Shards of a Broken Sun [Megaten/Shugo Chara/Exalted]

[PistachioCookies]

Which is to basically try to use existing motives and already gained insights to try and reduce the difficulty of what is being attempted to be learned, which is to maintain the integrity of the souls.

Click to shrink...

Problem is with training, it's not so much a matter of difficulty as time. And the times are fixed.

Training Psionics from 1-2 is going to take a flat 56 hours and 1 XP. Training Integrity from 2-3 is going to take 112 hours and 4 XP. I don't think any stunts, no matter how well written, will change these times or XP requirements (which would be what represents the difficulty). From what I can tell, the only thing that the stunts get to change is how many hours a day we get to use to spend on the training and how much overlap we get on the training times.

Given the training regimen from before.....

Amu spends 6 hours a day at school from 9 to 3 roughly speaking. Probably wakes up around 7-8 and only has an hour or so to primp herself up and eat breakfast before heading off to school. Supposedly, Amu then only gets 2 hours after school to train on anything, the rest of the time is taken up by non-training stuff. Ideally, we'd be able to fully leverage those 6 hours of school time to train but unless Amu cuts class entirely, she's probably going to only to get the equivalent of two more hours at most in that time (with one of those being her lunch break).

With only 25% overlap on Integrity, it will be a huge bottleneck over the the rest. Levelling three skills from Lv1 to Lv2 takes 168 hours, though if we get 25% overlap on them, it comes down to 126. But Integrity takes 112 hours - even with 25% overlap, it's still going to take 84 hours to finish. That puts the whole package at 210 hours to complete. If we somehow squeeze 2 extra hours per day from doing stuff at school, that means 53 days to finish - nearly 2 whole months, before we see a single level in any of the skills we tried to level. Without Integrity, it comes down to 32 days, which is nearly half the time in comparison, but if we're not training it at the same time, we won't have the EXP to emergency boost it unless we schedule nothing else but Illusion for training (since it takes 8 XP to SOS boost it from scratch).

And that's assuming we do, in fact, get 2 extra hours from trying to do stuff at school.

I'm thinking that if we attempt the slow route, we'll be heading into the next arc without any new levels. I don't expect more than a month of downtime before it gets dropped onto us.

 

[Quickshot0]

Problem is with training, it's not so much a matter of difficulty as time. And the times are fixed.

Click to shrink...

Lots of fair points. Though taking all those in to account, one can still try to improve the efficiency then by finding better overlap, or just finding something that might effectively do more integrity training in the same time. And one can also try to leverage extra gained knowledge in an attempt to reduce total amount time needed to train.

Of course whether @Baughn will agree with those I don't know, but we're pretty time constrained, so I can't help but want to try anyway.

So in that sense, one can try arguing that perhaps Amu already has thus already seen some integrity stuff lately, and also gotten some extra training in it due to recent events. Her soul abrasion surely means something in that regard after all.


Ah well, at worst I've merely thought up some more potential story details.

 

[Baughn]

Of course whether @Baughn will agree with those I don't know, but we're pretty time constrained, so I can't help but want to try anyway.

Click to shrink...

The listed training times are already insanely low, and assume a great deal of shenanigans on Amu's part. In-story that why you didn't have the option before—she had no reason to go nuts in that manner before. Now she does.

In other words, you're not getting a discount on them. But spending more than two hours per day shouldn't be too difficult; two hours is just the baseline she doesn't need to justify to anyone.

The listed times actually assume 48 hours per week, fwiw. Very much a full-time job.

 

[Quickshot0]

The listed training times are already insanely low, and assume a great deal of shenanigans on Amu's part. In-story that why you didn't have the option before—she had no reason to go nuts in that manner before. Now she does.

In other words, you're not getting a discount on them. But spending more than two hours per day shouldn't be too difficult; two hours is just the baseline she doesn't need to justify to anyone.

The listed times actually assume 48 hours per week, fwiw. Very much a full-time job.

Click to shrink...

Roger that, I guess you'll give no mercy on potentially having learnt some of it as well then?

 

[PistachioCookies]

Would've been real nice if it turned out that Dreamwalking could be used to turn sleeping hours into training hours. But apparently, it's not even supposed to work properly while asleep. Ugh, talk about a waste.

