard Rod (+2 Magic Attack) and Sprint Shoes (+1 Speed). Ramza as a Wizard with Green Beret and Sprint Shoes gives him 9 Speed, allowing a continuation of the 9 Speed setup in use since Golgorand.

I also buy a Reflect Ring and 108 Gems here. The Reflect Ring was intended to be used in the final battle, but it ended up being slower, so it never gets used. It does make an unwanted appearance in Chapter 4, though. 108 Gems is used in one battle in Chapter 4 to boost Holy damage.

Yuguo: CT 3 Flare. The ghosts will always die here, so the only worry is magic evasion and bad Zodiac on the undead Wizards and Time Mages.

Segment 20: Riovanes to End of Chapter 3

Riovanes (outside): CT 3 Flare. Rafa and Malak leave the battle when one of them (or both) are critical or dead, so Flare has to hit one of them. The Knights are still tough kills.

Riovanes (inside): Height 3 Flare (Wiegraf), Level Prime Flare/Level Prime Flare (Velius). Ah, the dreaded Wiegraf/Velius fight! CT manipulation is needed in the first part of the battle, since if Ramza uses Flare without moving, the game doesn't consider his turn finished until after Velius shows up. If Ramza moves and then uses Flare, his turn is considered finished. Also, Wiegraf transforms when gets to critical or dies. If he only goes to critical, Ramza's CT stays where it was. If he dies, Ramza's CT goes to 100 before the second battle starts.

Since Ramza starts the Velius battle with 100 CT, he can act, wait, and then immediately get a second turn. It's a good thing he does, since if Velius was able to get a turn, he would likely immediately use Seal on Ramza, ending the fight due to petrification.

Riovanes (rooftop): Height Prime Flare. Elmdor has 10 Speed, so he always goes first and he always uses Muramasa on Rafa. Ramza goes next and targets Lede with Flare. In this battle, when any of the enemies reach critical or dead status, they all teleport away.

Chapter 4: Segments 21 to 30 (total time saved in Chapter 4: 11:38)

This chapter has many new strategies:

Ramza using Bolt 3 at Finath
Ramza using Counter Magic at Bed Desert to win the battle solo
Recruiting Orlandu for more than just his Excalibur
Abusing the Item Duplication glitch
Orlandu killing Zalera and zombie Zalbag with critical Lightning Stabs to skip dialogue
Orlandu killing Kletian in one hit at Murond Death City (needs level 29 for an extra PA point)

Segment 21: Doguola to Zeltennia

Doguola: CT 3 Flare. Knights and Lancers are starting to have really high HP.

Bervenia: Level 5 Flare. The Ninja is fast, but too far away to have any impact. Meliadoul has 10% magic evasion from her Platina Shield. Too bad her infinite Reraise from Chantage can't save her when she teleports away after death.

Finath: CT 3 Bolt 3. The Chocobos here share Ramza's level, so Bolt 3 is sufficient to take them out. The enemy in the middle of the map can be a Uribo, and if it is high enough level, it will have 9 Speed and go first, causing a reset.

Zeltennia: Level 5 Flare. Delita's level is around Ramza's level, and in this attempt he is level 25, so there is a small time loss since he also gets targeted by Level 5 Flare. Zalmo has 25% magic evasion from his ever-present Elf Mantle.

Ramza equips 108 Gems and the Counter Magic reaction ability to make the next battle possible.

Segment 22: Bed Desert to North Wall

Bed Desert: Counter Magic Ice/CT Prime Holy. Balk has high HP, high Speed, and innate Magic Defend UP, so he can't be killed in one Holy. However, he can be killed on Ramza's first turn using the Counter Magic reaction ability. Shots from Balk's Blaze Gun count as casting spells that can be cast back at him through Counter Magic, and the additional damage allows one Holy to kill him.

North Wall: CT 3 Flare. Ramza re-equips his Sprint Shoes for the CT 3 strat. High HP enemies are still tough to kill with one spell.

Segment 23: Sluice

The Sluice gets its own segment due to the particular setup needed, the possibility of randomly losing the battle, and Orlandu (who is recruited after the battle) needs to be at least level 27 with Lightning Stab already learned, which doesn't happen very often.

Ramza changes back to Squire, equips Time Magic and Teleport, best fit's his equipment, and equips Red Shoes. Best fit normally puts Red Shoes on automatically, and it is very useful for this battle since they grant +1 Magic Attack and +1 Move. However, since I bought a Reflect Ring back in Chapter 3, best fit equips that instead. Since the segment is so difficult, I lose some time to manually equip the Red Shoes.

