Theorycrafting! (Because I like competitive metagames too much)

[SnowFire]

Djinn: Hrmm, I'll pass.  This is the *Theorycrafting* topic after all.  And I imagine MC has played vanilla FFT so much that it'd require much wackier stipulations to be interesting for her, or at least LFT or the like.  And on that note, if somebody was *actually playing*, I can't imagine denying anybody Gained JP Up as just a quality-of-life deal to make things more fun, even if just in C1/C2.  Similar to how FE7/FE8 drafts usually let everybody have Marcus or Seth for the first half of the game.

Alex: I can't speak for MC, but I think I'm just looking at this from different angles re: Chemist/Priest/Monk.  To be sure, you're correct that "You don't need Auto-Potion and Phoenix Down when every fight gets one-rounded", but I guess I personally am assuming something more like "a smart person's 2nd trip through FFT" rather than "somebody who does SCCs" for general competency level.  After all, if you're good enough to do SCCs, then it's hard to assess balance since you're expected to win anyway with 9 classes rather than 1.  In the same manner, a lot of easier-than-FFT RPGs often have powerful healing that isn't actually super-necessary if you really know what you're doing, especially in challenge / speed runs.  Yet I'd still rank such healing pretty highly, since if you're not crushing everything immediately, you can simply fall back on never dying instead.  FF5 White Mage was considered top-tier in the in-game rankings thread for example despite having no offense for a loooong time.  FF7 Restore Materia is among the best materia to equip despite most enemies dealing no damage, and the enemies that do deal damage usually being spoilable with elemental resistance if you can set up for it.  (In the same way, you don't need maximum-broken Ninjas or max-broken TMs for the classes to still be really useful and strong.)

I will agree that any of MC's drafts with Summoner (and no Calc for the opposition) probably favors the Summoner team, because Summoner is just plain one of the most dominant classes out there, with a Summoner SCC probably being on par with "your choice of your least favorite 14 classes".  Summoners don't even *need* MAU or Short Charge to be good.  That's why I mentioned that with MC's ruleset, I'd be tempted to run 3-5 Summoners regardless of whatever else I drafted if I somehow grabbed Summoner.  (My 3rd playthrough of the game - without reading FFTSB or GameFAQs or anything - was a "-1 deploy limit" run, and I had something like Summone/Time Mage, Summoner/Priest, Summoner/Oracle, and Ramza as Ubersquire/Geo w/ White Magic.  Yeah, 3/4 summons is good times.)

[Dark Holy Elf]

Really enjoyed reading your analyses, Alex.

I do agree with Snowfire's points though, overall, re the value of good healing. In particular, Mediator certainly isn't better than Chemist. Equip Gun is not better than Auto-Potion, and Invitation(/Talk Skill) is not better than Phoenix Down(/Item). That's honestly one of the easiest comparisons out there.

[Ranmilia]

Oh, to be clear, there, the only things I'm valuing from Mediator at all are Praise and Preach.  The job description reads "a small investment in this class gives your mages (and/or monks and ninjas) a permanent passive damage bonus, and makes your reaction abilities trigger more often."  Is +~30% damage on your Ramuh or Black better than Auto-Potion?  Yeah, absolutely, that enables a lot of quick kills that wouldn't otherwise be possible.  On Meteor... hm, I guess there is the argument that Meteors are going to OHKO or overlap 2HKO anyway, so the damage boost may be overkill if you're on Time Mage or TM/Priest as your only casters?  That does lower it somewhat.

And all of this is coming from a super Spike type of approach where it's assumed these two teams are played skillfully through the vanilla game and are competing with each other on... some metrics, of which I'm gravitating towards things like "no resets" and "fewest ATs to win" as being important.  There's a ton of stuff you can fall back on to never lose.  Mediator+Monk is an amusing example, 100% Gun Hamedo and maxed Martial Arts are pretty nice.  Move MP Up MP Switch Ninjas are hard to kill.  The Quickening exists, if it isn't banned.  But clearly no one is considering setups like that to be seriously competitive, even MC's drafts completely discount them in favor of "just kill things with magic!"  Not losing is not the same as winning, in a competitive scenario.  Summoners are never going to reset either, because all the enemies are going to be dead. 

I guess it comes down to expectations of player skill?  Snowfire says that high skill players are hard to consider balance for, while I'm operating on the opposite and feel that high skill players are the only ones who balance *can* be adequately considered for.

The 4JF format does change things up.  That requires a lot more thought and I'll probably tackle it later tonight!

[DjinnAndTonic]

Djinn: Similar to how FE7/FE8 drafts usually let everybody have Marcus or Seth for the first half of the game.

Wait what. People do Fire Emblem draft runs?!

[hinode]

In my experience four Summon Magic users+1 guy with item results in slightly less resets on average than a full team of Summoners. You usually have enough offense to faceroll anything either way, but the one guy with Phoenix Down usually gives you a healthy margin of error in case you accidentally let someone get KO'd by a stray ninja or whatever. Later in the game you'll also have ethers in case you REALLY need the extra MP for something.

I haven't actually used math skill in practice, but I'd imagine that 5 Mathskillers is gross overkill in literally every situation, even moreso than 5 Wizards w/ Short Charge 108 Gem Leviathan.

I guess it depend on how much you value the safety nets for resets vs absolute minimum turncounts when evaluating classes. Also whether we are assuming absolute perfect play on every single turn or allowing for some leeway on minor mistakes, like miscalculating enemy attack range once in a while. Personally, I just don't quite have enough FFT experience to handle every single circumstance when going all-in on mage offense, so I do value safety nets like Phoenix Down and Auto-Potion fairly highly.

Edit: Djinn, there's an entire subforum dedicated to them on Serenes Forest.

[SnowFire]

I guess it comes down to expectations of player skill?  Snowfire says that high skill players are hard to consider balance for, while I'm operating on the opposite and feel that high skill players are the only ones who balance *can* be adequately considered for.

The thing is, if "few resets" are the main metric, then past a certain point really good players all have 0 resets.  It's like 2 students getting a perfect score on the test: you can check things like "how fast did you finish the test", "did this hype require weird or nonintuitive bug/AI exploits",  and "were you forced to go grind up rare steals / poaches to study for the test to really guarantee a perfect score", but it's a little arbitrary.  I'd argue hard-focusing offense can help with the tiebreaker issues like "number of turns used" assuming elite players, but also potentially lead to resets and & more frequent crystalizations if you mess up.