But spending more than two hours per day shouldn't be too difficult; two hours is just the baseline she doesn't need to justify to anyone.

Click to shrink...

My gut says it will be at best one month before the next crisis is in full swing.

For us to train those 4 skills in that time, we would need to somehow squeeze 7 hours of training a day to hit completion before then, assuming 25% overlap. If we're training 3 skills, including Integrity, we still need 5-6 hours per day (with the same amount of overlap).

We basically need to somehow triple the baseline hours per day to hit that. Only way I can see 4-5 extra hours of training time magically appearing out of nowhere is if Amu went full truant and just skipped class for a whole month and this would cause so many problems, whatever she gained in training time would be lost in trouble of her own making.

It really does look like we need to make a choice between only buying 1 skill at a time to make sure we have the levels for the next arc, versus getting an overall bulk discount on training times for multiple but only having them ready by the arc after.

 

[Quickshot0]

The first week at least gives more time I guess? So I guess you're getting more time there. But once it's school time again hours will be harder to come by. Though one can argue she is still learning things at school, so hard to say it's entirely wasted.... hmm.

 

[PistachioCookies]

I just had this really terrible idea of having Amu set fire to her new school to get another 2 weeks of uninterrupted training time. If nobody knew she did it, she wouldn't need to justify anything. Which is a bad idea and just shows the bottom of the barrel I have reached in terms of coming up with ideas for more training hours now.

 

[Baughn]

Please don't burn the school, it did nothing wrong…

 

[Quickshot0]

Ah well, maybe we'll just have to accept we can't do all the things then. And there's always willpower buying... And maybe something else in future will come up to spend XP on as well aside of the quick buy option which probably needs things like integrity and such to make less risky.


Ah well, lets be positive and consider the positives of having more XP then one can easily use up. And for all we know things totally could be quiet for 2-3 months... maybe...

 

[PistachioCookies]

Ah well, maybe we'll just have to accept we can't do all the things then. And there's always willpower buying... And maybe something else in future will come up to spend XP on as well aside of the quick buy option which probably needs things like integrity and such to make less risky.

Click to shrink...

I mean we can - it's just the levels likely won't arrive in time for the next arc, probably only in time for the next one.

To ensure that they arrive for the first arc, we will only be able to buy 1-2. And Integrity likely won't arrive in time, since it takes 96 hours on its own (if weeks are 48 hours rather than 56) and apparently needs special care with training it, meaning the bulk of it probably can't be done at school. Moreover, if we buy more than a single 1 XP skill, we'll use too much XP to be able to buy Integrity as the emergency level-up.

If we don't mind playing the long game and going into the next arc without new levels, we can opt to buy in bulk now and get 25% discounted training times on all of them due to the overlap.

Or we can buy just 1 now (probably Illusion) and have it in hand. But obviously, there won't be any overlap time-discount on buying skills one at a time, so it will take the full 48 hours for it.

 

[Quine]

Emergency level-ups rely on us being in a situation where we "desperately need a specific skill or ability", and there's no guarantee we'll desperately need Integrity 3 during the next arc.

If things do go seriously wrong, in a way that Amu desperately needs Integrity to deal with, that probably means we end up hospitalized again. Other people might receive significant harm in that kind of situation, too. I hope we can avoid that.

 

[Quickshot0]

@PistachioCookies Indeed, so it's all a bit problematic on what to do. Though if we don't start on integrity now when we probably have the most periods of down time, when will we start on it?

Still if one has two or three skills partially done, then even if you quick buy one, then the cost and risk of it would apparently be reduced as well. Due to needing to lean less on an outside power I'd presume. So there is that aspect as well.


Most likely a perfect answer doesn't exist, so instead it becomes a balancing of risks. One can do a quick to learn power up and try to get through what ever comes up with that, and hope that no quick buy is necessary or that the side effects are mild. Or one can push for a longer term investment that hopefully will reduce risks and hope the current abilities are enough to get through any immediate issue for now, and else that the reduced risk/cost will get one through it with a quick buy.

So what does everyone think is the better strategy in this case? As I honestly am not entirely sure, both having to say for it. Gain a new ability quickly now to face what ever might come up, or a longer term investment for hopefully more gains later. The best one depends on what one thinks will happen in the next few months.

