Somtimes inspiration strikes in odd ways. Back when I was playing through the journey of my solo Thief Bernie, I thought more than once about going back and doing a re-run of the solo Black Mage which I had initially completed back in 2004. It's been almost a full 20 years since then and I'd like to think that I've gotten much better both at understanding Final Fantasy's gameplay as well as writing more informative reports for the website. I'd also have the chance to run the gameplay using Final Fantasy Renaissance, fixing the bugs in the original NES version and opening up new avenues for the Black Mage class to explore. Still, this would nevertheless be retreading old ground and I wasn't sure that it would be all that interesting to hear about me doing the same things all over again in a slightly different package. What I needed was a new spin on the Black Mage class that I hadn't tried before.
And that's when it hit me one night while I was finishing up with Bernie: what if I played a MELEE-ONLY version of the Black Mage class?! That is, a Black Mage that couldn't use offensive spellcasting and had to rely on their pitiful physical attacks for damage. It sounded completely ridiculous, like taking the Thief class and crippling it even further with a setup that had less damage, less health, and lacked the incredible running odds. And just to make things even more crazy, I wanted to do this as a solo character as well. 
Ludicrous as it might sound, I legitimately thought that I could make this concept function thanks to the bug fixes in Final Fantasy Renaissance. Weak though the Black Mage might be in melee combat, the class now gets access to a fixed TMPR plus eventually SABR from the Power Gauntlet on top of the fabulous FAST spell which has always worked. Were those tools sufficient to complete the game? I was about to find out.
One additional important note: I was defining this melee-only concept as no direct damaging spells from the Black Mage, and yes, that included instant-kill spells. No RUB or QAKE eliminating things instantly, that's a form of direct damage even if no numbers pop up. Status effects don't deal direct damage though and would be OK for this variant; I expected to be getting a lot of value out of the fixed SLEP spell at the very least. I had little to no experience with most of these spells (since it's almost always better to just kill things with Black magic in this game) and it was going to be fun testing them out at length. For naming purposes, I wanted to include a shout-out to my original character Blak without duplicating the exact same setup. I settled on naming this character Kuro which many readers will recognize is simply the word "black" in Japanese. Kuro, or more properly 黒, was about to begin on a long and difficult quest as my most handicapped character yet.
These are the starting stats for every Black Mage at Level 1. There's not much debating that the Black Mage has the worst base stats in the game and probably the worst stats in terms of scaling as well. Their Strength and Vitality are absolutely pitiful, stuck at the minimum value of 1 point apiece; every other class starts with at least 5 points in each stat. This would make Kuro innately terrible at melee combat while also giving him the worst HP growth in the game - fun stuff! Agility and Luck are actually pretty decent with starting values of 10 and here in Final Fantasy Renaissance that Luck will actually do something in terms of running away from combat. Too bad the Black Mage gets fewer guaranteed points in Agility and Luck than any other class in the game going forward; you didn't think that the class would get to retain its advantage here, did you? Then there's Intelligence which is far higher for the Black Mage than any other class and is always guaranteed to increase on each level up. This mastery of Intelligence was rewarded in the base game by being a completely useless stat; yes, base Final Fantasy does not use Intelligence in any way, shape, or form. Here in Final Fantasy Renaissance, it increases spell damage by (Intelligence / 2) percent which will sadly get wasted since Kuro won't be casting attack magic. I guess it will help improve the HP recovery from the Healing Staff later on, hooray.
So the Black Mage starts out the game as horrifically bad at doing anything other than casting direct damage spells. This situation then gets even worse over time as their stat scaling is inferior to every other class as well. For example, Fighters are guaranteed 147 stat point gains over the course of the game while Black Belts get 152 stat increases. Black Mages? They get a mere 103 guaranteed stat increases. The player pretty much has to do stat manipulation if they intend to play a solo Black Mage and I certainly intended to do it here. Black Mages also receive 2 points of Magic Defense with each level (Fighters get 3 points and Black Belts get 4 points) and a paltry 1 point of Accuracy per level up. This means that it takes 27 levels to reach 32% Accuracy where they unlock a second hit and then they will not get a third hit before reaching the cap at Level 50. Now this stat can be increased with the Accuracy rating from equipped weapons but sheesh, Black Mages really are not meant to be used as melee fighters.