[Dark Holy Elf]

Oh, to be clear, there, the only things I'm valuing from Mediator at all are Praise and Preach.  The job description reads "a small investment in this class gives your mages (and/or monks and ninjas) a permanent passive damage bonus, and makes your reaction abilities trigger more often."  Is +~30% damage on your Ramuh or Black better than Auto-Potion?  Yeah, absolutely, that enables a lot of quick kills that wouldn't otherwise be possible.  On Meteor... hm, I guess there is the argument that Meteors are going to OHKO or overlap 2HKO anyway, so the damage boost may be overkill if you're on Time Mage or TM/Priest as your only casters?  That does lower it somewhat.

I would first argue that Praise deserves little to no hype, as Cheer Up does a better job (maybe not relevant for draft theorycraft which arbitrarily pretends Ubersquire doesn't exist, but extremely relevant to how I'd score the jobs in reality). Still, even if you respect it... it and Preach really aren't great. e.g. to get a 30% damage boost to Ramuh, you'll need to land Preach 21 times... which given its accuracy probably means you'll have to use it at least 30 times, and that's assuming you set up for good internal zodiac. When are you going to do that? It really depends how we're scoring people here, but you yourself suggested "fewest ATs" as a possible candidate and it's obvious this is a losing strategy under such a format.

By comparison, Auto-Potion is just "learn it, now you have roughly triple the durability". (using mc's figures here) You can quibble her numbers, maybe it's only closer to double if you're being 2HKOed for instance. But double durability makes a decent case for itself against +30% damage (I suspect you'll argue it loses, but it shouldn't be a slaughter), and 400 Chemist JP plus some potion money is far easier to come by than 200 Mediator JP plus over 30 turns spent on actions which do almost nothing to help you win the fights they're used in.


I think we all agree that zero resets is a desirable outcome, and Chemist's tricks do go an awfully long way towards ensuring that (I agree with hinode that 4 Summoners + 1 Chemist would outperform a Summoner SCC.) They don't help much with some other metrics, such as speed (Auto-Potion does help decently, just not as much as Summon/Math/Meteor/etc.), so I'm certainly open to them being anywhere from like, #3 to #5 or so. It does depend somewhat on how you weight various aspects and what level of player skill you assume.

[metroid composite]

With Calc in the mix, there are now three battlefield nuke classes so they're going to be split.  This makes them less important to ban, but also means Team 2 has to be very, very on point with what they plan to do.  I can already say in their shoes I would not ban Squire, I would want to first pick Squire + nuke class.  When your competition is Math, you aren't winning the long game, so getting out to a strong early lead is important.  Gained JP Up does that better than Item, IMO.  If Squire is banned, then Summon + Time at least gives you Short Charge summons to work to, that's the best thing you can be doing without Math or Wizardry.  Chemist just does not make sense to me.

The thinking here was as follows:

1. Ramuh costs 200 JP.  You will get there just fine without Gained JP Up.
2. If you don't ban squire you have to pick it immediately, making your immediate pick Summoner/Squire.  There is no scenario where you get Summoner/Time Mage unless you give Squire or Chemist to the Calc team.
3. Phoenix Down is literally free, and a very good skill to have.  Auto Potion is very good and costs 400 JP--very much affordable without Gained JP Up.  Picking Chemist is a big part of the "dominating the earlygame" plan.
4. So...like in the "take Summoner/Squire" plan...is the ban Chemist?  Cause if you ban Chemist, the Calc team takes Priest/Time Mage, and now you have to pick up Monk if you want revival.

Picking Chemist early comes at a significant cost in offensive team power.  I feel like all of these drafts are taking it way too highly.  You don't need Auto-Potion and Phoenix Down when every fight gets one-rounded, which is honestly what should happen in a draft format with skilled players and only one ban per team.

Mmm...for what it's worth I was imagining that you would do something to make the battles not complete walkovers (like...low level caps, three party members, I don't know...something).

So Team 2 gets the better of this one, cleanly.  Team 1 feels all over the place.  What is their plan?  Holy and weak classes until Samurai comes online?  Let's see what MC said.  Dance?  Dance + Holy?  Is... is that even better than Wizard SCC, even when fully set up?  And Team 2 has all that other stuff?  This may be less close than the Math draft.

For what it's worth, team 1 put a lot of effort into making sure team 2 didn't have revival, but yes, this did screw them in offence.

Ok, it took a ban to do it but this looks more like what I was envisioning at the outset, with an even split of the mage classes and then a fight for support.  Monk/Ninja looks like a suspect pick here.  Is Team 2 really going to actually use those physical jobs over potential picks of Oracle, Mediator and Samurai?  Revive was mentioned, but really, who's ever learning and setting that?  Oracle at least would give Move MP Switch and sticks, if you're worried about running into situations where magic is somehow bad.

The thought was a mixed party, instead of an all-mage party, where you do run one Monk.  This also has the side-effect of probably getting HP restore for the entire party through spillover JP (in a party short on reactions).  Also, Chakra for MP restoration, which is relevant in a party whose offence plan is Short Charge Meteor.

I agree that there is a realistic argument for not moving in on Monk here, going without revival, and picking the stronger classes.

Oh, also, Bard going dead last when magic is in play, that probably shouldn't happen.  Ramza is still a forced male PC, Bard is still an acceptable-ish job for a mage to be in, and Move+3 isn't that big of a dip compared to some of the stuff teams have seriously considered.

Wait, wait wait, you're trashing Chemist because "fights should be over in one round" -- which ok, I wasn't intending for that to be the flavour of the draft, but I follow your logic.  But now you're...praising Bard?  What?  Bard is literally the slowest, grindiest class that ever existed.  I have some respect for Bards, but hell, Mimes end fights faster than Bards do.