 

[ShadowAngelBeta]

I was about to come in and bonk people for trying to fire bomb stuff since Baughn just brought it up on our discord, but I see its in the name of training time.

Carry on.

 

[Quine]

The maximum possible overlap is 50%, but that's for—let's say—practicing fine-control telekinesis by lifting barbells at the same time you're training strength, and assume the presence of a gym that doesn't mind the presence of a preteen lifting hundreds of kilos.

The way Amu would train integrity does in fact need a lot of attention.

Normally you should be expecting around 25%.

Click to shrink...

So, for clarification, what does it actually mean to train N traits at once with X% overlap? What does that do to the training time?


Here's what I'd expect overlap to mean. It doesn't quite line up with some earlier statements, which is part of why I'm asking.

I'd expect 100% overlap (which is past the maximum) to mean 1 training hour reduces the remaining time for all N involved traits by a full hour, and 0% overlap to mean that 1 training hour reduces the remaining time for each trait by 1/N of an hour.

For values between 0% and 100%, I'd expect that X% overlap effectively gives the same values as if you'd spent X% of the time training at 100% overlap, and the remainder of the time training at 0% overlap. So training 3 traits at 25% overlap for an hour would give 0.25 + (1/3)*0.75 = 0.5 hours of training progress toward each trait, for an effective 50% training speed boost.

This is what makes sense to me, but it would also mean that when training Integrity 3 at the same time as Illusion 2, Illusion 2 would finish first. That doesn't line up with earlier statements about traits finishing simultaneously.

 

[Baughn]

So, for clarification, what does it actually mean to train N traits at once with X% overlap? What does that do to the training time?

Click to shrink...

It's the reduction in training time, as a percentage of total... um, a worked example might be best here. This will all be staggeringly obvious once there is some output to look at, I hope, so-

Let's say Amu is attempting to train Integrity 3 at the same time as Dreamwalking 2, and you've somehow stunted them into a 25% overlap. The base training time for I3 is 2 (Amu: 8) weeks (96 hours); for DW2, it's 1 (Amu: 4) weeks -- 48 hours.

It's dreamwalking, so Amu gets the base 2 hours of practice per day, plus 30 minutes while asleep. Not a lot, but any more than that would be dangerous given the nature of the Dreamlands. Doing it while fully lucid is a large safety boost. Note that the stunt means the base 2h/day is already 2.5h/day, even for the part of this that's practicing Integrity.

If she was doing this with 0% overlap, she'd need a little over 8.2 weeks total. With 25% overlap, it's 8.2 / 1.25 instead; about 6.5 weeks. With a hypothetical 100% overlap there'd be no penalty for doing two things at once, and the training would finish in 4 weeks...

... but that assumes all the training finishes simultaneously, which wouldn't happen. DW2 has half the cost of I3.

= = =

The best way to think about this is probably like the research screen in a 4X game. You get a certain number of "research points" (free time Amu chooses to spend on training) per week, which can be allocated towards one or more goals. You get bonus income for stunts that let her allocate more time, whether that's "training in your sleep" or "convince mom and dad she totally isn't traumatized, honest, practicing martial arts for four hours per day is just a hobby". You get a multiplicative bonus for spending the time more effectively, e.g. by doing something that trains two related skills at once.

So for the calculation above, it becomes:
- Base 2h of training
- + 0.5 hours at night
- * 125% for the 25% overlap
= 3.125 effective hours spent training Dreamwalking & Integrity

Which then gets split in half, for 1.5625 hours, per day, for each of the two level-ups.

With a total cost of 48 hours for leveling up Dreamwalking, that makes it 31 days before she's done. At that point the 'research income' reverts to being 2.5 hours per day, all of which is assigned to Integrity.

... and no, you can't balance that out by spending less time on one than the other. Mostly because this spreadsheet is already complicated enough.

= = =

However... realistically this is a quest, not a computer game. I don't mind seeing lots of discussion. But if you're trying too hard to min-max then you're forgetting about the first rule: Vote for the story you want to read. Stunts exist for that reason, not to let you min-max better, and I promise I'm not such a harsh QM that you'll regret leaving a few percentage points on the floor.