I started off the gameplay by purchasing a Small Knife (5 Attack / 10 Accuracy) and the minimal Cloth armor for 1 point of Absorb, along with the SLEP spell for the only allowable offensive magic. I suppose that Kuro could have also picked up LOCK but that's truly useless ubervariant material there. Then I had to kill off the three accompanying White Mages which was surprisingly difficult to do and took several tries before Kuro could get it right. The biggest problem there was Kuro's utterly pathetic starting health of 25 HP which left him easy prey for even the most basic combat against Imps. These enemies have a basic Attack of 4 which means that they roll a number from 4-8 and then subtract out the defender's Absorb rating, which translated into 3-7 damage against Kuro's unprotected form. While they did miss about 10% of the time, 3-7 damage per hit was easily enough to overwhelm a solo target with 25 HP. Conversely, Kuro's own Attack with that knife was only 6 points which translated into 6-12 base damage. Imps have 4 Absorb so Kuro was dealing 2-8 damage per hit and typically needed two attacks to cut through the 8 HP on each critter. He was literally too weak to defeat the easiest random encounter in the whole game against 3 Imps. How in the world was this variant even going to be possible?
The answer to this conundrum was the SLEP spell. SLEP is a status ailment that puts the target to sleep if it hits successfully; it only works against monsters with less than 80 HP and it was bugged in the release version of Final Fantasy so that anyone put to sleep would always wake up on the next turn. This is how the spell was intended to work and now does work: first there's a roll to see if the target is successfully put to sleep, which has a base chance of 86% and decreases based on the magic resistance of the target (for Imps the calculation worked out to 78% success odds). Then for those who fall asleep, a number is rolled from 0-80 on each of their turns and if the number is less than their max HP, they wake up. Otherwise they stay asleep and since Imps have all of 8 HP the odds were excellent that they would remain in dreamland. Thank goodness that SLEP also hits all enemies on the battlefield instead of being single target which was downright amazing in this situation. Against the weak early game enemies with low health totals, SLEP was an utter godsend that made this variant workable.
Therefore SLEP transformed an impossible situation into something playable, albeit still quite difficult indeed. SLEP has the additional side properties of removing all evasion from the target so attacks will always hit something asleep and even increasing the Attack stat by 5/4 when hitting a sleeping target. I was using SLEP in every battle and it was quite noticeable that Kuro was hitting for more damage against a sleeping target, sometimes reaching damage numbers that otherwise weren't possible with his current setup. Now for the downsides to SLEP: any damage dealt to a sleeping target wakes them up and ends the effect so it wasn't possible to keep an individual target locked in sleep paralysis. In fact, the turn order is completely unaffected by sleep status so if Kuro acted first he could hit an Imp, wake them, and have that same Imp attack him later in that round! Final Fantasy's ancient interface was also not built to handle sleep status: I could see who was hit by SLEP when casting the spell, but NOT which Imps woke up naturally. The interface would say "Imp: woke up" without any graphical or text indicator as to which Imp that was. Thus Kuro had to stab blindly and hope he was hitting the right targets which often didn't work out as I would have planned.
The toughest part was making it to Level 2 even though it only takes 40 XP to reach the first leveling threshhold. I had to hope that Kuro would go early in the turn order and get off SLEP quickly followed by the Imps not waking up as he slowly stabbed them to death. It took three successful victories to reach the level up and I had to try a bunch of times for each one, saving at the Inn after each battle. Unfortunately Level 2 for the Black Mage is a terrible level in terms of guaranteed stats: Intelligence is the only stat guaranteed to increase (non-guaranteed stats have 25% odds to increase in Final Fantasy). I caught the biggest lucky break of the whole run when Kuro's very first time reaching Level 2 saw him hit in every critical stat simultaneously: he gained Strength, Vitality, and Agility plus had a great result on the HP roll for his "strong" level up, 4 HP when he could have rolled anywhere from 1-5!!! That was way better than I could have possibly hoped for and meant that I didn't need to grind a whole bunch of difficult battles hoping for the 1 in 16 odds to get Strength and Vitality to go up together.