"Bard is still an acceptable-ish job for a mage to be in"?  I hope you don't actually mean the class itself, because umm...50 MPM.  Tied for the lowest MP in the game (and the lowest MP growth in the game).  Can't equip robes, just in case you were planning to use your equipment to give them some MP.  Less HP than every single mage class (including Calculator).  Admitedly, their MA multiplier isn't bad; it's like...Priest level.  Except they have basically no equipment that boosts MA.  No robes.  No weapons that give MA.  No shield slot for Aegis Shield.  Their hat and their accessory is all they have to work with for MA.  I'd wager a guess that once Rune Blades become available, Knights have more MA than Bards.  *runs some calculations* yes, yes they do.  Often tied even before Rune Blades.  (Or ahead, actually, if you use C Bags).

17 MP at level 30!!!! (Admitedly, that's with bard stat growths; 27 if you have mage stat growths the whole way).  And again, can't boost that with robes.  Can boost it with hats.

Time Mage is a very strong pick to me, roughly equal to Wizard and Summoner.  My ideal composition is (strong lategame MT magic) + (strong magic support) + (strong earlygame carry). Time Mage, Summoner and Wizard each provide 2/3 of the puzzle.  TM lacks earlygame, Summoner lacks support, and Wizard lacks lategame.

Yeah, this is a fair point, and I think Time Mage drifted up a bit as I thought of this as more of a draft format.

Cause in terms of abstract class quality, Time Mage definitely hurts from competition.  Magic Attack Up is relevant for all four big offense options.  Short Charge is irrelevant for Math Skill and Draw Out.  Short Charge has twice the MP cost, and to really make use of it you need to pick up the 900 JP cost summons instead of the 200 JP cost summons that go with MAU.

That said, I think Time Mage also suffers from the very specific format restrictions--namely, this is a class where you want 3000+ JP.  You really want Squire, and you probably won't get Squire with it.  It's really awkward to get Squire with Time Mage.  Like...first pick Time Mage, and let your opponent get Wizard+Summoner--that's one way to get Squire.  Another way is to be team 2 and first pick Time Mage+Squire, and you know, let your opponent have Wizard or whatever.  And you're not getting Time Mage any later than that unless Calc isn't banned.

General thoughts: Drawing conclusions from these drafts runs a danger of circular logic, where we as drafters pick a thing high or low because we already think it's good/bad and then look at the results and go "thing got drafted high/low, it must be good/bad!"

Oh absolutely agree with this...but even with that being said, I feel like I learned some things.  Like...Squire went later than where I had put it on the tier list.  Chemist ended up getting picked higher and earlier (even though, granted, you feel this is the wrong direction).

General eyeballing job strengths in ban format:
Calc >> Summoner > Wizard > Time Mage > Squire > Priest > (Ninja and Samurai if you are moving in on them) > Mediator > Chemist > Oracle > Geomancer > Thief > Monk > Bard > Lancer > Archer > (Ninja and Samurai if you are not moving in) > Knight > Dancer > Mime

I guess I'm pretty skeptical of Priest's positioning.  Like...you value damage, and consider healing pretty bad, that's fine, but this means Priest is mostly where they are due to Holy.  Pretty sure Ninja is quite a bit better at damage than Priest.  (In fact, last I checked Ninja was considered the easier SCC, despite the fact that Ninjas are all about killing, whereas Priests do a lot of stuff).

I guess I'm also not sure why Geomancers are where they are.  The highest damage setups for mages?  Basically never Geomancer.  The highest damage setups for physical classes?  Not in Geomancer either.  Attack Up?  I think it's less damage than Martial Arts for Ninja if you're going that route?  Not 100% sure on that.  (Reminder: Attack Up doesn't boost the damage of Throw).

I also think your Lancer placement feels like...well maybe it needs a split the way Ninja/Samurai have two ratings.  Lancer is a class you have to dedicate a lot of training to to get any value out of it.  If you're not going to have a dedicated Lancer, then Lancer does almost nothing for you.  If you are going to have a dedicated lancer, sure, you can deal lots of singletarget damage.  Like...if you're not planning on having anyone in Lancer, it certainly does less than Knight (Knight has some decent inexpensive R/S/M abilities).  If you are planning on having a dedicated Lancer, then it does plenty (and certainly helps your offence more than Bard).

[metroid composite]

Queklain is Level 20.


To be clear, I think 15 for Orbonne is reasonable enough, it was 15 for Riovanes that I was questioning (beyond the lowest-level classes).


That said, no way is Time Mage 13 for Velius. Time Mage needs to hit Level 12 just to start casting Meteor, something they do in chapter 2. (Everything before Meteor involves lots of Haste and staff whacks so they actually gain large amounts of Exp early). Now, they finish battles in such a small number of actions after Meteor that they probably aren't 18 by the end of C3, that I'll grant, but they should be ~16 just based on getting one level every four battles post-Meteor (which sounds right to me).

Samurai/Squire I'm both really skeptical about and I'm certain I wasn't that low myself (despite not grinding), but it's not outside the realm of possibility. Keep in mind that if I recall your playstyle right, you tended to reset if you got in randoms so right there that's a fair bit of potential lost Exp, especially for Samurai which will absolutely backtrack for Kiyomori post-Yardow thus having 7 potential randoms (~2.3 on average? They're 1/3 chance, right?) in chapter 3.


Hmm, that makes me want to put this together...


List of potential randoms, for reference:
Chapter 1:
5 unavoidable (Mandalia, Sweegy, Mandalia, Mandalia, Mandalia)
1 if we backtrack for Silk Robe/Chain Vest
2 if, in addition to the above, we backtrack for Mithril Sword, Silver Bow, heavy equip upgrades, etc., after Miluda

Chapter 2:
4 if we backtrack for the first spear and/or the Ice Bow
3 if we backtrack for the Triangle Hat, Rainbow Staff, Power Wrist, and/or sticks instead of waiting for after Zaland
4 if, in addition to the above, we backtrack for Coral Swords before Zaland and Bariaus Hill [very unlikely]
2 if we backtrack for Green Berets and Wizard Robes before Zigolis and Goug
4 if, in addition to the above, we backtrack for katanas and balls
1 if we backtrack for Brigandine, Wizard Staff, and/or Bizen Boat
1 unavoidable (Bariaus Valley)
10 if we backtrack for Cross Helmets [extremely unlikly!]