This setup gives Ramza 7 Speed (so all the enemies go first), 5 Move (so the first movement of 7 squares has a higher chance of success), and access to the Quick spell, which instantly gives Ramza a turn when it hits.

The goal is to have Ramza stand on two specific squares (the switches to open the Sluice), making this the only battle where killing enemies is completely optional. Teleport and Quick allow this to happen after the enemy finishes moving after their first turn. Also, it should be noted that if Ramza is not alone, the Knights that start the battle standing on the switches do not move, so it is essential that Ramza be alone (or the only unit alive) when the Knights get their turns.

After the battle, Orlandu is recruited. He wields the Excalibur, which grants auto-Haste and strengthens Holy attacks, and has a wide array of skills he can use. The most important ability is Lightning Stab, which has great range and infinite vertical tolerance, and will be useful to end some battles in Chapter 4 early to skip dialogue. He also has to be a minimum of level 27 when I recruit him. The reason won't be obvious until near the end of the game, but he gains a Physical Attack point at level 29, and that extra point will guarantee a kill on a future boss.

Segment 24: Germinas to Limberry

Sorry Aeris, I don't need your flower, despite having hundreds of thousands of gil to spend.

At Zarghidas, there are some new items to buy. Also, this is the first segmented run where I make use of the item duplication glitch. This allows me to get a second Excalibur for Ramza. To execute the glitch:

1. Go to "Fitting Room" in the shop menu
2. Fit Orlandu with a weapon in his left hand that is worse than the one to be duplicated
3. Select "Best Fit" to restore his equipment to the original setup
4. Cancel back to the shop menu and leave the shop

This leaves Orlandu with his Excalibur but also creates another one in the inventory that Ramza can equip. Due to the increasing speed of the enemies, allowing Ramza to get to an effective speed of 12 with the Excalibur is very important.

Germinas: CT 3 Flare. For some reason, two of the Archers in this battle always have over 80 Faith, making them easy kills. The rest, not so much.

Poeskas: CT 3 Flare. More of the same.

Limberry (outside): Lightning Stab. Level Prime can hit Lede, but the Apandas' levels are random, so they may be hit also. Instead, a single Lightning Stab takes her out much easier.

Limberry (inside): Level Prime Holy. Elmdor has a LOT of HP. Luckily, I don't need to kill him to end the battle.

Segment 25: Zarela

There is a lot of conversation in this battle, and in order to skip it, Zarela must die before guest Meliadoul gets her first turn. Zarela has over 1100 HP, so Demi 2 is needed from Ramza. However, Orlandu's Lightning Stab only deals 400 damage normally, so he needs to get a critical hit to get enough damage for the kill.

Enjoy the long cutscenes before the long walk back to Igros.

Segment 26: Adramelk

Shoutouts to Claude for figuring out how random encounters actually work. For the longest time, it was thought that random encounters were 50/50 every time a green dot is crossed. It turns out that it is actually 70/30 in favor of no random encounter. Also, when crossing multiple green dots at once, the odds of success tend to be higher than if each dot is crossed individually.

Here, five green dots in total must be crossed to get to Igros. Crossing them all at once is about a 40% chance. However, I take a quick stop at Gariland to get a Thief Hat for Ramza and a Flash Hat for Orlandu. It's unfortunate, though, since Claude found out that if you go straight to Igros, you can never get an encounter at Mandalia if you make it that far.

Dycedarg: CT 3 Holy/Lightning Stab. Dycedarg has amazing defenses with innate Defense UP and Magic Defend UP and an Aegis Shield to block 50% of magic that can be evaded. Luckily Holy cannot be evaded, and neither can Lightning Stab.

Adramelk: Height 4 Demi 2/Height 4 Demi 2. Being a Zodiac Beast, he dies very quickly to two doses of Demi 2.

Segment 27: Murond

Outside: Lightning Stab/CT 3 Flare. Orlandu takes out two Mediators easily with Lightning Stab and CT 3 Flare cleans up the rest. Flare is nowhere near guaranteed to succeed, though, since the Priest has an Elf Mantle and the Geomancers can have a shield, a mantle, and Abandon (a Ninja reaction ability that doubles all evasion).

Inside: Lightning Stab. Orlandu's time to shine! He equips the Twist Headband for maximum power. Kletian eats a Lightning Stab and they all run away in fear.

Segment 28: Zalbag

Chapel: Critical Lightning Stab. Zombie Zalbag is Ramza's toughest opponent, since he is the only enemy with worst Zodiac compatibility. However, he does have good Zodiac compatibility with Orlandu, and a critical Lightning Stab takes him out and skips a lot of dialogue in the process.

