Scrapping Luck & implementing new Strength/Intellect Split

Yeti and I played a super fun game of spreadsheet theory emblem; basically, we wanted to examine the balance between stats, and assigned numeric values to each component’s usefulness to physical/magical/healing units separately, then took an average usefulness to calculate the standard deviation between stats.

Here’s a link to the spreadsheet for interested viewers; you can see Yeti’s awesome proposals too, and look at the numbers that go behind the math if you’re a nerd or something.

This is GBAFE (the colors are kinda arbitrary here, but they get consistent in the next two examples).

The basic conclusion here was that Luck, on its own, is a terrible stat; but that its attributes could be redistributed to better balance the other stats. Skill and Res also need general buffs, and Speed is obviously overpowered.

Our goal was to create our own proposals for new stat systems; I wanted to keep mine similar to the traditional FE style, in hopes that I could create something to implement in a hack. I’ve got two proposals, one that simply scraps Luck, and another that creates a new Str/Mag split.

Leave luck to the heavens, we have skill! Sell your Speedwings, kids! The reign of imbalance is over!

  • Attack Speed = ((Skill + Agility)/2 - (Wgt - Con, 0 if negative))
  • Cut out the Luck stat, redistribute its functions (Hit/2, Avo, Dodge)
  • Accuracy = (Skill x 2 + Def/2) + Boosts
  • Avoid = ((Agility - (Wgt - Con, 0 if negative)) x 2 + Res) + Boosts
  • Dodge (crit evade) draws from Speed now.
  • Staff Accuracy = (Skill x 5 + Mag (or Str/2))
  • Healing = (Staff Mt + Spirit/2)
  • Hybrid attacks (like magic weapons) draw from the opposite defensive stat now too.

inb4 “Arch its too complicated go home you’re drunk”

  • Physical Accuracy = (Skill + Intellect + Spirit/2) + Boosts
  • Magical Accuracy = (Skill + Defense + Strength/2) + Boosts
  • Physical Attack Speed = ((Speed + Skill)/2 - (Wgt - Con, 0 if negative)
  • Magical Attack Speed = ((Speed + Skill)/2 - (Wgt - (Str/2 + Con/2, 0 if negative))
  • Physical Avoid = ((Agility - (Wgt - Con, 0 if negative)) x 2 + Spirit) + Boosts
  • Magical Avoid = ((Agility - (Wgt - (Str/2 + Con/2, 0 if negative))) x 2 + Spirit) + Boosts
  • Dodge (crit evade) draws from Intellect.
  • Healing = (Staff Mt + Spirit/2)
  • Staff Accuracy = (Skill x 5 + Strength)

I explained the logic behind these changes later on in the topic.


luck is best stat and ur dum

So without making skl being too intrusive into calculations, here are 2 basic ideas that I had to increase the importance of skl by changing the mechanics:

  • Using skl as a different version of spd

You know how the doubling threshold is 4 AS? Imagine a different mechanic that has an activation threshold of 4 skl. When this mechanic activates, the unit doubles his str, or doubles his weapon’s MT (like the warrior weapons revealed in FE:if do on counters), or whatever.

Obviously this would be broken if stacked on top of the doubling system, so a workaround to this would be to give some classes the Pursuit skill and other classes the +atk-if-difference-of-4-skl skill.

  • Using skl for skills

Fire Emblem does this already, but skl is used for % activation, which is kind of dumb. Instead of % activation, imagine a skill meter mechanic that increments by skl points per turn or something and depletes when the unit chooses to activate a skill on command.

I’d rather not turn over all of the formulas and make them more complex by throwing in skl as a term. FE mechanics are elegant in their simplicity and one way to make skl useful (and spd less useful) is to make it actually do something other than affect accuracy. Imagine how much worse spd would be if it only affected avo. I’m fond of the first mechanic that I mentioned, but it could probably use some polishing.

My own idea was just using POW as a stat for magic and physical, and if a character becomes an OP hybrid that’s the designer’s fault. Remove LUK if you want, divvy its abilities into other stats if needed, but most importantly, create a new stat called SPI (spirit) to replace the hollow spot left by LUK’s removal.

SPI is used for skills in all forms, one way or another, with only a few rare exceptions. SPI determines the size of your mana or energy meter, which is used for advanced spells, or skills that can be used a couple times before the meter must refill.

High SPI means a bigger meter and faster growth. Assuming two paladins, both with Sol but one having 24 SPI vs the other’s 5 SPI, the 5 SPI guy can use Sol maybe once in a map, while the 24 SPI guy can use Sol twice, and he can even use it up to two more times before map end if he regenerates it properly or however the math works out.