Chapter 3:
4 if we backtrack for various new weapons at chapter start (sword, spear, bow) instead of waiting until after Goland
3 unavoidable (Zeklaus x3)
8 if we backtrack for Mithril Guns [unlikely]
2 if we backtrack for Platinum Sword and +30 heavy armour HP after Yardow
2 if, in addition to the above, we backtrack for Ninja Edge and Kiyomori

Chapter 4:
12 unavoidable (Yuguo, Grog, Finath, Bed, Finath, Dolbodar, Zirekile, Araguay, Sweegy, Mandalia, Mandalia, Sweegy)
Plus a bunch more if we do the Beowulf sidequest, don't feel like counting that up.


I think how much backtracking one does is open to player choice, though I'd typically rather not do any backtracks that involve 5+ battles at least, and a few more in there are certainly questionable too. Some obviously depend on the party contribution; the first backtrack of chapter 2 is a no-brainer if you plan on running a Lancer in the first half of the chapter, but if you aren't you will probably skip it.

So anyway, digging like two pages back in the topic to this...used this in a new set of calculations.  Also upped the expected number of actions for most fights to 5 instead of 3.  (Because my initial calculations, which assumed 3 actions per fight, was returning values people said were low; also I watched a random let's play, and 5 actions in a fight happened on Barius Hill).

Here's what I got

End of chapter 1: level 6

End of chapter 2: level 14

End of Chapter 3: level 22

Sluice: level 27

Before final dungeon: 35

Going into Altima: level 38


I will say, these numbers look kind-of high to me.  Although...I think my Geomancer SCC might have ended up around that level, just because they killed people with elemental so it typically took quite a few actions to kill enemies.



Obviously if we go by Alex's "battles are over in one round max" then these numbers are horribly, horribly wrong.  Numbers for that would be more like

Chapter 1: level 2 (can confirm, have beaten Chapter 1 at level 2)
Chapter 2: level 6
Chapter 3: level 10
Sluice: level 13
Going into Altima: level 20

[Ranmilia]

Ah, my rust is showing in regards to Bard, then.  I just skimmed the BMG and only noticed that its MA was okay.  Also I forgot Cheer Up existed.  >_> 

Geo and Lancer... I didn't put too much thought into their placements, no.  Geo depends really heavily on what the rest of the team is, and Lancer I was considering far enough down the list of things players would main as to be irrelevant, and only thinking of them as "maybe you have no reaction and somehow want to get Dragon Spirit?"   That's probably LFT influence seeping in though - Dragon Spirit is really 560 JP?!  I think there are enough spare actions to make faith raising at least somewhat useful, but maybe not that much...

Priest that high is due to synergy with other mage jobs, considering that magic is usually going to be the main strategy of choice.  If your options are just Priest vs Ninja in a vacuum, Ninja's better, but if you're looking for "second class to support Wizard, Time Mage or Summoner" then there's no real reason to give Ninja the time of day - Priest can help your earlygame while remaining relevant later.

Level caps or fewer PCs definitely change the game significantly, and make Chemist much more desirable!  I do feel like I've talked Chemist down a bit more than it deserved, now.  It does significantly help the earlygame.  A team lacking revival is just not a problem to me, though, as long as their offensive capacity is sufficient. 

The thing is, if "few resets" are the main metric, then past a certain point really good players all have 0 resets.  It's like 2 students getting a perfect score on the test: you can check things like "how fast did you finish the test", "did this hype require weird or nonintuitive bug/AI exploits",  and "were you forced to go grind up rare steals / poaches to study for the test to really guarantee a perfect score", but it's a little arbitrary.  I'd argue hard-focusing offense can help with the tiebreaker issues like "number of turns used" assuming elite players, but also potentially lead to resets and & more frequent crystalizations if you mess up.

Basically this, except I think this is vastly underrating how powerful FFT offense is and how easy it is to have 0 resets, which makes "how fast" the real competitive measurement in question to me.  Like... resets are kinda rare in the good SCCs to begin with.  Summoner + Item is certainly better than Summoner + nothing, but when you're adding not just two jobs together, but half the jobs in the game, with some choice allowed?  That's overkill, without restrictions of some sort.  Something has gone horribly wrong if a reset is ever in the picture. 

Uh, anyhow, sorry, didn't mean to be all contentious!

Edit since new post - as far as character building goes, I do tend to assume that some amount of JP gain is frontloaded and done in chapter 1.  Not necessarily "frog an enemy and grind your full skillset" but "poke your teammates with rods in between Delita murdering the enemy at a moderate pace."  The specifics are hard to work out without hashing things like "how do you get through 'banned' jobs in a draft anyway?"   One rounding fights starts in mid chapter 2, except for Summoner, and even then is somewhat of an exaggeration.

[metroid composite]

I will agree that any of MC's drafts with Summoner (and no Calc for the opposition) probably favors the Summoner team, because Summoner is just plain one of the most dominant classes out there, with a Summoner SCC probably being on par with "your choice of your least favorite 14 classes".  Summoners don't even *need* MAU or Short Charge to be good.  That's why I mentioned that with MC's ruleset, I'd be tempted to run 3-5 Summoners regardless of whatever else I drafted if I somehow grabbed Summoner.  (My 3rd playthrough of the game - without reading FFTSB or GameFAQs or anything - was a "-1 deploy limit" run, and I had something like Summone/Time Mage, Summoner/Priest, Summoner/Oracle, and Ramza as Ubersquire/Geo w/ White Magic.  Yeah, 3/4 summons is good times.)

While Summon is very very good, and I'm sold at this point on it being top tier outside of Mathskill...

Like...Summoner SCC is not considered the easiest SCC.  SCC easiness rankings, if I recall, go something like...

1) Chemist
2) Time Mage
3) Ninja/Priest/Summoner (I actually forget which of these are considered easier, but I think Summoner was considered slightly harder)

Summoner generally needs some support in the lategame (either Short Charge or Magic Attack Up).

The Quickening exists, if it isn't banned.