Reaching Level 2 took Kuro up to 49 HP which was infinitely safer than 25 HP. He also gained a point of damage and now he could safely win one battle with SLEP abuse before heading back to the inn. He was slowly bleeding out gold because each inn usage cost 30 gold and he was only gaining 18-30 gold per Imp fight, though the rate of income decrease was slow enough to be manageable. Unfortunately Level 3 gave Kuro basically nothing, all of 1 additional hit point along with Strength and Vitality points that did nothing yet and spell charges that he couldn't use. Level 4 was another "strong" level which pushed Kuro up to 74 HP and was finally sufficient to win two random encounters in a row before hitting the inn, thus reaching a positive gold flow. I used this level to save up enough money for a Tent, then vetured into the Temple of Fiends at Level 5 and grabbed the Cap inside for a second point of Absorb. It actually did make a difference against Imps, cutting their damage from 3-7 to 2-6 per hit. Kuro was ever so slowly getting stronger.
Once Kuro was stable in terms of fighting random encounters, I could start thinking about the first boss of the game. Garland is infamous in Final Fantasy circles for being a joke opponent, with your typical non-variant team able to walk straight into the Temple of Fiends and kill him at Level 1. Things can be a little bit tougher for solo characters though and I'd had a lot of trouble defeating Garland during my solo Thief playthrough. Things were even worse here for Kuro who had less damage and less health while doing this funky melee Black Mage thing. Garland has 106 health, a base Attack of 15 which translated into 13-28 damage per hit against Kuro's minimal defeses, and an Absorb of 10 which greatly cut down on Kuro's own strikes. I was targeting Level 8 as the first realistic moment where Kuro would have a shot at Garland. By forcing Strength increases on each level, this would take Kuro to 8 Strength and a total Attack of 10 after factoring in his Small Knife. He would therefore deal 1-10 damage per hit against Garland and it didn't take a genius to spot that the side dealing 1-10 damage didn't stand much of a chance against the side dealing 13-28 damage. How in the world would my solo Black Mage stand any chance?
Well, there was one advantage that Kuro had that Garland didn't have: he could drink Heal potions. Purchaseable at a cost of 60 gold apiece, Heal potions are the only recovery item in the entirety of the original Final Fantasy. They are almost always used to restore 30 HP out of combat but can also be consumed using the "Drink" command in-battle to restore 16-32 HP at a time. This uses up the character in question's turn and scales extremely poorly such that's it only ever worthwhile to do at the very beginning of the gameplay. However, Kuro was indeed at the start of the gameplay right now and drinking a ton of Heal potions would give him an edge that he desperately needed. I took all of the gold that I'd accumulated while laboriously grinding to Level 8, about 2500 gold in total, and converted that into 25 Heal potions along with a Tent to save outside the Temple of Fiends for repeat tries at the boss. By the way, it takes about 5500 XP to reach Level 8 and Imps pay out 18-30 XP per battle, just in case anyone was wondering how tedious this opening sequence had been.
Then it was time to face Garland:
Monster AI scripting is extremely simple in the original Final Fantasy and most combats can be boiled down to a simple mathematical formula. Garland literally does nothing other than attack each round, with hits dealing the aforementioned 13-28 damage and with him having about 91% odds to hit (expected damage per round = 18.66). Kuro was dealing a less impressive 1-10 damage per hit with 84% odds to hit (expected damage = 4.27) so he had no chance whatsoever under normal circumstances. Even landing a critical hit or two wouldn't swing the math of this combat and it didn't help either that the Small Knife only has 2.5% odds to crit on each attack. What did change things was the Healing potion math: 16-32 HP per potion consumed for an average of 24 HP restored. This was higher than Garland's expected damage so on net Kuro would get back roughly 6 health for each round spent drinking potions. In practice, the potion healing and damage taken from Garland mostly canceled each other out, and then occasionally Garland would miss and Kuro would get a big spike in health. The whole situation was incredibly awkward and expensive for Kuro's pocketbook but it did in fact work at giving Kuro more effective max health.
So I began running a series of attempts against Garland, tracking how much damage Kuro was able to do before he ran out of health and potions. Some of the attempts went badly from the start and led to instant resets while others had better RNG and I kept them going until all of Kuro's resources ran out. He made it over 80 damage inflicted twice and then had an attempt where Kuro landed a critical hit that faltered at 101 damage dealt, a mere 5 HP short of winning. That's when I decided to go back to Corneria and buy another eight Heal potions, now up to 33 in total, which cleaned out almost all of Kuro's stashed funds. Given what I had just experienced, it was a matter of time until Kuro had another similar stretch of luck and managed to win out. He did finally pull off the victory after about four more tries, albeit at the cost of drinking all 33 Heal potions, sheesh! It only took reaching Level 8 and writing 2500 words in this report to make it past freaking GARLAND, heh.
