Tuesday, March 29, 2011

FF1 balance changes

I didn't actually get any submissions for my FF1 balance contest.  This didn't hugely shock me as it isn't like everyone goes through a 'must play random old video game X' phase at the same time.  That said there were some interesting discussions in the comments so I figured I would post what I came up with anyway.

Magic Changes:

1.  Intelligence does something!  Intelligence adds to your chance to hit with spells.  This makes the Black Mage significantly better at casting attack spells than Red Mage / Ninja and that all that Int the Black Mage gets is useful.  Base success rate of spells reduced from 148 to 128 for players so that overall spell damage does not inflate too much.

2.  Many bad single target spells now hit all applicable targets.  I would have included Rub in this list but since several mobs cast it on the party and it is already brutal when they do this I don't think changing Rub is a good idea.  It means Rub is still bad for players to cast and devastating when used against them.


Fire, Lock, Lit
Ice, Tmpr
Sabr, Blnd, Brak


Lamp, Invs
Pure, AMut

3.  Several bad spells doubled in efficacy.  As part of this all buff/debuff spells no longer stack.  Specifically the dodge spells need to not stack so that ridiculous situations like the whole party being invulnerable to physical attacks don't occur.  I realize this makes the White Mage only party totally useless but I am okay with that.


4.  Changes in spell levels:  Pure to L2, Soft to L4, Mute to L6.

The idea behind these changes accomplish 2 things.  First, to make spells that were almost never worth casting actually decent, which is true for most of the changes.  The other major shift is to make the low level BMagic damage spells decent.  Low level BMages are total junk compared to melee because they get only a couple of decent actions before they are back to swinging uselessly with daggers.  Now BMages will actually have a few totally awesome actions before having to resort to useless swinging instead!  RMages might even learn and cast attack spells before Fir2 also.  Obviously beyond about level 10 this stops being very relevant, which is fine, because at that point BMages have enough spells they can contribute pretty consistently.

Thief and Fighter changes.

Mostly people agree that all the classes have good and bad points and are reasonable inclusions in a party except for Fighter and Thief.  Fighters are flat out the best to the point that it is probably almost as quick to win with 2 Fighters as it is with a 4 person party including 0 Fighters.  Thieves are unbelievable junk.  The primary reason for this is Fighters get the best attack progression, best weapons, best armor, best HP and best damage progression and Thieves get to be either average or terrible at pretty much all of these.  Changes:

Base Attack%:  5  (was 10)
Attack%/level:  1  (was 3)

Base Attack%:  10  (was 5)
Attack%/level:  6  (was 2)

With these changes Thieves hit about as hard as Fighters used to and Fighters hit about as hard as Thieves used to.  Fighters will still be massively tough and fantastic for filling the front line but for hard hitting physical attacks you will want either a Thief or BB.

Difficulty changes.

These last changes are designed to make the parts of the game that were simply too easy significantly more challenging.  I don't mind some parts being easier than others but in any fantasy setting the final boss is supposed to be a real challenge and the fight should take awhile.  Many of the game's dungeon bosses are simply too weak and die too quickly - it is common for a fairly challenging dungeon to be ended by a boss fight that lasts only 1 or 2 combat rounds.

Vampire, Lich, Kary, Kraken, Tiamat, Chaos:  Health increased by 100%.
Pirates:  Health increased by 900% to 60.


  1. What's the problem with white mages having the ability to become invincible? Is the game actually more fun if you can't do crazy things with silly parties?

  2. I don't have a problem at all with 4 white mages being invincible to physical. 4 white mages is obviously a terrible party even if you are invincible to everything! The problem is that in order for high level mitigation spells to be worth casting they need to mitigate a lot of physical damage. In fact they need to mitigate so much that a /item of White Shirt + a cast of Invis2 would make the party very close to invincible and that is easy to achieve even without a White Mage in the party at all and can often be done before a monster gets an action.