Oh, OK, one sec--Important rule that I hadn't mentioned:

Full team coordination isn't being considered

What I mean is like...picture a Vancouver co-op style playthrough, where you control one character, and all levelling decisions for them, and 1-4 other players of varying skill also control one character.  Pass the controller around multiplayer.  You can assume your teammates are competent.  But you can't assume a specific reaction/faith/zodiac/whatever from your teammate.  Even if you tell them to learn Critical Quick, they might be like "nah, don't feel like unlocking that right now."  You also might need to carry your team a little.

So...good luck trying to herd all the cats necessary to make The Quickening happen.  But even outside of The Quickening...an argument I've seen on the FFT board that I don't take too seriously is that "Raise has a 100% hitrate because you will have a flawless internal compatibility triangle and 70 faith on all your characters."  Yeah...in a Vancouver style co-op game that won't be the case; not for every character, at least.

Now, you might, might be able to talk someone into the second best setup after the Quickening (Countermagic Quick) since that only requires two people working together.  It hasn't happened, but it could...maybe.  But it still requires a lot of assumptions.  Quick's base hitrate is trash, so probably best internal compatibility.  97 brave is a must.  Enough MP restoration per Quick is a must (and that takes like...240 MP, not a statline you'll see at normal levels).  It requires someone learning Quick at all (800 JP ability) and someone else learning Countermagic at all (800 JP ability). 

[metroid composite]

Priest that high is due to synergy with other mage jobs, considering that magic is usually going to be the main strategy of choice.  If your options are just Priest vs Ninja in a vacuum, Ninja's better, but if you're looking for "second class to support Wizard, Time Mage or Summoner" then there's no real reason to give Ninja the time of day - Priest can help your earlygame while remaining relevant later.

So...here's the thing--I often end up feeling that Item has better synergy with mages than White Magic does (for healing).  Item is something you can do when you run out of MP.  Item can restore MP.  Item is instant.  These are all properties that are very good on a skillset to give to a mage.

Black Mage with Item is straight up the best setup for most of Chapter 1, and there's some debate as to whether Summon even beats that in Chapter 1 (since Summon is out of MP after one spell in Chapter 1).

In Chapter 2, you can make arguments between Summoner with Item and Summoner with Black Magic as the best secondary for summoner.  I know when I did a solo challenge with a bunch of extra restrictions (no yell/accumulate, no Auto-Potion, no randoms/grinding) I ended up deciding Summoner with Black Magic was better than Summoner with Item in Chapter 2.  Buuuuut that's a solo challenge, where Phoenix Down does nothing instead of being fantastic.  Regardless, White Magic was never on the table.  Ever.  Unlike Item and Black Magic which were both serious considerations.  Incidentally, Yin-Yang Magic was way higher on the "competing with black magic/item" list than White Magic, by the way, because Silence Song is sometimes an insanely good spell

[metroid composite]

The specifics are hard to work out without hashing things like "how do you get through 'banned' jobs in a draft anyway?"

Mmmm I dunno, maybe something like...

If you don't have a editor of some kind?  Go through the class without learning the skillset of the class.  And don't abuse the stats of the class if it's like...Wizard (equip no weapon instead of a rod to lower your MA a bit).

If you are using an editor?  Not too hard to strip out generic wizard abilities, and replace wizard stats/equipment with Time Mage stats/equipment, and have enemy wizards in story battles use the undead wizard class so that they can still do things.

Although ehh...that's not exactly what I'm going for either.

Short version: if it's a class that hurts to be in like Thief, it should still hurt to be in the class.  If it's a class you don't mind being in at all, like Wizard, it shouldn't be painful to get JP in that class.  Just don't abuse this by being like "I'll be in Wizard for this fight because it's hard"--it shouldn't give you more power than your most powerful job at that point, at least, not if that helps you out of a hard battle.  Wizard is the only class where it's pretty hard to strike that balance, though.  Without spells they suck.  With spells, they are probably better than your best class (at the point you'd be going through them).  I dunno, maybe learn only Poison or something?

Basically this, except I think this is vastly underrating how powerful FFT offense is and how easy it is to have 0 resets, which makes "how fast" the real competitive measurement in question to me.

Mmm...assume there's some extra restrictions that make 0 resets unlikely.  Like...running with a party of three instead of a party of five.  (I don't want to say party of one, because that devalues some skills like revival).  Or, for example...I know that the Vancouver crew has decided to steal Elmdor's gear every playthrough (not because they actually plan to use the gear necessarily, just because it's hard to steal).  And certainly assume minimal grinding (cause that makes things harder).

I will say that 0 reset runs are a bit awkward because they really discourage percentage based skills.  Like...if you are told to get a 0 reset run, you're probably going to avoid Yin-Yang Magic, which is normally a very good skillset, and you...likely still learn Teleport, but you probably value it a lot lower.

[Grefter]

Hmmmm depends on the flavour of the event as well?   We haven't really had any discussion on the relative strengths based on ability to unlock in the drafting section.  You could make a hack where job reqs for everything is Squire 1 for it.

That does fuck with what counts for balance of JP costs in base FFT though.

[SnowFire]

With MOBAs on the mind, thought of doing a theorycraft MOBA draft format equivalent for WAXF.  Which is...  less interesting than it is for FFT, because the things that skew a draft vs. the in-game rankings thread feel more artificial.  In an FFT draft, you can fairly reasonably draft "late" classes (Samurai, Calc, Ninja, maybe even Time Mage) and get a relevant bonus by C3 or so vs. hard fights, and if your early game is bad it isn't particularly helped much by drafting the likes of Knight or Archer.  In WAXF, the hardest fights in the game are all early, and the hard fights late - which do exist - have the earlygame carry classes & their skills still quite relevant.  So there's a huge incentive to just pile on earlygame classes, and even a class that literally said "you win" in the lategame probably shouldn't end up drafted that high (e.g. Emulator).  Items for everybody means there's also a bit less chance of getting brutally cut off on healing, and the synergy between classes is a bit more subtle (sure, you want Martial Mage to go with Emulator, but they're both perfectly fine alone.)  So yeah, if you just take the in-game rankings thread and add +1 to the first set of classes (+ non-Alexia base classes), -1 to the third set (+Alexia), and -2 to the final set, then draft high to low, you're probably fine.