At least the next boss in line would be extremely easy for Kuro. I sent him to raid the minimal potions held in Matoya's Cave, then ran away from random encounters until reaching Pravoka. There was a very minor equipment upgrade to be bought here in the form of the Gloves, taking Absorb from 2 to 3 points, along with access to the Tier 2 Black Magic spells that I'd be picking up next. First up was dealing with the Pirates and gaining access to the SHIP, a task that barely even registered thanks to the power of the SLEP spell. Kuro had long practice in using SLEP by now and the Pirates with their grand total of 6 HP were easy prey for the trip to dreamland. Eight of the nine Pirates fell asleep to the initial casting and they had all of 6/80 = 7.5% odds to wake up each round thereafter. Although a couple of them did wake up over the course of the battle, Kuro was killing one foe every round and they lacked the damage capacity to stop him before he murdered the whole lot. I'm really glad that SLEP works correctly here in Final Fantasy Renaissance as otherwise this would have been pretty nasty (as it was for my original character Blak).
I used the gold infusion from defeating the Pirates to purchase the next key spell for this run: TMPR. This was another spell which was bugged in the release version of Final Fantasy and literally did nothing at all when cast. Now it has its intended function: it adds 14 to the Attack value of the target. (My mechanics guide states that it's supposed to add 14 to both Attack and Accuracy, which it definitely does not do in Final Fantasy Renaissance, so either Ozmo implemented this incorrectly or the reference I have is incorrect. I think this is fine though: the lategame spell SABR adds to both stats while the early game TMPR only adds to Attack.) Gaining 14 additional points of Attack was absolutely massive for Kuro who had been stuck at 10 Attack against Garland. It more than doubled his offensive output! Once I had the chance to sail down to Elfland and purchase the slightly-better Large Knife with 7 Attack / 10 Accuracy, TMPR took Kuro up to a grand total of 26 Attack. That was 26-52 damage per hit against unarmored targets like the Sharks in the ocean:
The watery reaches of the inner sea were indeed the best place for Kuro to grind out additional gold and experience for the moment. He still didn't have anywhere near enough health to stand up to larger mobs of opponents (ironically the very encounters that the Black Mage class normally excels at defeating) but Kuro was pretty good at taking down isolated Sharks or small groups of Kyzkous for their big gold payouts. He also was forced to retreat or simply died against many of the random encounters out on the ocean; I don't want to overstate the power of the new tool available to Kuro. Nevertheless, the solo Shark battle was usually winnable and it paid out 267 XP which felt like mana from heaven after grinding against all those Imps at 1/10th the value earlier.
I kept Kuro near Coneria for a little while since the monster encounters on the SHIP were identical anywhere on the inner ocean - no reason not to stay right by the cheapest inn in the whole game. I did make one venture over to the Dwarf Cave for the 1000 gold inside which Kuro turned into a Copper Bracelet down in Elfland. Now up to a whopping 6 Absorb with the Cloth replaced, better than nothing I suppose. I was still healing and saving after every random encounter because so many of them went poorly for Kuro. He continued piling up gold with help from the occasional Kyzoku fights until I had 1500 in total. This was enough to purchase Kuro's next major spell:
The traditional FIR2 and LIT2 were unavailable for my Black Mage which meant that it was time to break out HOLD instead. I had literally never used this spell at any point in my many previous playthroughs of Final Fantasy and it was worth looking up how it functioned. HOLD is a status spell that inflicts paralyzation; as a quick reminder, the calculation to check if status spells succeed is to take the base value (148) and add the accuracy rating unique to each spell (64 in the case of HOLD), then subtract out enemy Magic Defense and divide by 200. Against an Ogre with its Magic Defense of 65, the odds worked out to (148 + 64 - 65) / 200 = 73.5%. Not bad! If the spell succeeded, the target would be stuck in the paralyzed status condition which Final Fantasy veterans will surely have encountered from various undead enemies. Paralyzed targets are frozen in place and can't do anything until it gets cured; player characters have 25% odds to recover each round while monsters have only 10% odds to break out. It's a rare place where the gameplay actually favors the player for a change.