    If you make Inv and Fog type spells weak enough that unlimited stacking really isn't feasible under normal circumstances then an individual cast is completely garbage and the spells become useless again. I don't mind niche parties having crazy gambits they can play but I think having spells whose *only* practical use is stacking up to total invulnerability status is a poor use of the limited spell pool. If you can suggest an implementation that makes Inv and Fog type spells useful to cast under normal circumstances (as in, competitive with Cure/Heal spells in a tough encounter) and also prevent parties from being invulnerable to pure physical bosses like Kraken on round 1 as well as continuing to allow 4 white mages to be invincible I will concede that your design is superior. The only way to achieve that at all that I can think of would be to implement new tech allowing each individual caster to have an active buff of each type on each target such that you would need several casters to achieve invulnerability. Again though a party of WMage, Rmage, X, X can use 3 copies of Inv2 and easily become invincible if the benefit of casting it is decent.

    The invulnerability of the 4 White Mage party is, I think, a small sacrifice to have more balanced options for all the other parties out there. They do still have a double strength Inv2 and Fog2 to make them really tough so they only lose out on fights where they really intend to win by hammering down for 50 rounds.

  3. I guess since the ability to become invulnerable is a bug that it doesn't really matter since even if you didn't put that rule in.

    I would argue that even the current invis (20% evasion) is better than cure spells, though. INV2 is 20% evasion to the party. The next tier down is 99 healing to an individual. Going from being hit 75% of the time to 55% of the time will prevent 27% of the incoming damage, so as long as the boss would have swung for 367 damage over the course of the entire fight it's worth casting INV2. If he isn't swinging for at least that much then what use is any white magic spell in the fight?

    This isn't to say INV2 is ever cast in the base game, but neither are healing spells. Not in combat, anyway. I don't think you're going to find a point where white magic is good in combat without becoming mandatory.

  4. Remember that many bosses do a huge portion of their damage from magic attacks and you don't know which sort of attack is coming. Also remember than 20% evasion is not 20% damage reduction - it is quite random how much damage it prevents while Cure/Heal spells are more predictable. If I have the white mage cast Heal every round it will significantly help my party stay alive no matter what the RNG chooses for the boss' actions. This is not true for the protective spells since they only stop certain kinds of attacks so they need to be stronger to make up for that lack.

    Of course the waters are muddied because if you happen to get your Inv2 up *before* the boss does his big swing it is pretty good... but if he pounds you first and then chooses to blast you with fire and ice after the Inv2 is up it did nothing. Cure/Heal are the opposite because using them preemptively will often result in useless overhealing and using them reactively often results in people being dead before the heal can land. My feeling is that for protective spells to be useful they must be noticeably better than healing spells if used very early in a fight of substantial length. In order for combats to not be trivial however they protective spells have to be weak enough that they are a bad idea compared to a heal when used late in a fight.

    Of course in the NES version the fights against bosses average 2-3 rounds so protective magic has nearly no room to work - by the time you get up a couple of those spells the boss might not get to attack again.

  5. Chaos can cast CUR4, so the fight against him can't last more than a few rounds. If it does then you can't win. If you are going to give him a ton more health, you probably have to remove that from his spell list.

    Do your fighter/thief numbers really make things better? I just find their results really random. Using a level 29 knight (with Excalibur), a level 29 ninja (with Katana) and a level 29 master (with no weapon) it takes the following number of attack commands to kill Chaos: Kn - 68, Ni - 39, Ma - 15. For Tiamat in the temple of fiends its: Kn - 25, Ni - 20, Ma - 7. For Kraken its Kn - 12, Ni - 9, Ma - 4.

    Sometimes the Ninja does a lot more than the Knight, sometimes they only do a little more. When the ninja and knight are close the Master outstrips them by a mile. If you increase the level to 30 then the ninja will get a 7th attack and do must better than the knight but the master will start being more than twice as good as the ninja anyway.

    In Origins they changed the critical hit system. If you make this change then suddenly Chaos is 68/24/12, Tiamat is 25/14/7, Kraken is 12/9/5.

    The combat mechanics in the game are bizarre and I don't think any amount of screwing with the hit values for various levels will make them work. Everything that comes out is a coincidence. Sure a fighter does 69 damage and ninja does 46, but some monsters have 8 absorb and some have 100. Some monsters have 24 evade and some have 144.

    For trash monsters almost everything dies in one or two hits to any end game character whether you make these changes or not (which also makes AoE magic kind of pointless.

