Labyrinth of Touhou

[Talaysen]

Apparently I forgot to make a topic for this.

http://www.rpgdl.com/wiki/index.php?title=Labyrinth_of_Touhou

[Reiska]

Alright, moving here from the wiki page as Talaysen pointed me

Regarding the mention of post-game stats on the wiki: I'm working on a New Game +* playthrough right now using a fixed party; as a side effect of this, it means I'll be able to see exactly how much total EXP I've gained along the way to the final boss (since there'll be a bunch of characters I never use and thus leave at level 1 forever).  

From this, I should then be able to extrapolate what level each character everyone in the cast will be with an equal distribution of EXP at endgame - the characters I never used will all have 80% of my total gained, so I just need to divide their EXP totals by 0.8, then on a new game + file I can set all characters to that precise amount of EXP (using CE) and level them up.  I'll still need to figure out an appropriate skill point distribution, though I suspect for the purposes of RPGDL it would be best to come up with a standardized one like the wiki page already uses; I'll see how my own develops as I go, though it'll probably be skewed a bit by my use of a fixed party (due to me not throwing skill points at characters I stop using later in the game as better replacements come along).

* LoT's New Game +, unlocked on completion of the post-game content, gives the player the options of what to carry over from the cleared save file; you can choose to retain (or not retain) EXP totals and levelups, skill levels, equipment, and character unlocks.  The run I'm doing carried over the character unlocks but nothing else, so I started with all 40 characters unlocked at level 1 with no skill levels or equipment.  The Maidens of the Kaleidoscope forums have a save file available for download at the beginning of the game in this state, though it is slightly glitchy, with two of the postgame characters having incorrect affinities and status resistances - as I understand it this is a game bug, but I need to actually finish the postgame myself too.  Luckily, it can be corrected with a bit of hacking as the intended values are known and available.  I've notified the supplier of that save, anyway, so they'll probably have it fixed sometime soon.

As I noted in my comment in the wiki, I do have a list of all characters' base stats at 200,000,000 EXP before the application of any skill levels or level-up bonuses; however, 200,000,000 was a number I simply chose arbitrarily when I made the list.  If I had to wager a guess, 200,000,000 EXP would represent deep postgame but probably not quite powerful enough to take on the game's superboss on 30F, but that's a bridge I'll cross when I actually get to 30F ;)  Nonetheless, 200,000,000 EXP puts the characters in the range of level 325 (Rinnosuke, the slowest leveler) to 413 (Chen, the fastest, who is way ahead of everyone else).  I can toss that stat dump up on my webspace if anyone wants to see it, but I would concur that the end of the normal game is probably a better point at which to sample character stats for the purposes of RPGDL; due to the way the stat mechanics in LoT work (exponentially), the differences between the characters will become increasingly exaggerated as levels climb higher.

[Talaysen]

Welcome to the RPGDL forums, Reiska!  Always nice to see new people around here.  :)

End of normal game is pretty much always what is used for the RPGDL.  Postgame stuff is generally not allowed (though some people do allow it, and sometimes the definition of "postgame" is disagreed on).  Nonetheless, having postgame stats is interesting just to see how the characters shape up then, and there's always a possibility of those stats being used in one of the board tournaments here, since they don't always adhere directly to the normal tournament rules.

I don't remember what I did for level up bonus distribution and I can't even figure out what I was doing looking at my spreadsheet...  I think it depended on the character's growths (higher growth => more level up points allocated) or maybe percentage increase in stat after allocating.  I'll see if I can remember later.  For skill point distribution, I just levelled up every stat until it required more than a certain base cost to increase further.  I think I used 60* or something for endgame stats, but since you get more skillpoints in postgame that would be significantly higher.  You don't have to use this distribution if you don't want, though.

* How I came up with this seemingly arbitrary number:  I took the total amount of skillpoints you gain throughout the game, then divided it by 12 (the maximum number of people you can put in the party).  With that number, I took a character (probably Reimu since she's first on the list), and kept putting skillpoints into the lowest cost stat until I ran out.  The lowest cost at that point was in the 50s, so I rounded up to 60 and used that.