There's also the issue of puzzle maps.  And what counts as a puzzle map.  Like..  does the "escape the capital" map where they very clearly want you to use Berserker(s) count as a puzzle map?  How about the boss dude who can only be hurt by Felius physicals & magic?  I'm going to make a good-faith effort to draft some puzzle-relevant classes but assume that in-game, if a map becomes actively anti-fun without the class du jour, you'd just cheat through it and not worry TOO much.

Usual disclaimer that I'm not an XF expert (Elf is that way) and only played through once!  (but hey, I can theorycraft MTG despite not playing Constructed in ages...)


1. Elementalist
2. Dandelion Shot, Martial Mage

Obvious.  Get your top-tier offense.  Elementalist goes first since you can deploy multiples of 'em.  Needed for some puzzley maps anyway that want you to use magic and make physicals suck.

3. Sacred Slayer, Fantastica
4. Emulator, Secutor

Sacred Slayer is itemless healing, Widespread, & turn shenanigans.  Fantastica this high is weird since the class is crap, buuuut it is one of the first 4 (and thus ultra-relevant) and offers Slow Down for bosses and MP-busting for Katrina & a few other bosses.  I *guess* Team B takes Emulator at this point since they also have Martial Mage, and grabs Secutor for earlygame as well as still being relevant the rest of the time.

5. RangerStormrider, Arcanist
6. Gadgeteer, Excavator

No Secutor = Ragnar is now more important.  Arcanist is backup itemless revival & good for various quasi-puzzle maps anyway where Teleport is important.
Team B grabs the remaining "first 4" class and also gets Excavator for Switch (vs. Katrina) and various puzzle maps, since Arcanist & Excavator are usually the best at those.  (There's also the Invisible Stalkers map...  Team A has Widespread Elementalists to kill them already, so Team B would like Detect.)

7. Grappler, Royal Fencer
8. High Cavalier, Strider

Get Accelerate & Alexia hype.  Team B hate-drafts Alexia-support (Trail / Intrude) because that is probably better than taking "real" picks at this point.  EDIT: Also Drop Kick is nice to throw on B's Secutors, per Elf.

9. Halberdier, Enigmancer
10. Extremist, Sentinel
11. Nightstalker, Berserker
12. Geomancer, Doge

Big pile of who-cares here.  I guess Engimancer is nice when combo'd with Sacred Slayer Widespread for IFF very lategame, and Debilitate is nice off Extremist, and Sentinel helps vs. the undead-rushing-into-town mission.  Oh and there's that early boss with tons of HP where only Felius can do damage to him whee that's fun (but also maybe doesn't count).  Geomancer gives you Shut Out which is useful for the final battle and that's about it.

Ends up with:
Team A: Elementalist, Sacred Slayer, Fantastica, Stormrider, Arcanist, Grappler, Royal Fencer, Halberdier, Enigmancer, Nightstalker, Berserker
Team B: Dandelion Shot, Martial Mage, Emulator, Secutor, Gadgeteer, Excavator, High Cavalier, Strider, Extremist, Sentinel, Geomancer, Canine

When stripped down to base classes + sets 1 & 2 (aka what you'll have for the stretch between near the end of C1 and the first half of C2 which owns you in the face)...

Team A: Elementalist, Sacred Slayer, Fantastica, Stormrider, Arcanist, Halberdier,
Team B: Dandelion Shot, Martial Mage, Secutor, Gadgeteer, Excavator, Sentinel, Luci

Yeah, both of these seem solid enough for surviving.  Team A blows you up with magic while slowing the boss and having Ragnar (and uh Felius I guess) hold the fort with physicals, and uses SS healing / turn manipulation.  Team B has a Secutor frontline with Sacrifice / Blast for magic, uses Gadgeteer healing & MM turn manipulation, and has Sentinel ZOC.  (Although...  Asgard C2 seems like he might be a bit of an issue with no Lightning & no Slow Down.)   Team A is definitely better in the early running since Team B blew an early pick on Emulator, but Team B at least gets to own the lategame with that.

[Dark Holy Elf]

Minor correction, Ragnar's class is called Stormrider.


FWIW, what I would do with an XF draft is what some FE10 drafts (and other FEs?) do... if you can't (or do not wish to) win a battle without using a certain class, well, the draft rules are relaxed so you can use said class, but at a penalty for each instance of the class you use. Obviously the larger the penalty the more certain puzzle-happy classes benefit.


I'm a bit skeptical that Team B would want to pick two special characters with its first go (since that means only two PCs out of your six will actually get to use anything from the first five classes chosen), although they are both excellent jobs to deny to Team A so maaaybe. Still, letting A get its hands on Elementalist/SS/Fantastica is brutal and something that they never recover from. The fact that I'm not certain there's a better opening pick for B makes me think that Elementalist is kinda OP, perhaps more than I thought even. With that magic stat and dirt-cheap Elementalist EQ they synergise sooo well with any other class that uses strong, magic-based skills.

I'm a bit skeptical about how high Excavator went. Switch has its uses certainly, but... that's mostly when combined with Fantastica's MP killing, which Team B doesn't have. Switch doesn't actually beat Katrina, since you can't lower her HP to zero. I guess Katrina 2 (who casts spells) can be manipulated into having, like, the exact HP equal to the MP cost of one her spells, then Switch, then she'll cast a spell and you win? Never tried this so it might not work. Still, eh. If I'm Team B I take Strider at pick #6, it synergises well with Secutor (accuracy boost, Drop Kick) and in particular it's range 3 that doesn't require Rush, the only way Team B can do this before its late pickup of Emulator attack spells. EDIT: except Blast, but that's only one PC, and suffering from a lack of higher-Mag carriers for a while.

I feel like Enigmancer should have gone earlier, but I'm not sure where. 4 move mage is good, Devastate is good. At the latest, I think Team A takes it at pick #5 instead of one of the two specials. Team B may take it earlier just to wall the complete domination of generic mages that A is working on... but it's hard to see where, B does not want A to get its hands on any of the classes it picks as its first four except maaaybe Emualtor, and Emulator's probably still a better pick than Enigmancer since it has a lot of the same strengths, and more.

There's actually a slim chance Crossbreed may not be last! He greatly increases the chance Team B gets past the first Samille/El Jackson fight without a penalty. ... nah. I can't see Team A taking him over any of its last four picks.