    A fighter doesn't take much less damage than a ninja. Tiamat hits a fighter for an average of 182 damage a round and hits a ninja for an average of 226, only 34 more damage. If they put elemental resists on the pro ring instead of the ribbon then the average damage difference would be only 18. Chaos swings at a knight for 226 and at a ninja for 239. Is it worth daking 13 less damage a swing from Chaos to do half damage and not be able to cast FAST?

    If you want to make the combat system work, I think you need to do a substantial rewrite. Number of attacks should not be bound to chance to hit. The power of spells should be considered based on the number of spells you can cast, the number of fights you'll do in a dungeon and how good just attacking is. If a ninja attacks Chaos for around the same damage that NUKE does, then why do you have a Black Wizard? Are enemies really dangerous enough to need a fighter leading your party? How good is it really to have one if their purpose is defense (aside from dragon armor being your fourth ribbon)?

    So much of what is good and bad in the game is wrapped up in a deeply intertwined web of strange decisions. There being no fourth ribbon is a huge part of why every party should have a fighter. If it weren't for that black belts would be just as good or better. I played the game with four black mages and while it was sometimes hard they won the game in a cakewalk at level 28 which is pretty average, health and armor really didn't mean much at all (though it was seemingly impossible to get past Tiamat with all four alive I only needed three to kill Chaos anyway).

    If your changes were implemented, what would change in the player experience of the game? How many more times would I have been able to replay and enjoy it? Would Ziggyny be replaying it right now with a single thief as a challenge?

    In order to make fights longer you need to do more than increase monster health, you need to change the way the whole game works to make having longer fights make sense. The enemies are often just a brutal as the players, casting spells for huge AoE damage, and healing is next to useless. I also think that Ziggyny is right that it would be very hard or impossible to find a point where white mages were neither required nor useless.

  6. Fighters take a lot less damage than other classes in the early game due to having ridiculous armor absorb values. That advantage shrinks once the party gets access to Gold Bracelets and then gets huge again once the Fighter gets the Dragon Armor/Aegis Shield. They still take less physical damage, but the difference isn't very extreme, as you said. They do take a boatload less magical damage though due to the Ribbon issue. Obviously this is because of the very specific itemization the game has but it does work out that the Fighter is a fantastic defensive character for the whole game. Keep in mind that they also have way more hitpoints than everyone else until the BB/Master catches up at very late levels.

    Here is the thing about the BB... they are ridiculous. Totally awful early game, solid midgame, completely overpowered late game. I don't like that design much but I felt like it wasn't worth changing since I was aiming for brevity rather than perfection and I imagine some people love that kind of progression.

    The Ninja/Knight damage progression isn't perfect. There are times when the Thief isn't that far ahead and times where he is way ahead. The game has breakpoints that make that inevitable. However, over the course of the whole game the Thief and BB will do the highest single target damage on average with the Fighter clocking in at around half of their damage. Note for this calculation I end the game at 30 - the fact that the BB does 1750 a swing at level 50 doesn't factor into it.

    You are completely right that the game is really swingy. Bosses do very dangerous things. That said, when I fight bosses at what seems like an appropriate level (since the dungeon is manageable and drains a lot of resources but isn't low % to get through) the bosses have absolutely no chance - they die in 2-3 rounds and nothing they do matters. If they lasted twice as long then defensive spells and healing would actually be very useful against them. As it is when I FAST my best thug normally the boss dies in one swing, two at the most. Doubling fight length would make all of the bosses really dangerous instead of having them be easier than many trash packs, which is the current situation.

    You are right that doubling the health of Chaos is the most problematic by a mile. CUR4 is a huge issue. Doing 4000 damage in something like 8 turns is extremely doable but challenging - it would make crazy parties like 4 White Mage or 1 Thief harder for sure but well designed parties should be able to beat it. I totally buy that 3000 health might be a better number though.

    I do think that these change would improve the game such that it would have been played even more than it was. The improvement would have been marginal, at best, but it would improve. Thing is that anyone who wants a crazy challenge can still get it. 1 Thief is still going to be terrible (not quite as much), 1 WMage is going to be even harder and you can find any number of crippling penalties for yourself if you want to. The big gain is this: There are many more ways to be nearly optimal. I like being optimal and trying to play as well as I can and doing that in FF1 as it is involves always using a few specific setups. I think both I and many others would enjoy being able to play many more different groups that are arguably optimal. I would enjoy having more different spell setups that are powerful and competitive so there would be more thought and consideration given to spell choice throughout the game. I think my changes make the number of ways to play the game really well increase drastically and leave the number of ways to play the game really badly still legion in number.