Also, for the record, the stats mechanics in LoT aren't exponential with level, just quadratic.  Which means that the stat differences will be exaggerated as you said, just not as much as you implied.

Anyway, if you need any help with the entries for postgame characters, I am willing to help out.

P.S.  Oh right, regarding your use of a fixed party.  I should explain how the RPGDL usually does this.  Essentially, what we normally assume (and mind you, some people disagree, nothing in the RPGDL is set in stone as people have different interpretations on how character stats should be considered) is that one fixed party is used throughout the game, but that EVERY CHARACTER is in that fixed party WHENEVER POSSIBLE.  So in the case of LoT, you would find out how much EXP characters get if they were in the party the entire game (well, some would say from the point they join until the end... but IIRC characters start with an appropiate amount of EXP when they join so it makes no difference in LoT), and then assume they get that much EXP.  Same for skillpoints (which is why I only divided by 12 instead of 32 when I did skillpoint distribution).  It sounds like this is what you were planning on doing anyway.

It's actually really easy to do (and the most inutitive way to do it) for LoT, but can be more complicated for other RPGs.

[Reiska]

Welcome to the RPGDL forums, Reiska!  Always nice to see new people around here.  

End of normal game is pretty much always what is used for the RPGDL.  Postgame stuff is generally not allowed (though some people do allow it, and sometimes the definition of "postgame" is disagreed on).  Nonetheless, having postgame stats is interesting just to see how the characters shape up then, and there's always a possibility of those stats being used in one of the board tournaments here, since they don't always adhere directly to the normal tournament rules.

I don't remember what I did for level up bonus distribution and I can't even figure out what I was doing looking at my spreadsheet...  I think it depended on the character's growths (higher growth => more level up points allocated) or maybe percentage increase in stat after allocating.  I'll see if I can remember later.  For skill point distribution, I just levelled up every stat until it required more than a certain base cost to increase further.  I think I used 60* or something for endgame stats, but since you get more skillpoints in postgame that would be significantly higher.  You don't have to use this distribution if you don't want, though.

* How I came up with this seemingly arbitrary number:  I took the total amount of skillpoints you gain throughout the game, then divided it by 12 (the maximum number of people you can put in the party).  With that number, I took a character (probably Reimu since she's first on the list), and kept putting skillpoints into the lowest cost stat until I ran out.  The lowest cost at that point was in the 50s, so I rounded up to 60 and used that.

Also, for the record, the stats mechanics in LoT aren't exponential with level, just quadratic.  Which means that the stat differences will be exaggerated as you said, just not as much as you implied.

Anyway, if you need any help with the entries for postgame characters, I am willing to help out.

P.S.  Oh right, regarding your use of a fixed party.  I should explain how the RPGDL usually does this.  Essentially, what we normally assume (and mind you, some people disagree, nothing in the RPGDL is set in stone as people have different interpretations on how character stats should be considered) is that one fixed party is used throughout the game, but that EVERY CHARACTER is in that fixed party WHENEVER POSSIBLE.  So in the case of LoT, you would find out how much EXP characters get if they were in the party the entire game (well, some would say from the point they join until the end... but IIRC characters start with an appropiate amount of EXP when they join so it makes no difference in LoT), and then assume they get that much EXP.  Same for skillpoints (which is why I only divided by 12 instead of 32 when I did skillpoint distribution).  It sounds like this is what you were planning on doing anyway.

It's actually really easy to do (and the most inutitive way to do it) for LoT, but can be more complicated for other RPGs.