But yeah, Elementalist too strong. Dandelion Shot, Martial Mage, Enigmancer, Emulator, Sacred Slayer, and Fantastica are all good jobs to start with, and getting Elementalist makes every single one of them notably better.

[metroid composite]

Alright, so going back to the levels...didn't get a particularly negative response about the re-calculation with elfboy's random encounter numbers and assumptions of 5 actions per fight.

So...looking at levels at the store checkpoints I'd picked out...

Sand Rat Cellar: Level 3, (20 actions)
Meeting at lionel: Level 10, (70 actions)
Orbonne: Level 17 (120 actions)
Bethla: Level 27 (169 actions)
Before final dungeon: Level 35 (213 actions)

Actually, let's assume 1 extra deep dungeon levels or a sidequest fight before the final dungeon, just so each of these checkpoints are 50 actions apart.  And ooh, minor turn miscalculation, should actually be 170 for Bethla (this is working out eerily well; other than the final checkpoint, I've done no manipulation to get these all exactly 50 turns apart o_O).

Bethla: Level 27 (170 actions)
Before final dungeon: Level 36 (220 actions)

Incidentally, one part that surprised me is more levels gained between Orbonne and Bethla than levels gained between Bethla and the before final dungeon, despite the same number of actions.  Yeah...turns out a lot of those actions in the second half of chapter 4 are just the more-than any other chapter random encounters, which happen at party level and thus give slightly less exp.  Anyway, while these numbers feel high to me, I'll proceed with these if nobody else feels they're weird.  Using these to both:

A) calculate what JP setups are possible with these number of actions

and...

B) Calculate actual damage numbers for setups instead of speculating

[Dark Holy Elf]

Idly, how did you end up handling the randoms (especially the backtracks that may or may not happen)?

[metroid composite]

Idly, how did you end up handling the randoms (especially the backtracks that may or may not happen)?

I took your number of 33% chance, and then assumed 3 actions on enemies at party level.  I ignored all backtracking potential.



Now to take a brief break from things people actually care about (FFT) to...things people don't!


I'm going to attempt to calculate Twilight characters in D&D stats.

So...what do we know about Twilight characters?

Edward across multiple instances is reported at moving about 1000 mph (through difficult terrain like forests).  Various examples--running from Denali Alaska to Forks Washington in under an hour (800 miles).  Running from Port Angeles to Forks through a forest in "minutes" (56 miles), and notably faster than they were driving (well over 100 mph).  Making a 5 mile trip through a forest while carrying a human on his back in "seconds".

And some stuff that is easier to quantify in D&D terms.  Able to move completely silently to human eyes.  Able to move so fast as to basically appear out of thin air to human eyes.

Strength feats are less clear, because nothing short of fighting other vampires tests his limits.  Certainly a minivan is no problem (roughly 2000 kg).  Nor is lifting a 1m boulder and throwing it a long distance (also roughly 2000 kg).  Or snapping off a tree branch that is two feet thick, and swinging it to knock down the original tree (probably 600 kg on the tree branch).

Hmm...well, let's start with the easier one.

Bella Swan:
Strength: 8 (She's very uncoordinated, and not athletic, but takes books all over the place without noticeable encumbrance so she's not ridiculously weak).
Dexterity: 6 ("Significant Klutz" yeah, pretty much).
Constitution: 8 (Well, she's not "unusually prone to disease and infection" that a 6-7 score would imply, but certainly "incapable of a full day's hard labour" for 8-9)
Intelligence: 16 (Yeah, Bella is unusually smart; she figures out the vampirism, picking up on clues that other humans missed.  She was in advanced placement classes.  She nerds out with Jane Austen for fun.  She tricks a vampire who can predict the future.)
Wisdom: 7 ("Seems to have almost no common sense"...or sense of self preservation.  This extends to more than just vampires and werewolves, and includes walking down dark alleys).
Charisma: 13 (She's a nobody in Phoenix AZ, but her charisma stands out in Forks enough for her to be moderately popular to the point of making several girls jealous...sooo better than an average human).

Alignment--Chaotic Good?  Definitely good (her defining character trait is "martyr") and likes to prove old rules and traditions wrong, so more on the chaotic side.

(Total pointbuy cost: 1 LOLOLOL.  Actually I'm surprised it's above 0.)



So...moving on to vampires...there's a slight problem that there's no movement bonus for ability scores.  All creatures with 30 base movement speed can move 6 miles in an hour when hustling.  This is a lot less than 1000 miles per hour.

Well...ok.  There is one way to gain movement in pathfinder, and that's to be a monk.  Normal 30 ft pathfinder characters can move 120 feet when running at full speed in a 6 round turn.  That's 20 ft/second, 6 m/s.  21 km/h.  13 mph.  So...would need to go from 30 ft pathfinder character to like...3000 ft.  Roughly a level 900 Monk then? >_>

OK, strength--it seems fairly clear that 2000 kg qualifies as a light load.  Strength load numbers are exponential, so...strength score of 45?  Something around that.

Alright, so...dexterity.  What's the DC check for "with no training, jump out of a third story window, and land gracefully on the toes of your feet so that you don't snap your stiletto heels"?  Hmm...yeah I don't really have an answer to that.

OK...so there's DC checks on jumps.  But the problem is, characters with high base movement get + bonuses to jumps; at 3000 ft move speed, the bonus to jump would be like...+1200 skill bonus.  This allows a long jump of 1200 feet, or a high jump of 300 feet.  Yeah, these are pretty similar to numbers in the books.  Well...let's pretend that pathfinder did have a way to translate high dexterity into movement, and that it matched this jump conversion.  This means roughly 2400 dex.  >_>.  Yeah, sure, whatever; I really can't think of a time any of them failed anything that could be considered a dex check.  Ever.

Constitution.  Let's see...the description of 35 is..."Nearly immune to any level of fatigue, illness, disease, or infection- such a creature's stamina is practically god-level" hmm...well they literally don't sleep, don't get fatigued, and are immune to disease and infection.