  7. First of all, your changes don't make the one white mage party bad, they make it not possible. They also don't leave the one thief party being bad, they make it a relative cakewalk.

    I also think that you underestimate SABR as a buff spell (probably because it did literally nothing in the original game). It is definitely competitive with FAST and better if the enemy has high Defense in some cases.

    But also, why focus on hit% as a way of balancing thieves and fighters the way you want? In my last post I compared fighters with excalibur to thieves with katana, but in reality the thief is going to have masmune (since it would be nuts not to give it to the guy with 7 attacks). By making fighters so miserable at attacking you are removing party combinations as well as adding them. Bringing two fighters becomes a terrible decision, since the second fighter, with a defense and no suit of dragon armor, will be worse than a thief is now (with the exception of Fi/Fi/WM/BM where fighter 2 gets Masmune).

    But if your plan is to give the second fighter Masmune, why bring a fighter at all instead of bringing a Red Mage who will be better at fighting, have spells, and be just as tough as a fighter would in that slot? You could ask why you are bringing the first fighter instead of the red mage, and the answer is this: Excalibur and Dragon Armor.

    If the only reason to bring a fighter at all is because of Excalibur and Dragon Armor then maybe the problem isn't fighters but it's Excalibur and Dragon Armor.

    A level 30 fighter has 67 damage and Chaos has 100 armor, that means his normal hits deal 9.25 damage to Chaos. If he was using the defense instead (the next best weapon at 15 less damage) then his average hit would be 1.11 damage. Does it make sense to have the best weapon do more than 8 times as much damage as the second best? Even factoring in criticals, Exacalibur still does 14 to defense's 2.5. (This example also demonstrates how SABR can be a more powerful spell than FAST)

    I think the damage and absorb values would be a much better place to attack the discrepancy. Giving the thief more than twice as many attacks as the fighter just creates huge variability in their performance depending on enemy absorb. Modifying attack and defense values would also allow the black belt to be addressed.

  8. You make a really good point about the issues created by having a class that attacks a bazillion times for not much damage. In changing thieves to be good I was trying to make the absolute minimum number of numerical changes while achieving the overall damage values I wanted. Having thieves attack progression stay the same but having them hit a lot harder would work fine too but it seemed more challenging to implement so I went with the simplest solution that roughly accomplished what I wanted.

    If I was trying to do a really serious rewrite of mechanics I would definitely tackle the insanity that is high level black belts and some of the absorb vs. damage value wackiness. SABR in particular is listed as buffing attack but in the code it appears to buff attack and hit% - and also doesn't do anything at all. I was using the SABR value of just buffing attack which makes it quite good to cast (and worth a 7th level slot) but if it buffs both it becomes really absurdly nutty. Of course with the huge number of attacks that thieves/BB get with my modifications it might well be totally overpowered.

    I was thinking that increasing thief damage would be tricky since multipliers to damage post armor don't solve the issue of absorb discrepancies and adding extra damage pre absorb either needs totally new mechanics or a massive str boost to thieves. One thing in favour of the 'thieves can equip two weapons' plan is that this would actually give thieves much higher damage values which gets away from the absorb problems.

    I think that the second Fighter is definitely a place where these changes make a party that was once top tier now not so much. The second Fighter is still good for defence since he has great armor throughout the game and does take twice as many attacks as the 3rd and 4th character in the party. He isn't nearly as impactful as the first though, not even close. However, the same can be said of the second copy of most classes - their equipment gets diluted and their effectiveness is less noticeable because their role is already covered.

  9. I think a second fighter would be basically a trap. He would be totally fine for a large part of the game (the part where you are buying your gear) and then you'd pretty much have to fight Chaos with three characters.

    By contrast, the second and even third ninja still do decent damage (and can cast fast on the ninja with masmune), any number of black belts remain powerful, and I can say from experience that more black wizards is just fine.