Thanks for the welcome

I'm pretty sure how LoT handles the starting experience for new recruits is simply giving them 80% of your total gained experience to date, the same as a character you've recruited but not placed in your party is given, though I can't say I've tested it exhaustively.  In any case, yeah, LoT makes it easy to figure out; in a normal run, since you can grab Mystia immediately after finishing the game before doing any postgame events or battles, and assuming I'm right about how it figures new recruits' EXP, you'd simply divide whatever Mystia's starting EXP is by 0.8 and you'd have the endgame total for maximum active use.  (In practice, all of your characters will probably be slightly below that point on account of finishing fights dead and such, which will give them 80% of the EXP for that fight and any subsequent ones until you pop back into Gensokyo, but in any case they can't have *less* than 80% of the maximum.)

My fixed party NG+ run (which, for what it's worth, is slogging around floor 5 now) is using a team of Meiling, Mystia, Chen, Rumia, Yuyuko, Youmu, Orin, Renko, Kanako, Minoriko, Nitori and Suwako.  I must say, figuring out how to take out some of the bosses without Reimu tossing defense buffs around constantly is proving interesting.  Renko *can* buff (and buffs ATK/MAG at the same time), but her buff also eats half of the rest of the party's active gauges, which can make it decidedly dangerous to use if timed wrong.  In it, my current skillpoint distribution is just 15 levels to primary attacking stat, 10 levels to HP/SP, 10 levels to speed, and then either 5 or 10 levels to each of defense/mind based on how good the bases are for that character.  (Exception Renko, whose attacking stats you want to be as low as possible due to the way she works - she doesn't actually have any real attacks, and her buff does friendly fire damage!)  I haven't really touched TP yet since I'm still in the part of the game where SP-per-fight is low, and affinities for now I'm just working on bringing up to at least 100.

If I were going to do a fixed run in normal mode, I'd probably just use the starting 5 plus the first 7 recruits.  If my memory of the game serves me well, that works out to Reimu/Marisa/Remilia/Sakuya/Patchouli/Meiling/Chen/Wriggle/Cirno/Minoriko/Youmu/either Alice or Rumia depending on what order you do events in.  Which, actually, wouldn't be a bad party at all; Minoriko's the game's best healer for the most part, Reimu supplies all the AoE buffing/healing you could ever want, and you have plenty of powerful damage dealers in that bunch.  Long-term I suspect Sakuya, Cirno and Wriggle would be the weak links in that team, but Wriggle's poison is useful all throughout the main game, and I guess Cirno can paralyze stuff or something.  The party's biggest weakness is its lack of targeted offensive buffs, but since three of the characters have powerful self buffs, this is somewhat mitigated.

If I had my choice of Alice/Rumia, I'd probably go with Rumia.  There are some boss fights I wouldn't want to imagine trying without Demarcation.

Random addition: Went ahead and quickly played with 200M exp base stats I have (the ones with no levelup bonuses or skillpoints applied, just the raw baseline stat growth of each character before the player's input gets involved), and threw together a few things:

Averages - HP 54951, SP 774, ATK 26133, DEF 19169, MAG 25890, MND 22622, SPD 881

Ranges:
Level - 325 (Rinnosuke) to 413 (Chen)
HP - 23064 (Patchouli) to 111388 (Komachi)
SP - 325 (Youmu) to 1367 (Patchouli)
ATK - 5317 (Patchouli) to 58258 (Flandre)
DEF - 5301 (Patchouli) to 34816 (Tenshi)
MAG - 2840 (Yuugi) to 42368 (Patchouli)
MND - 2654 (Flandre) to 45008 (Patchouli)
SPD - 508 (Patchouli) to 1650 (Chen)

As you can see, Patchouli is either best-in-game or worst-in-game on every stat, truly a character of extremes.   Also: Neither Flandre or Patchouli are actually the strongest attackers in the game, despite having the highest raw stats (measured against the final boss's final form) - that honor goes to Yuka if you're considering all 40 characters, and Marisa (if you give her equipment to get her up to 1760 SP to fully power her Master Spark) if you're considering only the basic 32.  Without Marisa's Master Spark fully powered, she gets overtaken in turn by Nitori.  That said, IMO, Master Spark should be fired as soon as Marisa has the 160 SP to fire it again - charging it up doesn't add all that much damage in the long run for how long it takes for her SP to refill, and a 160 SP Master Spark is still the 6th highest damaging attack in the game against the final boss.