 

[Quickshot0]

So for instance... you can vote for the story where Amu is trying very hard to do what she can, because she feels like she wasn't good enough after the last disaster happened. And so she tries to overlap her training time a bit to make the most of it. Of course this means she's spending more time on training and less on friends though and might get a little frazzled by it all, but it's for a goal she believe in and she'll make it up afterwards.

Something like that?

 

[Baughn]

Precisely.

 

[Quine]

If she was doing this with 0% overlap, she'd need a little over 8.2 weeks total. With 25% overlap, it's 8.2 / 1.25 instead; about 6.5 weeks. With a hypothetical 100% overlap there'd be no penalty for doing two things at once, and the training would finish in 4 weeks...

... but that assumes all the training finishes simultaneously, which wouldn't happen. DW2 has half the cost of I3.

Click to shrink...

For two traits, this is equivalent to what I had in mind. For three or more, though, is it still just "/ 1.25"? If so, adding more traits to the training pile doesn't improve our efficiency unless it raises the overlap value, or gives us more training time.

However... realistically this is a quest, not a computer game. I don't mind seeing lots of discussion. But if you're trying too hard to min-max then you're forgetting about the first rule: Vote for the story you want to read. Stunts exist for that reason, not to let you min-max better, and I promise I'm not such a harsh QM that you'll regret leaving a few percentage points on the floor.

Click to shrink...

It's more the inherent harshness of the scenario that's driving me. This is, as your sig says, "A tale of the ending of the world."

That said, I'm definitely keeping the narrative in mind when optimizing training time. That's why my training concepts have leaned more toward "background" effort and interaction with Ami over high-intensity attention-splitting training regimes. I'm doing my best to keep Amu engaged with the people and environment around her while still training efficiently.

 

[Baughn]

For two traits, this is equivalent to what I had in mind. For three or more, though, is it still just "/ 1.25"? If so, adding more traits to the training pile doesn't improve our efficiency unless it raises the overlap value, or gives us more training time.

Click to shrink...

Indeed. Anything else would be exceedingly exploitable. If you can justify why it should increase the bonus that's another matter, though keep in mind that Amu is already going to be <REDACTED>.

It's more the inherent harshness of the scenario that's driving me. This is, as your sig says, "A tale of the ending of the world."

That said, I'm definitely keeping the narrative in mind when optimizing training time. That's why my training concepts have leaned more toward "background" effort and interaction with Ami over high-intensity attention-splitting training regimes. I'm doing my best to keep Amu engaged with the people and environment around her while still training efficiently.

Click to shrink...

That's fair, and I can't really say much other than "don't worry too much". I have a story to tell, which I can't tell if Amu dies.

… Saaya had no such protections, but at this stage there's a plot for her as well. I'm sorta used to things going better than expected, though.

One thing I will say…

The Conception, as displayed in canon, was entirely evitable. The underlying problem, however, was not. The world will end, but the date and method are still up for grabs. Anything else would be unfair.

It's a pity Uiharu isn't here. A flower-bedecked middle schooler would be just the thing right now.

 

[Quine]

Indeed. Anything else would be exceedingly exploitable.

Click to shrink...

Well, that changes the efficiency math substantially. We'll want to max out at overlapping 2 traits in the same training hour in most circumstances.

 

[Quine]

Hmm... hey, @Baughn, how much attention do we need to pay in class to avoid having our grades drop? Depending on how much attention that is, I think we might want to consider taking Lore 2.

Our training concepts so far have only made very limited use of class time for training. I think one of the big reasons for this is the idea that we need to pay attention to what we're learning. But what we're learning is mostly Lore, and we can learn Lore, or anything else, much faster than school teaches it.

Lore 2 is probably enough to learn the entire middle school curriculum, except maybe the language classes. Probably a good chunk of the high school curriculum, too. If we wanted to take it full speed, we could get it done in a week or two - we can easily justify plenty of time per day on it, but we'd quickly hit the point where our classes do nothing for us, and we might want to avoid conspicuously blazing through upper-year textbooks during class. (Maybe if we took Clairvoyance 1 at the same time, and tried to read those textbooks with our mind...)

After taking Lore 2, if we zone out during math or history class training psi instead, it doesn't hurt our grades, because we already know the entire next 3 years of material. (It wouldn't get us out of tests, or homework, though.)