HOLD was therefore an extremely useful spell for locking up enemies as the odds were so poor for them to recover once the status successfully landed. While it's true that sometimes Kuro would get unlucky and the monster would pop right back into the fight, normally they would be paralyzed for the duration of the combat. There was even a bonus effect here: like the sleep status condition, hits against a paralyzed target have their Attack value increased by 5/4. The one real drawback was the single target nature of HOLD as it could only hit a single enemy at a time. This also meant that it was a slow-acting spell, with Kuro needing to burn a round casting HOLD and then buff himself with TMPR before he could switch over to attacking. I found that the Ogre/Creep encounters around Elfland weren't good places to use HOLD as there were simply too many monsters present. The Creeps did too much damage to be ignored and Kuro didn't have the time or spell charges to paralyze three or four opponents at once. Instead, I continued to fight in the ocean where Sharks were the best foes for leveling. Sometimes Kuro would get a Shark + Red Sahag battle or the 1 in 64 double Shark battle, situations where he could freeze both enemies with HOLD and rake in the experience afterwards. I concentrated on gaining more levels while saving up gold once again.
Several levels later, Kuro had the 4000 gold needed for his first Tier 4 Black Magic spell:
It was another status-inflicting spell that I wanted, this time CONF for the confusion status effect. CONF had the same spell accuracy rating as HOLD but it put monsters into the confused state where they would attack one another or even themselves rather than attacking Kuro. The confusion effect was bugged in the release version of Final Fantasy where monsters would always cast FIRE when confused for some reason, however now it works perfectly in Final Fantasy Renaissance as they use their actual attack value for hitting themselves. Obviously this was super useful - what's better for an offensively-challenged variant than having the enemies kill themselves? - and it could even be combined with HOLD as needed to stun particularly dangerous foes. The biggest advantage of CONF was its multi-targeted nature as it hit all enemies on the screen instead of one at a time. Now Kuro finally had a solution for those combats with lots of monsters present on screen rather than needing to HOLD them all individually. The biggest downside was significantly higher odds for monsters to snap out of the effect: 25% odds per round instead of the 10% odds on HOLD. There also was no Attack bonus against confused targets for what that was worth and, as with SLEP earlier, no way to know which enemies in a group were breaking out of the confusion effect and which ones remained dazzled.
Over time and with a lot of practice using these spells, I came to think of CONF as a temporary way to get enemies off Kuro's back while HOLD was the tool of choice for the most dangerous threats. CONF also didn't really defeat that many opponents since monsters don't hit that hard with their blows (otherwise they would kill the player characters too easily), it was more of a distraction while Kuro stabbed them to death after using TMPR. CONF was still fantastic even with its limitations though and it allowed Kuro to fight the more powerful enemies around Elfland for the first time rather than simply pass through to shop at the stores. Since the damage from CONF was also tied to the monsters themselves, not anything that Kuro himself was doing, it opened up the potential to go hunting for big game:
Yep, those were the Giants encountered on the infamous Peninsula of Power otherwise not accessible until much later in the gameplay. For those unfamiliar with this bit of Final Fantasy lore, there's a very small 4-tile encounter zone northeast of Pravoka where the developers accidentally mapped the encounters from Lefein and can be used for various grinding strategies since the extremely strong enemies pay out so much gold and experience. There were five different battles to be found here: the Zombull/Troll encounter was unwinnable since Zombulls have status immunity, no chance of hitting them with HOLD or CONF. The FrWolf battle had too many targets present for Kuro to have much of a chance unless he would get insanely lucky. The Tyro encounter is the 3 in 64 odds battle (along with the even rarer Wyverns for the 1 in 64 encounter) and I didn't even see them pop up. The fifth and final encounter is against 2-4 Giants and this was the one that I focused on with Kuro. Giants have 240 HP, an Attack of 60 for 54-114 damage against Kuro, and enough Defense that he couldn't even scratch them before applying TMPR. Their Magic Defense value of 120 wasn't especially high, however, giving HOLD and CONF 46% odds to hit - not terrible! In fact, the spells seemed to hit at a much higher rate than that so I wonder if Ozmo has added Intelligence scaling to the spell to-hit calculation formula. Maybe I'll be able to get an answer to that on the FFR Discord at some point.
