World of Magic - Information Guide to all 23 Elemental Magics

Guides and Tutorials Arcane Odyssey Guides
121

True but the point isn’t necessarily to shift the order of the interactions, just to provide a more practical representation of how it could function ingame.

Yeah I’m not just gonna use that equation, I’m aiming more for the maximum damage output a single user of the magic could use, so it would take into consideration the fact some statuses that can’t be on the same target and go off a smaller amount of magics, as well as taking into consideration you only get 3 starter magics so I’ll most likely only use 3 magics rather than every possible status stacked onto one. This isn’t really to address the interactivity of a magic, rather the damage output that one could achieve with that. For interactivity I’d use something else entirely.

122

Revisiting this topic because I’ve started on my personal measurement (highest interaction combo from a single player without environmental influence), I am starting to be more curious about what measurements could be made for more accurately determining the exact interactivity of the magic, or in other words just a better alternative to SEDIT.

I think I really like what you used for RS SEDIT, where the total gets divided by the square root of the amount. Maybe the radical could be adjusted and rather than square rooting (^0.5) it could be another number used to root the amount, but regardless I think this is a valid way to reduce the insane impact that having multiple low-quality interactions creates.

I also think it would be valid to weigh the impact of ridiculously high multipliers, such as Gold which literally becomes one of the most interactive magics from just a single interaction with frozen, and rooting would seem like it would be valid in this case as well. For example, powering each multiplier by 0.956 would result in Gold’s 90% value becoming 74%, and a more moderate number like 30% would become 26%.

EDIT: I figured a better approach to this would probably be to divide the interaction increase by something like 30, root the answer, and then multiply 30 by the new answer. For example for Gold this would look like 90 ÷ 30 = 3 ----> 3 ^ 0.7 = 2.16 ----> 30 × 2.16 = 64.8. The numbers here could be adjusted but this would be the basic concept. This does make smaller values bigger but it at least makes sure numbers like 70-90 don’t exist.

In combination with the strategy you used for RS SEDIT, this would decrease the impact that a ridiculously large amount of mediocre interactions would have, but also decrease the impact that a single insane interaction would have.

More things would probably have to be weighed rather than just amount and strength but this would strike me as a more reasonable approach than the regular SEDIT.

Also it should be named from SEDIT because it just makes it sound unnecessarily complicated.

1 Like
123

I wouldn’t recommend going down that line - tweaking the root, adding more roots or adding extra multiplication/division steps is a form of Cherry Picking. You can torture the data to give you any result, and the more steps you add to “improve” the result the less you’re synthesising a useful statistic and the more you’re just making a tierlist.

Highest Interaction Combo is excellent in this regard - it’s very obvious exactly what it represents, and so it’s easy to see what its flaws are.
It doesn’t account for a magic’s option variety, but that’s ok because it’s not trying to, and that’s clear. Standard SEDIT is good at this for the same reason.

RS SEDIT is a weird middle ground. It’s better than SEDIT at being a “this is how good a magic’s synergy is”, but that’s not what it actually represents. RS SEDIT is the sum total of interactions, weighted by their squares and a couple of 0.5x conditions.
I think it’s simple enough to be ok, but I could see why someone might disagree - it definitely straddles the line of being a magic number instead of something clear that can be understood.

^0.92^0.7^(1/0.7) SEDIT doesn’t really represent anything.
Perhaps it’s a flawless rating for a magic’s interaction potential, but we’ll never know. The system is too complex to judge the desirability of its behaviour - this means it’s impossible to compare it to some other system (say, ^0.94^0.8^(1/0.8) SEDIT) without just saying “this version better fits my subjective view of how interactive the magics are”. It’s little more than a tierlist.

Your issue with gold’s 90% frozen multiplier is a good example here. The fact we’re trying to reduce the impact of gold’s 90% interaction without affecting the rest of the table too much is a good indication we’re tierlisting instead of making a better system.

What exactly is wrong with a 90% interaction? If Shadow had a +90% Soaked interaction, it’d easily be the best synergy payoff magic in the game - it’d certainly deserve to be ranked as one of the most interactive magics.
The real issue with gold’s 90% getting full weighting isn’t that it’s 90%, it’s that it’s a clearing interaction on Frozen - a status which is very hard to apply (combination status interactions only) and very brief (all opponents can clear it within a second).