Frankly, Yeti has convinced me of the folly of str/mag splits, but altering LUK to be more like SPI makes less sense than just swapping LUK out for SPI and adding LUK’s bonuses to other stats as Arch proposed.


1 Like

I get trying to show off what can be changed in a rom hack, but if it ain’t broke, don’t fix it.

If making speed less op in GBA FE is the goal here, then why not make skill items or something.

For example, a character holds a Luna scroll, which enables the Luna skill at a percentage of luck÷2.

Actually, skill items like this would be very useful in enabling stat balance, and bring skills into GBA FE.

or I’m just spouting more nonsense again

Did someone ring? :smiley:

1 Like

There’s only one formula that I’ve added Skill to, and that’s Attack Speed. Otherwise I’ve made various substitutions in the first proposal. Admittedly the second gets more complex by adding alternate physical/magical formulas for certain calculations. Other than that, I swapped the x 5 multiplier in Staff Hit to be skill based (but it’s still factoring in the same stats). It’s not like I’ve splattered Skill all over the game, it literally got one addition (which simultaneously helps balance Speed, as well).

Is there merit to “elegant simplicity”? Yes. But simplicity sometimes complicates the balancing act, too. I’m basically of the opinion that properly balancing an Str/Mag split would require creating some variety of alternate formulas for physical and magical calculations.

The numbers say it’s broke compared to what our changes have been able to do, comparing the standard deviations of their overall usefulness.


Knock down the base hit of everything by 20-25% and increase hit to be base+3*skl+luk

1 Like

literally the only part of this thread I will actually use in my future games

re: ‘ain’t broke don’t fix’, at least from my point of view it’s at least a little broken and could be fixed: think about how you feel when you get an Hp+Skl+Lck+Res level (“wow, thanks game none of these even do anything useful”), and then consider a Str+Spd+Def level (“10/10 would take basically every time, now I can 1rko that enemy I couldn’t before/can double enemies I couldn’t/get 3rko’d instead of 2rko’d”).
Obviously the worse stats are still ultimately helpful, sometimes you want a unit that can take hits from magic or negate enemy crit changes; but it’s very clear they’re significantly underpowered, and while it doesn’t have to be perfectly balanced there’s no reason not to think about making them a little more balanced.

Anyway for the sake of completeness here’s my take on it:

  • Physical Atk = Str
  • Magical/MWeapon Atk = Int
  • Healing = Wis
  • Physical Def = Def
  • Magical Def = Spi
  • Accuracy = Spi*1.5 + Str
  • Avoid = Wis*2 + Int
  • Critical = Wis/2
  • Dodge (CEva) = Spi
  • Attack Speed = Spd
  • Staff Accuracy = 30 + Int*5 + Str
  • Staff Avoid = Spi5 + Distance2

The goals in mind were to make Str and Mag both do something for the units that can’t use them for attacking, so a mage getting Str in a level up doesn’t feel like a complete waste, as well as nerfing Spd and using the stuff I took from it to buff other stats. Also, I tried to keep the formula about as simple as in IS FE, and just shuffle the pieces around, and actually managed to cut a stat entirely.

Strength is basically Str, but also gives a small amount of hit/attack staff hit that magic users can benefit from. Intelligence is Mag without the staff healing component, but with a small amount of avoid added that the physical units care about
Defense is just Def, that alone is good enough for a stat. Speed is just Attack Spd, also good enough same as Def.
Wisdom is a new stat, and basically a combination of Mag, Skl, and Spd. It boosts healing for staff users, is the source of ciritcal for combat units, and also provides the majority of avoid. Spirit is roughly Res and Lck, it provides resistance to magic and is the source of ciritcal avoid, while also providing the majority of hit.
Hp and Con are the same as in GBA FE.

I’d still like to give Str more use for magic units or Int more use for physical, but there’s not much I can add to them that wouldn’t just make them even better for the units that use them to attack (physical units already want Str that does nothing but boost their damage, making it also give hit just makes it more desirable)

Basically, I think of it as somewhat analogous to one of the Pokemon series’ changes; remember the “elegantly simple” old days, when the move’s type determined whether it drew from Atk or SPAtk (or Special, if you wanna get really OG). They added further distinctions to each move in Gen IV, and it made the game tons better.

Or you could do something like what Yeti proposes, scrapping Skill, Luck, and distributing things accordingly to new stats. I just don’t think mages really want more hit, so Strength seems still pretty useless and this was the best way I could think of to expand the scope of possibilities without making up entirely new mechanics to balance everything (which seems more complicated in and of itself than simply splitting the formulas).

This is a matter of contention. I don’t think the physical/special split made the game better at all.