Anyway, Kuro typically was smashed into a pulp where he fought against these Giants since he would die in a handful of hits while the Giants needed lots and lots of attacks to perish. If he could get off the initial CONF spell quickly, however, and the Giants had poor luck at breaking out of the confusion effect though, well, eventually they would destroy themselves with reflected damage. I had brought a Tent along for a save point and I could keep trying the battles until the RNG worked out in Kuro's favor. Sure enough, I was able to win a few of these Giant encounters and cash in on the huge rewards:
Jackpot! 
I typically don't spend much time abusing the Peninsula of Power since it's basically a programming error and not necessary to level your characters or outfit them with equipment. This was absolutely the perfect variant character to make use of the area, however, a Black Mage who had almost no direct damage but lots of status twinkery. And it wasn't as though I was making use of the FIR2/HRM2 cheese that most players employ here, Kuro was having to think on his feet and keep his opponents tied up in a fog of paralysis and confusion to make any progress, turning them against themselves rather than unleasing elemental magic. These battles were a real challenge and I had a lot of fun figuring them out while they simultaneously sped up the grinding process of getting to some ungodly level before venturing into the Marsh Cave.
I was mostly focused on gaining experience but the Giant fights did yield an equal amount of gold as well. It was enough for Kuro to max out his item stash with 99 Heal potions as well as all the Pures and Tents that he would ever need. More importantly, Kuro was able to purchase his other critically important Tier 4 Black Magic spell:
Ahhh, the FAST spell: the one spell available to Kuro that actually gets used in a non-variant game! FAST is well known for being the best buffing spell in Final Fantasy, granting double the number of hits to the recipient and letting your party's Fighter or Black Belt annihilate whatever boss of the moment might happen to show up. In fact, this is the biggest reason why status spells like HOLD and CONF are so pointless in normal gameplay: why bother when everything dies in a single round to the party's melee fighters? The developers clearly didn't assign enough HP to the bosses in this game which is why nearly all of the various remakes buff up monster health in some fashion. Anyway, I was able to figure out the math behind how FAST works in Final Fantasy Renaissance through trial and error. The spell doubles the Accuracy rating of the target and then cuts in half the Accuracy rating needed for each additional hit, from the default 32% to a lowered value of 16%. This allowed Kuro to get three hits instead of the expected two hits from FAST, as he had a base 28% Accuracy which was doubled to 56% and then that was enough for three hits at the lowered 16% per hit. Note that the spell did not work this way in the original Final Fantasy and merely doubled the hit total so Ozmo definitely tinkered with the math under the hood on this spell.
Kuro quickly settled into a routine for gaining the levels that he needed to make further progress with the main plot. He would set out from Pravoka, walk over to the Peninsula of Power and then save with a Tent, followed by trying his best to win a single combat against the overpowered enemies. Then I'd save again with another Tent and head back to Pravoka to restock spell charges before repeating the whole sequence over again. As Kuro piled up additional levels, he began to have better odds to survive against the Giants and the Frost Wolves (he still had no chance against the Trolls and Zombulls). Whereas originally Kuro was winning maybe 1 in 20 of these combats, eventually he improved to something like 1 in 4 odds. This was largely due to having more total HP available so that he didn't die in three or four monster attacks. I also had more luck than I expected against the Frost Wolf packs; they hit about as hard as the Giants while having much less health so they were pretty good at killing themselves as long as Kuro was diligent about reapplying confusion status.
When Kuro had less than 2000 XP needed for the next level, I'd break out of this pattern and go fight easier targets to avoid overshooting the mark. I absolutely needed to keep manipulating his stat gains on each new level and that meant getting Kuro as close as possible to the next level, then fighting quick battles against Imps over and over again until rolling the necessary stat gains. Kuro had to force Strength and Vitality points for each level which was a real pain as the Black Mage has few innately guaranteed points in these stats. It was 1 in 16 odds to hit both stats when they were missing from the guaranteed list and Kuro had to force that low-odds dice roll over and over again. He had to do this on Level 11, Level 14, Level 15, Level 17, Level 18, and Level 21! And while sometimes these stat gains arrived quickly, at other times it could take ages on end before Strength and Vitality finally appeared together. I had to re-roll the level up stats over 40 times for one of these levels and over 50 times for another level. That was more than half an hour of real-world time just sitting there resetting the game over and over again. Sir System Error mentioned trying to force good HP rolls on the "strong" levels in addition to Strength and Vitality but that was more repetition than I was willing to stomach. This was plenty bad enough without bringing the HP gains into the picture as well.