Rather than reducing the 90% interaction just because it’s 90%, we should reduce it because it’s wildly impractical. How exactly we weight “this interaction is impractical” is a difficult question to answer, but figuring out simple and consistent rules is the game here.

Honestly I figure “0.5x for combo status, 0.5x for clears” is pretty good already. Gold’s frozen interaction is only worth a ~22% interaction in my RS SEDIT calculations - gold is ranked high thanks to its two repeatable 30%s.

124
Note: You should probably just skip this hidden info section because I feel like what I addressed in this single quote below is much more practical than everything else that was written and that everything else is probably more of just what I think about the direction we could take this system (and also because reading into this quote and looking at the chart I start to see that some of what I said is much less relevant). However, I feel like what I said still has some degree of relevancy so if you want you can open this point but for the most part it's much less practical and much more of just an essay you can bore yourself reading if you have time to burn and a life to waste.

That wouldn’t be cherry picking because the results aren’t necessarily being suppressed. Rather than just trashing this point however, I’ll replace “cherry picking” with “disingenuous,” “bad data” or some other term that I feel would be more accurate to make sure your point is not lost within semantics. Anyways, keep in mind SEDIT isn’t a set stat. SEDIT is something that we have made up in order to describe interactivity, and so we can change that in order to be more and more accurate with what we want personally. Yes it would be much better if it were a more of a practical stat than an algorithm but the issue with this is that as soon as you try to change this into a relatively simple equation it just loses so much value that it becomes much more of a useless number than an actual ranking.

tweaking the root

Also the root wasn’t even something essential to the equation, this is just something that we’ve added onto it to make it more accurate, so in a way that’s already doing what you’re describing, which I think is fair.

The problem is that the stat isn’t useful. It’s being synthesized to become more useful because the stat in it of itself basically means nothing. Also along the line of making a tier list let’s divide the semantics from the point again. I don’t think that the comparison to a tier list is necessarily wrong but describing it as a tier list implies that there is no mathematical process in order to determine the results, which there is. The idea that it becomes more and more subjective however, is what you are trying to address, and I think that’s a fair point, however, I feel like the problem with keeping it as a very easy to calculate statistic is that it just becomes more and more useless, and if we were to just say “well this magic has good interactions and this magic has bad interactions” I would honestly say would likely be more useful in this case, however now it’s now devoid of any logical processes used to determine the data and is entirely based off of “how do we feel about this?” What I’m doing here with the measurement I described is sort of doing that to an extent, but it’s still an algorithm used in order to rank them rather than a bunch of statements that a person makes, which is though not as valuable as raw data, is still more valuable than a statement, and in a situation like this where raw data isn’t really that valuable, in my opinion we’re left with no choice than to make it more of a “mathematical judgement” than an empirical measurement.

The issue with this is that by keeping the 90% figure the same we’re changing this to be less of “how interactive is this magic” and more of “how effective are these interactions.” This is now limiting something else someone would want to find with a stat measuring interactivity, which would be how many interactions does this magic actually have. With RS SEDIT it becomes much closer to “how effective are said interactions” which is incredibly harder to judge based on a simple equation, and yes with what I’m doing here it does become more subjective but when you try to make a judgement off of pure, cold hard raw stats rather than trying to judge it from a more human perspective it becomes very inaccurate for what someone is trying to achieve.

I feel like one of the main issues here that we're probably most conflicted on is what should be weighed as more effective, the practicality of the interactivity, or that in combination with the quantity of interactivity, so I feel like that's the main confliction here. The best way would be to isolate that into multiple stats, which I would 100% agree with if not for the fact that we're literally overflowing the amount of stats we have for one single piece of information. Yes it would be more practical to have each thing divided into different stats, but when we have