[Reiska]

Just a (belated!) update: I am in fact still working on playing through this, although I switched to a regular, non-fixed party run.  It's slow going, but it's up to 14F (final being 20F) now.  Been pretty busy, both with real life and with other games ;)

My actual endgame skillpoint distribution is going to be all over the place (since there are some characters I no longer use, and thus won't give skill points to anymore, resulting in inequal distribution).  Luckily, I'll have a means of figuring out the total number of skill points earned throughout the course of the game: I can write down all characters' skill levels, load a clean NG+ save file that carries nothing over (so all it has is all characters unlocked at level 1 and no skill points allocated), use a memory editor to give myself 1 billion skill points, and then level everyone's skill levels to exactly where they were at endgame.  From that I can derive exactly how many skill points were earned, and accordingly come up with an equitable distribution for the purposes of all 32 characters getting a fair shake.  I'm leaning towards, for the purpose of producing the character comparison, not allocating any skill points to elemental affinities - this follows the logic of the SRW J topic, choosing to emphasize the uniqueness of each character's baseline protections.  After all, I can make Cirno actually resistant to fire with enough skill points, but is that appropriate thematically? Not really.

Incidentally, I pulled the stat calculation algorithms the game uses from its Japanese wiki as well, so I could probably reverse-engineer your bonus allocations with some effort. ;)  In almost all cases, I've been putting all levelup bonuses into a single stat for each character (whichever I feel complements their party role best); so far, the only exceptions have been Wriggle and Tenshi, which have gotten a 50/50 split of DEF/MND.  The allocations so far (by stat):

HP/SP: nobody
ATK: Remilia, Sakuya, Chen, Cirno, Youmu, Yuugi, Aya, Komachi, Suwako, Nitori, Suika, Flandre^, Orin^
DEF: Meiling, Wriggle (50%), Tenshi (50%)
MAG: Marisa, Patchouli, Alice, Rumia, Sanae, Ran, Reisen, Eirin*, Mokou, Yuyuko, Kaguya^
MND: Reimu, Minoriko, Wriggle (50%), Iku, Tenshi (50%)
SPD: nobody

* this was probably not a good idea
^ I haven't yet recruited this character, but I'm reasonably certain this is the build I'm going with
There's certainly some room for debate as to how to allocate some characters - Eirin, for example, is more commonly given a def/mnd build, but I didn't need any more tanks so I specced her offensively (which, IMO, is so far proving to be a mistake), and some people also advocate a MND build for Patchouli.  Fortunately, since I have the stat formulas, I can easily derive what anyone's stats "would have been" had I given them a different build of levelup bonuses - so for the characters which are most disputed in what is a proper build for them, I can present them with multiple builds.

At this point, there's only two characters whose builds I'm not yet clearly decided on, and they're the last two you get: Yukari and Rinnosuke.

Yukari can put out very good single-target offense if she's built for it AND if both Ran and Chen are also in the active party (note that keeping Chen in the active party during boss fights is often a very bad idea).  This is compounded by the fact that if you only have one active, Chen is always the better choice for Yukari's offense, though the practical difference is small unless her target has extreme MND.  Yukari is also one of the best support characters in the game with a clone of Reimu's defense buff and a spell that is essentially an AoE version of FFVI's Quick; however, no amount of leveling her DEF or MND will make her into a first string tank (but it will make her into a second string tank like Wriggle or a specialty MND tank like Iku).  Another strike against offensive Yukari is that, like Eirin, her AoE attacks are garbage (but better than Eirin's at least) and so her offense is boss fight oriented, and boss fights are typically the sort of place where Yukari will be supporting more often than not.  With a mind to postgame, also, the utility of Yakumo Ran + diminishes more and more as having Chen active becomes more and more of a liability (she has a hard time contributing to the postgame boss in general, as Flight of Idaten doesn't pierce huge defense well).  I'm leaning towards a defensive build for Yukari, I think, but I have a while to decide yet.