Eventually though, Kuro did reach Level 22 which was the one I had been targeting for heading off to the Marsh Cave. This was another "strong" level up and brought Kuro's max HP total to 278 which I hoped would be sufficient. My Black Mage had started with 1 point of Strength and Vitality, then gained a point in each on every level, leaving him with 22 points in each at the moment. This brought his base Attack up to 19 with the Large Knife, which would increase to 33 Attack with TMPR and then double to 4 hits with the FAST spell, Kuro having picked up his second non-FASTed hit at Level 18. He also had a rather decent 23 Luck which would give him 23 / 37 = 62% odds to escape succesfully from combat. Better yet, his level was high enough that a decent chunk of monsters would now run away from him, as I had observed Ogres and Creeps now fleeing in terror from this bizzare melee spellcaster. With 99 Heal potions and 41,000 unspent gold in his inventory, Kuro was as ready as he could be for the game's first true dungeon.
Kuro indeed proved to be ready for the Marsh Cave as he had few issues making it to the Wizards down on the third basement level. He had a solution of some kind for almost all of the random encounters that could appear inside the dungeon. Shadows were extremely weak opponents who could only inflict the pointless Dark status condition while Scums were basically the same story only attuned to poison status instead. Gargolyes and Scorpions had a novel solution: tag them with the SLOW spell to knock them down to a single hit per attack. This dropped their damage down to a manageable level where Kuro could fail a whole bunch of running attempts and still get away with health to spare. WrWolves actually ran away from Kuro since they have low morale and he was ridiculously overleveled for this point in the game. The two combats that I most feared were the Crawl/Bone pairing and then the old Ghoul/Zombie stun-a-thon encounter. Kuro had excellent success against the former enemies by casting HOLD on the Crawl monsters and then running away while the beast was stunned. Only the Ghoul/Zombie undead fight was an unbeatable encounter where Kuro could only run and pray for good luck. This one group of monsters wiped him out frequently but nothing else could consistently stop him.
That was aside from the Wizards themselves, unfortunately. The guardians of the stolen CROWN were the bosses of the Marsh Cave and they were no easy pickings for poor Kuro. Wizards have 84 HP, an Absorb rating of 16 that cut out most of Kuro's damage, and an Attack rating of 30 - plus the ability to hit twice! That was 24-54 x 2 damage against Kuro's weak protection and when both attacks landed his health dropped in a real hurry. At least the Wizards have low Accuracy and missed a lot but this was still pretty bad. Worst of all, Wizards have immunity to status effects in Final Fantasy which effectively ruled out everything in Kuro's bag of tricks. 1 in 200 odds to freeze them with HOLD, 1 in 200 odds to land confusion status, etc. It was a real shame because this another battle where the SLOW spell would have worked like a charm.
Shorn of his status trickery, all that Kuro could do was buff up his melee attack and swing away. Therein lay the problem: he needed one round to cast TMPR, then another round to cast FAST, and all the while the Wizards would be bashing away on his frail magely self. This battle can spawn 2, 3, or 4 Wizards and I had to hope that Kuro would get the 2 Wizard draw as any higher number of foes was unwinnable. Even against a mere pair of these opponents, I found Kuro running out of health too quickly before he could swap over to the attack. Each Wizard was averaging 50-70 damage per attack and Kuro had to absorb four such attacks before he could even start swinging back. Furthermore, the Wizard fight was down at the bottom of the Marsh Cave and there are no save points in Final Fantasy, plus two out of every three boss fights would have the unwinnable 3 or 4 Wizard setup. Long story short, I ran attempts at this battle for a full hour without success, with the screenshot above capturing the only time that Kuro managed to kill a single Wizard. He just didn't have enough health at 278 HP; it was clear that I needed a few more levels for a larger lifebar.