It’s interesting that you bring up Pokemon, because it has the same problem of physical attack not doing anything for special attackers and special attack not doing anything for physical attackers, and because of the EV system, very few Pokemon can do a mixed moveset.

But I mean, this is one of those things where if you try to make str useful for magic users by conceding a tiny bonus, then it seems tacked on, and if you try to make str useful for magic users by making it actually matter, then it just doesn’t really make sense.


I would argue that it did make the game better; because you aren’t forced to run mixed sets/go without STAB on certain Pokemon (Feraligatr, for example) when their type doesn’t match their optimal attacking stat, which is an improvement generally speaking. You’re right about this, though: having two stats that do nothing but add attack value seems polarizing in this context too. Ideally, the physical/special split would have come with new mechanics to justify the stat’s use in builds of that prioritize the opposite attacking stat. That’s basically my approach here. It seems like there is too little to add for either Str/Mag, without splitting the formulas, to justify both stats for both types of units.

Hopefully this won’t be too lengthy of a digression.

From a main series gameplay point of view, sure, the physical/special split made sense, and it added some depth without adding too much complexity. From a competitive point of view, the physical/special split contributed greatly to power creep because not only did Pokemon have access to STAB moves of their better attacking stat (which is fine), but they also gained access to a bunch of coverage moves that also run off their better attacking stat.

It used to be that Pokemon were limited in what they could cover without going mixed. With the new mechanics, Pokemon no longer needed to go mixed to get better coverage than they could get before. It became harder to design teams that could prepare for all potential threats, and the remaining Pokemon that did have the luxury of going mixed while maintaining coverage in gen IV (Salamence, Garchomp) had to be banned.


Just add more magic users and take away access to Pure Water and Barrier. Thoughts?

I definitely understand what you’re talking about here; because Garchomp and Salamence defined the upper range of power, the series responded with power creep to make that relative power lesser than it was in Gen IV (instead of taking care to limit coverage options on the prioritize attack stat, or devise other benefits to raising the stats that improve the overall viability of mixed builds). Perhaps increasing the rarity of “prepared for anything” setups is ultimately for the better, but, as implemented, the change seems stark. It’s interesting, and I think there are lessons to be learned from these shortcomings.

That same sort of polarizing effect can be seen in Fire Emblem’s traditional implementation of the Str/Mag split. One response might be to buff weapons (i.e. offering the Forge for any weapon, or making weapons unbreakable/more powerful outright) so that the few hybrid classes get more power to compensate for their weaker stat’s performance to improve their viability. FE faces an opposite problem; hybrid classes aren’t utilized enough, but that’s also because they aren’t terribly viable naturally. Units spend T1 generally as either physical/magical, and then promote to a hybrid, thus having a deficiency in their other attacking stat that makes using their hybrid status extremely situational, and thus relatively useless.

When the attacking stats contribute nothing but power, they become too polarizing. This is the basic problem with Str/Mag split, and hybrid classes don’t justify its existence in the way that mixed setups in Pokemon once defined their meta-game. Another option could be simply to add more hybrid classes with higher base stats/promo gains, but that would only install a more equitable distribution to a flawed system (if Pokemon dedicated 7th gen to making a ton of really good mixed mons, it’d just be more power creep, and the balancing cycle would repeat to nerf the mixed builds, which now require a buff against plain builds, rinse, repeat, rinse again, and maybe they’d figure it out by Gen XIV, or maybe not–FE’s gone five games straight without addressing this at all).

I wanted to bring balance to the force, not leave it in ruin.

From the topic dedicated to sharing “more questionable design decisions”

Inflating the magic user distribution would mean you’d have to inflate Res scores to compensate for the increased commonness of magic (if it were reciprocated between the player and enemy, or if this applied to just the enemies, either way it requires a bump in Res growths/bases, and thus deflation in Res back to the same relative value you started with). Taking away the possibility of a temporary boost makes the stat more valuable though, but how often are those effects used? Maybe they could stand to be nerfed, but removing them doesn’t feel entirely constructive either.

Fire Emblem has this problem where it is very hard to balance different offense systems because ultimately you are playing against an AI and the enemies are generally worse than you. The strategic aspect isn’t the same as in Pokemon.

Suppose you have two units, physical unit and magical unit. They are pretty decent units, probably one is better than the other, because from what we’ve observed from Fire Emblem, either physical units dominate (most games) or magical units dominate (Awakening).

Next you introduce hybrid unit A. You think, in order to balance hybrid unit A, it’s going to exchange spd for the flexibility to attack either with a weapon or tome. Well, great, but hybrid unit A is worse than your existing units because it’s too slow.