Although on the other hand "Most undead creatures do not have a Constitution score" -- yeah, this might be more accurate.  Probably this.

intelligence -- definitely superhuman.  The brain has to process at a higher framerate just to make the sheer speed possible--which creates an appearance of intelligence (basically able to have a minute's worth of inner monologue in less than a second).  But reports also include being able to juggle more ideas simultaneously.  The long lives also tend to make the vampires have more raw knowledge than most humans.  That said...still within reach intellectually; humans outsmarting vampires is rare, but happens.  Something like 22 int?

Wisdom -- Well...discounting abilities (Alice, Jasper, Edward) and discounting the kind of skill checks wisdom is based off of (perception...; vampires get huge bonuses to that due to their eyes, but that's a physical trait, not wisdom)...no, vampires aren't especially wise.  To the point that when Edward can't read a mind, he doesn't know if a faster heart rate means a positive or negative emotion.  I think this is a 10.

Charisma -- ooh, vampires are pretty good at this.  To the point of "basically getting free suggestion spells".  Not flawless, people can turn on them, it's possible to resist them, but certainly "Renowned for wit, personality, and/or looks" -- which is the 24-25 description.  Most people also have an instinctive fear of them, but causing fear is also charisma based.  Let's err a bit high--unlike intelligence, where sure, vampires think faster than humans, but there's still big dumb vampires that prefer smashing things to intellectual pursuits, so it's conceivable that a smart human might be smarter than a dumb vampire...but with Charisma this doesn't happen (see: Rosalie as a human).  So like...30?  Still low enough that humans can sometimes resist persuasion.


So something like...

Strength: 45 (or more?  Unclear what the limits are)
Dexterity: 2400 (Admitedly, this is Edward/Neonate numbers.  Other vampires will be lower, but still in the four digit range).
Constitution: - (undead)
Intelligence: 22
Wisdom: 10
Charisma: 30

[Dark Holy Elf]

With the caveat that I'm not familiar with Pathfinder specifically (but you did say D&D, so...!):

Dex doesn't govern movement speed at all in D&D systems. Maintaining speed over a long distance is based on Con, while max speed isn't related to stats at all: it is class/race-specific (which makes sense... a peregrine falcon with unremarkable Str/Dex/Con is far faster than humans with far superior stats). Jumping is also never based on Dex.

What Dex primarily governs is how hard you are to hit, how fast your reaction speed is, and certain skills such as acrobatics (landing from a three-story fall without breaking your heels, as you note), moving quietly, and using your hands for delicate tasks. I'm certain Twilight vampires are still excellent in this regard, but 2400 is completely ridiculous. Just based on your rough descriptions I would assume they are something closer to 19-25, but more information would be helpful.

You're probably overrating the strength somewhat too. Leafing through the 5e monster manual... a kraken (gargantuan creature, can lift ships let alone 2000 kg objects) "only" has a strength of 30, so it seems unlikely that Twilight vampires are better than that. The stat appears to be strongly logarithmic past a certain point.


(Actually, in general, D&D stats tend to work a bit differently at certain ranges. 3-18 is supposed to cover the range of human experience with a distribution you'd expect from rolling 3d6. 1 and 2 cover "worse than any human" which can be a wide range in some cases, e.g. animal Int stats. Then higher numbers cover everything from the barely super-human to godlike creatures... but for the most part, don't go much past 30 if at all, and are scaled to cover this range.)

[AndrewRogue]

What's the DC check for "with no training, jump out of a third story window, and land gracefully on the toes of your feet so that you don't snap your stiletto heels"?  Hmm...yeah I don't really have an answer to that.

I do!

As of 3.X epic rules...

DC 45 Tumble check allows you to treat a fall as if it were 30 feet shorter. If you want to add extra credit, DC 60 allows you to treat a fall as 40 feet shorter. And a DC 100 check allows you to ignore fall damage completely.

I will also note that DC 120 Balance checks allow you to stand on clouds.

[SnowFire]

A) There's a lot of issues with super-speed, in that "it doesn't really work" is the short answer.  But it's cool, so we have a magic wand that says it DOES work.  I don't think attempting to assume the mechanics of *actually* moving very fast apply, and if they did, aside from superhuman "Con" from the strain it'd place on a body (hey, vampires are magic, so we already have an out here!), it's hard to read too much into "super intelligence / processing speed".  Imagine a super-genius piloting a car at 300 mph through traffic on a highway.  Even if they are really, really perceptive at trying to plan a safe route, they ultimately have like a tenth of a second to make adjustments to not crash.  Being a normal super-genius doesn't help that much, and god forbid there's anything like fog around (something fairly common in Washington state!), so even like Int 40 (futuristic supercomputer!) only goes so far.  What would help more is expanding the super-speed magic wand to include just "naturally" avoiding all obstacles when speeding up.

B) Bella is 18 Charisma.  Not because the author is remotely convincing at making her authentically charismatic (a la, say, Kefka or Balthier), but because everybody just wants to be friends with her on meeting her or just generally thinks she's Important and should be Paid Attention To despite being a random mortal high schooler (e.g. for the villains).  That's a side-effect of 18 Cha; you carry a reality-warping field of "everybody looks up when you enter the room."  In all seriousness, this is one of the nice things about roll-play rather than role-play; sometimes a shy player wants to try their hand at a super-charismatic character, so sure, even if you can't charm them off their feet yourself, go ahead and roll a Diplomacy check...  okay, yes, Bella knows the right things to say, and her brown hair is shimmering at just the right angle from the sunlight, etc. etc. you're fine and your character just naturally knows what to do even if you don't.

[Grefter]

hat's a side-effect of 18 Cha; you carry a reality-warping field of "everybody looks up when you enter the room."  In all seriousness, this is one of the nice things about roll-play rather than role-play; sometimes a shy player wants to try their hand at a super-charismatic character, so sure, even if you can't charm them off their feet yourself, go ahead and roll a Diplomacy check...  okay, yes, Bella knows the right things to say, and her brown hair is shimmering at just the right angle from the sunlight, etc. etc. you're fine and your character just naturally knows what to do even if you don't.

So... accurate representation of the books then?

[SnowFire]

That's the point, yes.  (Met has Bella at 13 Charisma, you'll note.)

[Grefter]

Which I also think is accurate.  In a world where average Charisma is 5, a 13 is about right for a comparative effect just by sheer inverse properties of the crowd's negative charisma..