Rinnosuke, on the other hand, is even more of an enigma.  None of his attacks are particularly bad at all; built for attack and against typical opposition, Shining Stars of Traumerei will generate the most damage, though it's a row attack AND composite and carries all the usual pitfalls of both.  The Word for World is Forest also does solid damage and is a proper AoE, and it's not composite.  Start of Heavenly Demise, his most expensive attack, is actually weaker than either of those two and thus is a worthless piece of crap.  His lone single-target attack is his weakest against typical opposition, but it carries fantastic defense piercing relatively speaking, so against the postgame boss it becomes his best.  On the other hand, built for defense, Rinnosuke can be a passable slot 1 tank or a great slot 2 tank (though only in boss fights, given his abysmal TP).  I'm leaning offensive here; defense piercing can be at a premium sometimes.

[Reiska]

Idly, detailed stat mechanics.  These came from the Japanese wiki and were translated for me by Deranged, who also did the bulk of the work for the game's actual translation patch as I recall.

HP = ((Current Level + 6) * Growth Rate + 12) * Parameter Multiplier
SP = ((Current Level / 8) * Growth Rate + 100) * Parameter Multiplier
ATK = ((Current Level + 4) * Growth Rate + 3) * Parameter Multiplier
DEF = ((Current Level + 4) * Growth Rate + 1) * Parameter Multiplier
MAG = ((Current Level + 4) * Growth Rate + 2) * Parameter Multiplier
MND = ((Current Level + 4) * Growth Rate + 1) * Parameter Multiplier
SPD = ((Current Level / 32) * Growth Rate) * Parameter Multiplier + 100
EVA = who cares? but it's ((Current Level / 12) * Growth Rate + 2) * Parameter Multiplier

Parameter Multiplier is a bit different for each stat, but it's where levelup bonuses and skill point level bonuses come in.  It starts at 1.00 for all stats on a fresh level 1 character with one exception: for Remilia's ATK, it starts at 0.84.
For HP, you add 0.03 for each level gained, 0.02 for each levelup bonus assigned, and 0.04 for each skill point level assigned.
For SP, you add 0.01 for each levelup bonus or skill point level assigned; it gains nothing for just leveling up.
For all other stats, you add 0.02 for each level gained, 0.02 for each levelup bonus assigned, and 0.04 for each skill point level assigned.
Item bonuses are also added here, I believe.

No rounding is performed; all decimals are truncated.

Characters' growth rates can be found on the character pages on Touhou Wiki (http://en.touhouwiki.net/wiki/Labyrinth_of_Touhou/Characters)


progress update: http://www.twitch.tv/reiska/b/298854887

[Reiska]

Final exp total: 7710445
Final skill point total: 4019090 (allocation for each character: 334924)

Deleted the incomplete redo I had here because I'm never going to get around to finishing it and it had some math errors anyway.  However, I *will* use this topic (in the next post) to dump a whole bunch of info about LoT equipment for those who have permissive views on DL legality of equipment.

[Reiska]

In this post, a big LoT infodump.

The current stat topic is really badly overleveled.  I don't really have the time or energy to redo it, particularly with how squirrely some of this game's interps can be, suffice to say it's around 30-40 levels over what I'd consider endgame.  As noted in the post above, on my last cleared file my final EXP gain total was 7,710,445 and my total skill point gain was 4,019,090 (allowing a budget of 334,924 skill points for each character, calculated at 4019090/12).

Optimal allocation of levelup stat bonuses for any character is almost always putting them exclusively into one stat (chosen from ATK/DEF/MAG/MND).  What's best for who is a little debatable in some cases.  HP, SP, speed and affinities are a complete waste of levelup stat bonuses under all circumstances.