RS SEDIT POSITIVE
RS SEDIT NEGATIVE
RS SEDIT COMBINED
INTERACTION COUNT
INTERACTIONS COUNT MODIFIED
Combo Boost

and then not only that, but added onto the already existing

Damage/Damage w/ DoT
Speed
Destruction
Destruction × Damage
Clash Average
Cleared Statuses
Conversions

for a total of

Damage/Damage w/ DoT
Speed
Destruction
Destruction × Damage
Clash Average
Cleared Statuses
Conversions
RS SEDIT POSITIVE
RS SEDIT NEGATIVE
RS SEDIT COMBINED
INTERACTION COUNT
INTERACTIONS COUNT MODIFIED
Combo Boost

it just becomes an absolutely overwhelming amount of information, especially when so much of that is literally to describe one thing, so I think that at the point that we have so many measurements it just becomes more obstructive to the regular person reading “hmmm what magic should I pick” which just blows. If there was a way to accurately describe each piece of information without having to overwhelm the reader, or, anyone operating the data, then I would absolutely separate everything into separate stats, but then when there are already so much it just seems more useful to use a rather subjective equation in order to get your results rather than a hard-on “we added up these results without changing literally anything and this is what we got”


It’s not even just about readability too, I find it more pleasing to just have just a measurement for “interactivity” rather than having them all separated into different categories. It just makes things more simple and easy to deal with.

So summing it up I think our major conflictions come down to simplicity vs complexity and objectivity vs subjectivity, and basically that's just where we are seeing disagreements. I think we perfectly agree on what exactly is being done here and what exactly are the issues here, but we just don't see eye-to-eye on how those issues should be weighed.

Now, onto the more practical stuff (this is an edit I didn’t originally plan to just spew out everything and leave the most practical points for last)

I saved this point for last because I wanted to read over exactly what you said before addressing this head on, I do think that taking a look at the RS SEDIT it doesn’t really seem too far off from what I would determine the direction I would want to go, so perharps I would work with rather than something more complex. I do still disagree with Gold being third from literally 3 good interactions and one mediocre one, but other than that it doesn’t actually seem too bad from how I would probably rank them.

One thing that might be bothering me about this though is how this does only take into consideration how the magic interacts with other magics, and not how other magics interact with it, but I feel like this would be fairly simple to solve without too much “skewing.”
Yes the actual document lists the SEDITS for each status effect but I would like to see this represented within a single measurement, if not, added onto each magic rather than just being listed for the status effects. It makes it more convenient for analyzing the magic itself rather than having to go look at the statuses as well.

125

bro when i opened this document my entire computer lagged to death

1 Like
126

The Magma, Plasma, Fire and Explosion magics don’t mention their interaction with the Snowy status effect (-5% damage for Magma/Plasma, -10% damage for Fire/Explosion) in their Magic Damage Interactions section

127

man sorry for the bump but

change wom survey to ao survey

128

I plan on changing it when TGR is released because the main game is still WoM.

129

alright

130

i found a small mistake

“All elemental magics share the same cast times, size, and magic energy costs.”
first paragraph, size affinity exists now so maybe Meta/Fluect forgot to edit this

1 Like
131

I don’t think Meta is updating this until TGR, but that is probably something that will be fixed.

1 Like
132

This guide will not be updated until Arcane Odyssey releases. There will be an entire new document posted once the game releases.

2 Likes
133

Art of purity when

134

Art of what?

135

There’s a secret magic called art of purity at the end you just need to highlight it

Screenshot_20210415-201551_Chrome
Screenshot_20210415-201551_Chrome1920×1200 212 KB

Screenshot_20210415-201545_Chrome
Screenshot_20210415-201545_Chrome1920×1200 278 KB

137

@Meta why isn’t Ash’s dot over there idk if it’s a stupid question.

138

Wym, like dot the word or DoT/damage over time? Ash doesn’t have a damage over time it has clouds.

If you mean dot as in like an actual dot than idk what you are referring to?

139

the ash clouds aren’t considered DoT? oh ya it’s just how long you stay in the clouds mb

140

image
is the ash cloud damage also included in the magic damage tho

141

Nah we only include the multipliers with the DoT at the side. Clouds don’t consistently deal a tick every time a blast hits so I don’t think it makes sense to include them.

Also there would literally be 0 way to include cloud damage reliably since they follow a different formula from all other damage, you can’t represent that as a multiplier or anything really other than an equation.