    I'm just not convinced that you've created more close-to-optimal parties. Yes, thieves are worse than fighters in the base game, but if Fi-Th-WM-BM much worse than Fi-Fi-WM-BM? For end game boss fights the thief gets to cast FAST on the fighter, so he's not really much of a detriment. For early game bosses the thief is pretty bad but it's not hard to win anyway. And for the first few dungeons being able to run away more easily might be an advantage worth having if for no other reason than to save time.

    I don't think this debate will be settled, but I'm still not convinced that the fighter/thief rebalancing would make the game better, just different. If the goal is to have more class combinations be "close to optimal" then I suppose you may have done that. I went through the 89 class combinations that contain at least one fighter and/or thief and I would say that with the original game 72 of these class combinations are either very good compositions or compositions that will have no trouble winning despite an obvious flaw. With your system I rate 79 of them the same.

    I think it's a very minor difference, I feel that the implementation is inelegant (thieves ending the game with well over 200 hit while no one else can reasonably get even 100 will feel very strange), and I think it avoids the real problem which is the way equipment and encounters are designed.

  10. I totally buy your accusation of inelegance. As I have said though, this wasn't a solution I considered perfect but rather one that would be easy to implement and understand and which did not rely on new tech within the game. Redoing the way equipment and encounters are designed in a big way is certainly going to be the best solution if I was allowing that... but given the design constraints I decided to set it wasn't allowed. If you have a better solution that fulfills the ideals of quick to relate and not requiring new tech I would be happy to hear it.

    I think you are placing an excessive amount of weight on the Chaos fight and ignoring the early game. Honestly I rarely have any issue at all with Chaos and I just push him over without much difficulty. I regularly find the Marsh Cave, Ice Cave and Castle of Ordeals to be the hardest parts of the game and for those challenges the Thief doesn't have FAST and has utterly wretched armor and attacks - his combat weakness is a serious problem. Note that running away in those dungeons is useful but I would trade in the Thief for a Fighter in a heartbeat regardless. Keep in mind too that the Ninja FASTing people on boss fights is often pretty terrible - when you have a RM or BM they FAST the Fighter anyway so the Ninja is really just FASTing himself so that he can hit as hard as a Fighter without FAST up.

  11. Alright, I think you are right that I make too much of a big deal of Temple of Fiends bosses when the Wizards in the marsh cave is probably the hardest fight in the game. But using your formulas a thief does quite a bit less than a fighter to a Wizard at level 8 and less than a third of the damage a figther does to Astos at level 9 with the best equipment available. Assuming you get only a longsword and not a silver sword the fighter still does more to Wizards and more than twice as much to Astos. So your fix helps quite a bit to make the thief competitive against wizards, but a fighter would still be better, and the thief still lags well behind Astos.

    There are only 40 weapons and 40 armor in the game, so I don't think overhauling this is a massive undertaking, especially not when the majority don't need any change. If you want to balance the classes better I'd do the following:

    Use FF Origins Critical system (A..2A instead of (2 * A..2A) - D

    Reduce Fighter starting strength to 10.

    All attacking classes always gain 1 strength per level.

    Hit% no longer determines number of attacks, attacks are gained at
    specific levels (I'd need to do calculations to figure out which, but probably around the same levels a fighter in the base game would expect more attacks).

    Fighters get 2 hit per level while thieves and black belts get 3, they all start with 10.

    Weapon damage for Fighter-only and Masmune weapons reduced (Long Sword 15, Silver Sword 17, Excalibur 37, Masmune 38, there may be some I've missed that don't normally get used in the current game)

    Increase Excalibur's crit to 10

    Reduce defense value of high defense monsters (maybe not IronGol), usually by around 2 + the amount of weapon damage lost by the best fighter weapon for that monster

    There is a Fourth ribbon somewhere in the game and increase ribbon absorb value to 6 (this is a bad kludge fix - resists should be better distributed on available equipment and characters should be able to carry a selection of armor instead of everyone needing a ribbon)

    Black Belts/Masters do not get double attack for unarmed, their unarmed attack damage is 12 + Level * 1.4 and their unarmed critical chance is their Level

    These changes would make the game about as easy for any party as it is for a two-three fight party currently, so increasing monster health might be desirable. I'm not really sure this would make the game more fun, since I think it would make choosing between the three fighting classes pretty irrelevant. It would increase balance, but you wouldn't feel like you were playing a different game if you played again with a different mix of beaters.