#math!!

12x = 55 (1 - 3)

12x = 55 (-2)

12x = -110

x = -110/12

x = -9,1666... = -9 1/6

:)

Hi! Do you want niche calculations? SO, in Fire Emblem Awakening, the growth rates of child units are decided by the formula of

(Father's growth rates + Mother's growth rates + child's base growth rates) / 3

This means there are a LOT of possible variations for every kid. One kid in particular has more than most: Morgan.

This is because of Morgan's fixed parent: Robin.

(God I fucking hope I got my numbers right, I got a spreadsheet and everything)

Robin's growth rates

Robin is subject to the Boon/Bane system. Upon character creation, players can select one stat Robin excells in, and one they suck in. There are 8 stats, and a boon and bane cannot be of the same stat. These boons and banes will actually alter Robin's growth rates with up to 15% on boons and -20% on banes.

The total number of Robin's possible growth sets is simple: 8*7 = 56

8 is the number of possible boons, and 7 of the banes said stat can then have. So 56.

Morgan's possible secondary parents

In addition to Robin's many growth rates, as avatar, Robin can also get married to every single playable character in the game (excluding Einherjar).

For a male Morgan, this includes 24 possible fathers.

For a female Morgan, this includes 23 possible mothers.

So the growth rates are just 56*24 or 56*23, right? Haha. No.

Second Gen Units

Robin can also marry other second generation characters. Ethics of that aside, it adds another cinch to Morgan's stat set. Because the Second Gen Units also have flexible growth rates, calculated with the same formula at the start of the post.

We are, of course, excluding Robin as a potential parent for all these kids, because they are married in this scenario.

There are 6 male 2nd Gen Units, and 6 female 2nd Gen Units.

All 2nd Gen units who can optionally be Chrom's children (Inigo, Brady, Kjelle, Cynthia) have 13 possible fathers, all others 12.

The only outlier here is Lucina, who only has 5 possible mothers for plot reasons.

For male units = 12*4 + 13*2 = 74 For female units = 12*3 + 13*2 + 5 = 67

So there are 74 possible growth rate permutations for male 2nd Gen Units, and 67 possible growth rate permutations for female 2nd Gen Units.

Total possible fathers = 74 + 14 + 4 = 92 Total possible mothers = 67 + 15 + 2 = 84

This means the total possible father variations for male Morgan are 92, and the possible mother variations for female Morgan are 84.

The actual numbers

For male Morgan:

56*92 = 5.152

For female Morgan:

56*84 = 4.704

TLDR: In total, for both Morgans combined, there are 9.856 possible growth rate sets.

That's all.

note that I won't get my score for another month, so I can't exactly back this up BUT here ya go

know the inferences. you'll have the long part of the formula, but you need to know to take p-hat plus or minus that long part (as an example with a one-sample z interval)

"long parts" that are given on the formula sheet:

(I'm not going to add alt text, just because stats formulas are so confusing to describe without symbols, but please let me know in the notes or dm me if you need me to--I'll send links to pages with the formulas or something. we'll figure it out 👍)

you have to know what to do with these to get each formula. the first thing I did when I opened each booklet was wrote extra formulas for each inference test: one- and two-sample z- and t-intervals and tests (eight formulas when I was done)

remember that the t-table is for confidence intervals and the z-table is for p-values (you will have to use the t-table for z-intervals and the z-table for t-tests) (note--the actual difference is that the z-table assumes you know the population standard deviation, where the t-table uses a sample's SD. The z-table is therefore used with an assumed null hypothesis, i.e. z- and t- tests, while the t-table applies to z- and t- intervals without a null hypothesis). The table is in the back of the booklet for the multiple choice, and it does say "stop" at the bottom of the last question. you are allowed to flip past to the tables.

hypotheses for proportions (z inferences) use p = __ ; mean hypotheses (t inferences) use mu (the weird "u" looking thing; mean of the whole population). do not put x-bar; that applies to the mean of a sample and not population.

also know what your p-value means: it's the probability of obtaining the data you used in the calculation (or more extreme--don't forget "or more extreme") assuming the null hypothesis is true

write out your conditions and show them being met. even write that 60 is greater than or equal to 30 if n=60 and you need to prove that it's an approximately normal distribution (note approximately and not exactly)

put your conclusion in context. mention what's being tested, not just numbers

type I and type II errors are something you'll want to put some time into--they were on both my 2021 exam and the 2012 one ap released as a practice. you'll also want to know the chances of each one happening (one is the significance level used and the other is some weird formula, and I don't remember which is which)

type I error: reject the null hypothesis when you shouldn't have

type II error: fail to reject the null hypothesis when you should have ("fail II reject")

- on that note, never accept the null hypothesis when writing conclusions, just fail to reject it

regression lines and "model error" came up and I had no clue about either. I still have no clue if "model error" is even a statistics thing. neither were in my free response section, thank fuck, but regression lines and that number summary thing your calculator gives you was in the free response for 2012, unless I'm mistaken

chi-tests came up twice on my test: on one multiple choice it was asking what test was appropriate with the given circumstances, and a chi goodness of fit test was an option. the second time was in the final free response question, and it was the chi-test about correlation between two variables. I got lucky, because it didn't actually want me to run one and I could bullshit my way through the question, but make sure you know them. put z- and t- inference tests at a higher priority, though

one of my free response questions was entirely free of calculations; it wanted me to explain the purpose of placebos, the importance of random sampling and assignment, etc. I think I got lucky again there, but make sure you have a firm understanding of experimentation procedure

you'll probably be fine without worrying about probabilities of events and compliments and independent v. dependent events and everything (PA, PB, PA union B, those things). they were a free-response on mine and I drew a flow chart and had no issue whatsoever, it was really simple. this one probably shouldn't be a priority unless it's something you personally struggle with

spend a bit of time with transforming data. ex: mean 1 foot, sd .5 foot --> mean 12 in, sd 6 in

last but far from least:

theyre definitely not trying to get drunk ahahaha theyre partially on duty in that pic ahahahaha

probably a handle between the two of them? about 20 shots each should do the job

a1- high alcohol tolerance

—I’d be here 6 years writing all that fic!