My personal beliefs about what levelup stat each PC wants in the DL are as follows.  Generally, the stat they want in the DL is the same stat they want in-game.

Reimu: MND, she benefits more from the extra durability than the extra healing power since the heal is pretty nearly overkill already.
Marisa: MAG, anything else is a waste.
Remilia: Probably DEF?  ATK is a little better in-game, but DEF might be better in the DL...
Sakuya: ATK
Patchouli: MND.  Patchy's pdur is unsalvageable, but pumping her MND can get her pretty close to effective magic immunity in-game thanks to subtractive defense and Patchy's MAG is solid enough at baseline anyway.
Chen: ATK, doesn't benefit enough from anything else.
Meiling: DEF.  Crazy pdur.
Cirno: ATK, while she's partly composite Diamond Blizzard is purely physical.
Minoriko: MND (see Reimu).
Youmu: ATK or DEF?
Alice: MAG.
Rumia: MAG, doesn't benefit enough from anything else.
Wriggle: DEF or MND.  Possibly the only character that might benefit from splitting levelup bonuses.  She does all her damage with poison so her attack stats don't matter.
Yuugi: ATK or DEF
Aya: ATK
Iku: MND (like Patchy, essentially walls magic for the most part).
Komachi: ATK.  She's honestly a bad tank, boosting her HP is a trap.
Suwako: ATK, her money spells are physical.
Sanae: MAG or MND?  I probably lean toward the latter but she actually has problems keeping up healing-wise.
Nitori: ATK
Ran: MAG
Reisen: MAG
Eirin: MAG.  Probably gameworst PC.
Tenshi: DEF or MND.  Pick one and pump it, in-game I'd lean towards MND because you have Meiling for physical walling.
Mokou: MAG
Flandre: ATK
Rin: ATK (see Cirno)
Suika: ATK or MND, I lean pretty strongly towards the former
Yuyuko: MAG
Kaguya: MAG or MND?  MND argument is similar to Patchy's
Yukari: DEF or MND, pick one - MND is better in-game, in the DL I'd favor DEF because she's going to be relying on her buff anyway
Rinnosuke: Probably DEF

Aftergame characters:

Renko: MND.  She doesn't really benefit from anything well here (you want her MAG as low as possible, her ATK/DEF suck), though.  Should be considered a DNR if postgame PCs are even on the table.
Maribel: MAG
Utsuho: MAG
Kanako: MAG
Yuuka: MAG, probably.  While she actually has great pdur she probably needs the MAG for her offensive game to work
Mystia: ATK
Keine: ATK or DEF?
Shikieiki: ATK

Optimal allocation of skill points is really, really psyduck and is the main reason I abandoned redoing the game's stat topic.  If I were going to approach it now, I would probably just completely ignore skill points, even though this is somewhat hurts how well in-game is reflected.  One could argue that inflated levels could compensate for this, but the levels would have to be pretty notably inflated for this to be workable (probably more than the existing topic), since one level of skill point bonus in a stat is effectively equivalent to gaining 2 experience levels for a stat not getting levelup bonuses, and 1 experience level for a stat getting them.  A good argument for ignoring skillpoints entirely is that it preserves the uniqueness of the cast members.  The only other distribution I think I'd consider would be splitting the skill point allocation evenly between HP, SP, ATK, DEF, MAG, MND and SPD (47846 skill points to each stat). 

In-game, skill points can be used to close PCs' elemental resistance holes; I'm not inclined to respect this in the DL for the purpose of maintaining uniqueness, and would take PCs' elemental affinities at their base values and consider them unaveraged.  You could, however, include them in the skill point allocation; this would reduce the SP budget for each stat to 25763, and would generally blunt the extremes of elemental weaknesses for the cast.

Equipment is universal and there are no shops; everything is found in chests or dropped.  Functionally speaking, all equipment is accessories, as there are no slot distinctions.  Each PC can equip three items.  I'll edit a full list of equipment in with acquisition methods and effects when it's not 6 in the freaking morning.