Segment 29: Orbonne (UBS 4) to Hashmalum

UBS 4th Floor: CT 3 Holy. We're entering the home stretch now! This battle is extremely difficult to win in one turn. Six high HP, high evasion enemies necessitate the use of Holy over Flare. I figure that just hitting all of the enemies through their magic evasion at once it about a 1% chance, and Flare's lower power would not guarantee many kills if it does hit.

Rofel: Level 3 Holy. Despite all of my experience planning, I couldn't avoid Ramza being level 30 here. Oh well. Holy does just enough damage to kill Rofel.

Kletian: Lightning Stab. Level 29 Orlandu has just enough power to OHKO Kletian with a single Lightning Stab. The enemy Ninjas can mess things up if they have Elemental secondary.

Balk: CT Prime Holy/Lightning Stab. There is a very subtle difference here from my previous runs. If Ramza acts before moving on his first turn, the damage done to Balk will trigger dialogue. If Ramza moves first and then acts, the dialogue waits for Balk's next turn, which thankfully never comes after Orlandu's killing Lightning Stab. Also, MP Switch finally becomes useful, as it saves Ramza from a deadly Bolt 3 from Balk's gun.

Hashmalum: Level Prime Demi 2/Level Prime Demi 2. Ramza no longer needs to maximize damage with Holy or Flare, so he goes for maximum speed (Thief Hat, Sprint Shoes) in order to get two turns before Hashmalum wastes him with Meteor. The second cast targets Ramza since he leveled up after his first action, but it couldn't be helped.

Segment 30: Altima

The final battle has two possible battle strategies that kill both Altima forms quickly. The first uses Math Skill (as usual). Two Demi 2's kill the first form before it can move, and three Demi 2's plus an Orlandu Lightning Stab kill the second form, although it allows Altima one turn to uselessly punch Ramza.

The second strategy uses the Reflect Ring that I bought way back in Chapter 3 and does not use Math Skill. With the proper setup, Reflect can allow spells with an Area of Effect to hit the same target twice in one cast. This allows for the first form to die to two Demi 2's cast on the first turn, and the second form to die to four Demi's cast before he gets a chance to move. There are, unfortunately, many drawbacks to this setup. Because this strategy was not known to me until I was already well into the run, I did not learn Demi and Short Charge way back in Chapter 1 when I was learning abilities since I did not think I needed them. Also, both forms of Altima have reaction abilities that activate when targeted by magic (Absorb Used MP and Face UP, respectively), but don't activate when targeted by Math Skill. Altima has four chances for a reaction to activate at a 70% chance each, wasting about 3 seconds each time it does. The combination of teaching additional skills and reactions going off causes this strategy to be very slightly slower than using Math Skill.

Possible improvements:

This run, despite accomplishing an excellent time, still can be improved:

1. There are many minor menu errors in the run that are individually insignificant, but probably cost a few minutes over the course of the run.
2. A new formation was discovered for the Gariland battle after this run was completed, and that formation could potentially save some time completing that battle.
3. I was unable to find a solo Goblin for one of the grinding segments in Chapter 1, costing about 1 minute.
4. More testing should have been done on the Reflect strategy for Altima. Buying the Reflect Ring too early caused a small time loss (1 second for buying it, 2 seconds before the Sluice, where Best Fit equipped the Reflect Ring instead of the preferable Red Shoes). If the Reflect strategy can be faster, it may be better to buy Reflect Mail instead of the Reflect Ring.
5. Saving costs time (about 10 seconds if done from the world map, about 5 seconds if done in between a series of battles), so combining segments would save time.
6. Manipulating the camera better in battle (zooming out saves time when the screen shifts to different units between turns, especially if the units are far apart from each other).

Huge shoutouts to Claude, who started doing RTA runs of FFT that eventually surpassed my previous segmented time of 4:22:16 and inspired me to do this (final?) update to my segmented run. He currently (as of 1/19/15) holds the RTA world record of 3:58:26 for this category. He is directly responsible for many of the improvements seen here, including the solo Goblin grinding/trap at Mandalia, getting to Level 23 for 9 Speed as a Priest before Golgorand, using Orlandu in Chapter 4, manipulating the camera, and others that I'm sure I've forgotten.

Also, shoutouts to my many Twitch viewers who kept me entertained during all of my resets, and to the others also attempting FFT RTA runs in various categories (elmagus, Tide and johannhowitzer).

Thanks and enjoy the run!

Farlem

Return to the Game List, the FAQ, or the Home Page.