Thus it was back to the Peninsula of Power once again, sigh. The bad news was that Kuro would't get another "strong" level until hitting Level 26, four more levels away. He would be getting a grand total of 6 more HP on each level before that which wasn't likely to make much of a difference. I had no choice but to grind out those additional four levels which averaged a little over 20,000 XP apiece. This would have taken an ungodly long time without the Peninsula of Power and I'm incredibly glad that it existed for this variant! For that matter, Kuro became quite adept at these random encounters and had especially good luck at taking down the FrWolves with repeated CONF spells. He even managed to get the rare Tyro encounter a few times which was surprisingly easy to win: HOLD for paralyzation and then TMPR/FAST to cut down the dinosaur for a cool 3387 XP. Only the Troll/Zombull pairing remained unwinnable since the zombie bulls had the same status immunity as the Wizards.
This report can hit the fast forward button and skip past those additional hours of grinding XP. Kuro now returned to the Marsh Cave at Level 26 with a new total of 329 HP. That should be enough to survive one more attack from the Wizards which I hoped would be sufficient to win. I noticed right away that more monsters than ever were running away from Kuro now that he had even more levels under his belt. The enemy running formula in Final Fantasy is Morale - 2 * [character level] + (0...50) < 80 so at Level 26 that meant take the enemy's Morale rating, subtract 52, add a random number from 0-50, and then run away if the total result was under 80. Crawls for example have a low Morale of 106 and the math worked out that they would run about 50% of the time when it was their turn to act. The Wizards themselves only have a Morale of 126 so it was theoretically possible for them to retreat as well (about 12% odds). Sure enough, when Kuro confronted the Wizards at his new higher level, some of them wanted no part in the combat:
Yes, they can actually run away even in this supposed boss fight! 
Bosses have no special coding in the original Final Fantasy, no "heavy" flag as in Final Fantasy 5, and Sir System Error has managed to get Chaos to retreat in the final battle with some of the Final Fantasy Renaissance wonkiness. In this particular fight, Kuro was facing three Wizards only to have one of them turn tail and run on the first turn of combat. I buffed up Kuro with TMPR and FAST while the other two mind flayers attacked him, only for Kuro to kill the first one followed the the last remaining Wizard retreating on the same turn. Kuro didn't get gold or experience for the ones that ran but that hardly mattered. He then had to make it back out of the cave and fortunately ran into nothing dangerous while making his exit. It was kind of shocking how smooth and easy this was compared to the previous attempts - what a difference a few levels made.
Kuro had spent almost none of his spell charges passing through the Marsh Cave and therefore I headed straight to Northwest Castle to face Astos, saving outside with a Tent for safety of course. Unlike the plight of the solo Thief, Astos was not an opponent that concerned me with Kuro. The dark elf lacked the same protection against status conditions held by the Wizards:
Have fun not moving, buddy! Astos has a high Magic Defense rating of 170 which lowered the to-hit rate of spells like HOLD and CONF to roughly 20%. That was difficult but not impossible to pull off, and in fact Kuro would land the pictured HOLD on the very first try (further evidence that Ozmo has likely adding Intelligence scaling of some kind to this formula). This gave Kuro the space he needed to cast the TMPR/FAST pairing once again and bring his damage up into the respectable zone. Astos relies on his massive 40 Absorb rating for protection and yet these buffing spells actually let Kuro cut through those defenses without having to rely on critical hits. They took Kuro up to 35-70 base damage, 1-30 after subtracting out that boss Absorb, along with four hits per attack. I was seeing a lot of "4 Hits! 30 Damage" printouts as Kuro hacked away. Astos broke out of the paralyzation after a few turns, then fortunately missed his RUB spell followed by missing his SLO2 spell as well. I could have countered that with a second FAST if it had been needed, of course. Astos lacked the time to cycle around to anything else too dangerous in his spell rotation before Kuro's knife flurry finished him off:
It was a fairly easy victory on the first try, with only the RUB spell having any real danger for Kuro. He even could have countered Astos' FAST spell with his own SLO2 if necessary. It was a great victory to be finished, finally, with this portion of the gameplay. I suppose it's worth pointing out that Kuro had to be Level 26 to complete this whole sequence, roughly the same point where a non-variant party would be beating the whole game. Beggars couldn't be choosers though and some of the toughest parts of the gameplay were now in the rear view mirror. The Earth Cave was next on this list but will have to wait until the following page of this report.