Alternatively, we can slow-mode it. Just stay a few weeks ahead of the material in class - it'd take us minutes a day. We wouldn't reach Lore 2 any time soon, but we'd still have class time largely free for inconspicuous training. We could also slow-mode Linguistics 1 and stop paying attention in English class.

As another alternative, we could try to learn the entire middle and high school curricula and test out of the school system entirely, but I don't want the social consequences of leaving the school system right now.


The narrative of this has elements I like and elements that are kind of weird. The immediate narrative of this would probably be, Amu decides to buckle down and study ahead, so incidents she has to deal with don't set her back in class. She buckles down and studies, and somehow, it all comes to her so much faster than before. Terrifyingly fast - she couldn't even read a textbook this fast before, let alone learn the material! She slows down either immediately, if we slow-mode it, or after about a week or two if we go full throttle.

Either way, she's left wondering what her limits really are, when she puts her mind to something. I think I like having her wondering this. On the other hand, taking Lore 2 like this shines a big spotlight on just how utterly absurd these training rates are. This is far, far beyond Amu's canon academic performance. Any psion we control can boost Lore at the same rate, too, as long as they have the XP for it.

 

[PistachioCookies]

The Conception, as displayed in canon, was entirely evitable. The underlying problem, however, was not. The world will end, but the date and method are still up for grabs. Anything else would be unfair.

Click to shrink...

In the previous quest, Polaris was still around and JP's was still preparing for the Septentriones, at least before Amu partially blocked Conception. Whether it's Hell Week or Hell Month or all 7 Septentriones coming at the same time, I assume the Devil Survivor 2 plot isn't quite as evitable as stopping the cult from triggering Conception (triggering it early, anyway).

Well, that changes the efficiency math substantially. We'll want to max out at overlapping 2 traits in the same training hour in most circumstances.

Click to shrink...

....Not exactly, at least not with the example given. We apparently can get 30 extra minutes a night because we're doing Dreamwalking and Dreamwalking is doable while sleeping, but I don't imagine any other skill can be trained during that time period (no matter the effiency). Meanwhile, it is not doable during school hours (OTOH, skills like Illusion are trainable during those hours).

Since daily training hours are not individually allocated to skills but apparently pooled and averaged between all the skills being trained at once, doing Dreamwalking in addtion to just Integrity and Illusion (or in place of Illusion) will give us an extra 30 minutes of training time per day.

If our 2 skills were Illusion + Integrity then we'd be getting, say, 4 hours of training time a day if a stunt was able to open up 2 hours of school time for training. This is effectively 5 hours, after the 25% overlap bonus. This would make it take (48 + 96) / 5 = 29 days to complete.

On the other hand, if we swapped Illusion for Dreamwalking (assuming that we can still train Integrity in school hours) we're getting 4.5 hours daily instead, effective 5.6 from overlap. So those two particular skills take 26 days instead.

That assumes Integrity can actually be trained during school hours. If it can't, due to Amu having to pay special attention and it must be trained outside like Dreamwalking and forced to share the same 2 after-school hours, we might as well do Illusion at the same time too. Since if they are forced share the same hours, doing DW + Illusion is only going to give us 2.5 hours (3.1 w/ overlap, takes 46 days to finish) while adding in Illusion opens up an extra 2 hours during school and then we actually finish all three FASTER in 35 days, due to the extra time it opened up, even though the whole package now takes 196 hours instead of 144.

So, bearing in mind Amu should not be burning down any schools or going delinquent and torpedoing her character, academic studies or the narrative for the sake of pumping iron, this is my vote and the training stunt I'm going to propose:

[X] [Focus] Ami
[X] [Focus] Kana

[X] [Training] Plan: Do As Ami Does
-[X] Dreamwalking 2 - Ami needs some love and attention after the incident. Not from the counselor. Spend some bonding time with her leading up to bedtime every evening and maybe you'll also be able to figure out how her lucid dreaming powers work at the same time, and you can try and replicate it while you go to sleep every night too. Maybe Kana will also be able to provide some tips, given she seems to have some way to influence Yui's dreams whenever they get bad.
-[X] Illusion 2 - Ami pulled something out of dreams for your birthday, perhaps you can do the same and give her a reciprocal present on her next birthday. Also, all your school stuff got crushed along with your desk by the rift, so you now need new stationery. Fortunately, Miki allows you to pull functioning pens, pencils and other writing implements out of thin air. Make some before going to school every day and then see how long you can keep them working throughout school once separated from Miki. Though, you'd best hurry up and buy an actual set too, just in case they don't last too long.
-[X] Precognition 2 - Your capacity for schoolwork is less than 100% given everything that happened. This is not a problem for Ami, as Miki exists to help her cheat. Conversely, this means Miki is not available to help YOU cheat, but fortunately you have another solution. Much like Tsukasa and that not-so-quack fortune teller from TV, you believe you may also have some ability to see the future thanks to Dia, if nothing else. While at school, attempt to use Precognition to shore up your flailing academic grades by predicting the questions and answers to everything the teachers ask everybody, whether in-person or on the quizzes. While outside school, conjure a deck of Miki-sized playing cards for her and try to guess what she draws next whenever she randomly shuffles the deck.
-[X] Integrity 3 - It might be possible to speed up Saaya's recovery if you knew more about how she was afflicted. You can think of 2 experiences you have personally suffered that might come close - either the strain that comes from borrowing someone else's Chara, or the hollowness you feel from keeping Ran and Su out for too long. Miki and Ami always enjoy having Ran and Su out and about and you're fairly sure Utau wouldn't mind too much if you borrowed Eru for a little bit either - as much as you can manage each day, anyway. Angel's Cradle might help Ami not have nightmares about demons, if nothing else.

Hopefully, this opens up several hours from school and several more after school beyond the baseline 2, as I assume the 2 are coming from the 3-5 "school club" period of time and spending time with Ami before bed should be somewhere around 8/9 to 10/11. In addition to the 30 minutes of sleeping time from Dreamwalking.

If we are able to get 6 hours or so from this, we might actually get it all done in a month with 25% overlap.

 

[Quine]

Since daily training hours are not individually allocated to skills but apparently pooled and averaged between all the skills being trained at once

Click to shrink...

I don't think that's how it works. I think that's only for when we're actually overlapping training time - performing an activity that trains multiple traits in the same unit of training time. If we spend sleep time training just Dreamwalking, I don't think that gets pooled with whatever we do during the day.

Since if they are forced share the same hours, doing DW + Illusion is only going to give us 2.5 hours (3.1 w/ overlap, takes 46 days to finish) while adding in Illusion opens up an extra 2 hours during school and then we actually finish all three FASTER in 35 days, due to the extra time it opened up, even though the whole package now takes 196 hours instead of 144.

Click to shrink...

And this kind of thing is why it wouldn't work like that.

 

[PistachioCookies]

I mean according to this:

So for the calculation above, it becomes:
- Base 2h of training
- + 0.5 hours at night
- * 125% for the 25% overlap
= 3.125 effective hours spent training Dreamwalking & Integrity

Which then gets split in half, for 1.5625 hours, per day, for each of the two level-ups.

Click to shrink...

That seems to be how it's done.

I mean, how on earth could we be training Integrity while asleep? There's no way the half an hour we're getting during that time overlaps with anything other than Dreamwalking but we're still getting the 1.25 multiplier on it anyway.

 

[Quine]

I mean according to this:

That seems to be how it's done.

I mean, how on earth could we be training Integrity while asleep? There's no way the half an hour we're getting during that time overlaps with anything other than Dreamwalking but we're still getting the 1.25 multiplier on it anyway.

Click to shrink...

I believe that example was assuming we did find a stunt that allowed Integrity training in our sleep. After all,

With a total cost of 48 hours for leveling up Dreamwalking, that makes it 31 days before she's done. At that point the 'research income' reverts to being 2.5 hours per day, all of which is assigned to Integrity.

Click to shrink...

in that example, after we finish Dreamwalking 2, the 0.5 hours a day of sleep time go fully toward training Integrity 3, instead of cutting off.

But even if all training hours get thrown into a pool and counted equally toward all traits being trained, the overlap rating is still a measure of how much our training activity actually overlaps. If thrown in the pool, time spent on non-overlapping training would dilute our overlap rating. Instead of X hours at Y% overlap training whatever and 0.5 hours of sleep time training Dreamwalking at 0% overlap, we'd get X.5 hours of training at less than Y% overlap.