Then you introduce hybrid unit B. Aha, hybrid unit B isn’t useless any more; I’ve given it more spd! Except now it’s better than your existing units because it can do the same thing as either of them. What’s more, either magic is going to dominate because it’s just intrinsically better, or weapons are going to dominate because the unit has a higher str stat.

So these are some very black-and-white scenarios that lack nuance, but the reason why hybrid units are less interesting than they seem on paper is because frequently they’ll be much better at attacking from one side than the other, at which point they’re basically like a non-hybrid unit except with a little trick.

Ooh, ooh, this is where I swoop in to answer these sorts of questions because I’m an amazing Fire Emblem player. Pure Water and Barrier are the only ways to prophylactically prevent status (besides, like, res rings), so they have a huge strategic role. Neither diminish the value of res; frequently you will need the use of such items on high-res units to reliably protect against status. Going from 2 to 9 res when your opponent has 18 mag isn’t going to help that much.

That’s basically my point, yes. As constructed, having two stats that only contribute offensive scores are too polarizing.

My “scrap Luck for Intellect stat” proposal seeks to address these issues. The root of the problem is that Strength and Magic, traditionally, contribute nothing to units who don’t use the stat for offense. Thus, hybrid functionality becomes nothing more than a neat trick.

One solution from IS was to use Str as a weight buffer in FE9/10; magic classes really want this, since their Constitution scores are typically lower. Mages would appreciate a low-growth weight buffer, but this consequently makes the stat too overpowered for physical units. Keeping Constitution as a constant for physical units, while splitting magic weight between Con and Str, allows us to have the best of both outcomes without the downside of either.

The way I’ve done it, splitting weight calculation into separate physical/magical formulas, gives added purpose to Str for only mages (physical units rely more on Str than mages do on Mag), and adding Dodge to Int gives that stat a multi-functional purpose. Unfortunately, the stats were seemed of relatively little value to classes prioritizing the opposite offensive stat.

Attack speed itself seemed too powerful; such that it either had to be an independent stat, or divided amongst two. Skill gets its necessary buff by splitting AS, while Speed gets a nerf. Two birds, one stone. However, that made Skill a little too good, relatively, so it seemed ideal to take away the x2 multiplier it gets in the accuracy calculation. With attack speed and avoid formulas already split into physical/magical, adding accuracy into the mix didn’t seem like too far of a stretch.

Spirit also factors into avoid in Luck’s place, which buffs the survivability of healers and mages to a lesser degree. It shouldn’t be a major change, since high Res classes typically got characters with high Luck growths, but consolidating that function helps make this stat more relevant for physical units (who, in most cases, would be better served trying to avoid magic attacks than tank them–Spirit helps them on both fronts).

This is, I suppose, the point where things get complicated. Mainly because I broke the “rule of two” (any component, ideally, should only be impacted by two stats at the most). Physical units get increased hit, crit evade, and magic weapon damage from the Intellect stat. Magical units get a 1 point of weight buffer for every 2 Strength, which in turn helps their avoid, and they get half of Str added to hit. Then, the opposite defensive stats for each factor into hit as well (part of good defense is consistent offense, after all); half of Spirit to physical hit, and half of Defense to magical hit.

Yes, I’ve broken the “rule,” but incorporating all four of Str/Int/Def/Spi, plus Skill, means that all of the main stats sans Speed are, at some point, factoring into hit. Perhaps this part is too complex; but if complications are necessary (which I think they are, in this case), asking a player to remember that their opposite offense/defense stats factor into hit is less cumbersome than many possible solutions.

However, I also understand that these would be significant changes in several respects. Part of the appeal to Fire Emblem’s handling of magic is that it functions exactly the same as any other weapon, and my proposal complicates things inherently by changing that fundamental fact. At the same time, there isn’t enough capacity with the system as constructed to make the Str/Mag split work, which necessitates an innovation like splitting formula into physical/magical variations.

I fail to see how Pure Water and Barrier don’t diminish the value of resistance. Both items cause you to take less damage from magical attacks. I think we can agree that if something temporarily gave you +7 defense that would be pretty powerful. Why doesn’t this logic extend to resistance?

If, on enemy phase, several magic users attack a unit, that unit could take quite a bit of damage, or they could make use of a resistance boosting item and take 7 less damage from each enemy. Less damage is less damage no matter how you slice it. And taking 7 less damage is rather significant in my book.

This seems like shaky logic to me. Just because there are more magic enemies does not automatically mean resistance bases/growths must go up. Beginning chapters typically have all physical units. Does that mean you must boost defense to match this? No. Some units just have terrible defense, and those with more defense are more valuable (assuming all other factors being equal).