Having a Recipe for Random Pulls Is Just Common Sense, Right?
"Hmm."
...Resource income's been pretty lackluster lately.
More precisely, the farming yield had dropped off compared to how hard I'd been grinding in the absolute early game.
I'd figured it was because of the event prep, but why was it still the same even after the event ended?
...The protagonist kid's been making the girls do all kinds of things for a while now. Maybe they just don't have the time?
Whatever scheme he was cooking up, he hadn't told me a single thing about it.
...Well, he was effectively above me in rank, so I couldn't exactly refuse his requests.
Still, with Minty joining the team, the gap had mostly closed, and we were running at roughly the level I'd expected — so it wasn't a major problem.
"Take it a little easier."
"Ah, Aroma. Thank you. I'm almost done."
I took the hot chocolate Aroma handed me and savored the drowsy warmth for a moment.
Haah. The sweetness of these marshmallows was irresistible.
I didn't particularly like Jasmine as a character, but I had to give her credit for this one drink preference.
...Don't tell me they transmigrated me into Jasmine just because we share the same food tastes?
"I've got a rough grasp of everything. Now I just need to compare it against the plan and draw a conclusion."
"...Even if you say that, I have no idea what you mean."
"I was just talking to myself."
"Okay. Hang in there."
I'd already drawn up a plan before, but just in case, I needed to compare the resources earned from this event against my projections and calculate the margin of error.
Only then could I figure out how much spending would actually be advantageous right now.
And once that calculation was done, I'd run the win-rate projections for the Deadly Battle within those limits. If the win rate barely changed, I'd cut spending on the parts that weren't flush with resources — that sort of thing.
I had to invest resources with cost-efficiency in mind.
Level 50 for everyone should be fine, at least.
The current level cap was 50; it would unlock to 70 during swimsuit season. The event resources would scale up accordingly.
So dumping levels into a character I wouldn't be using later — namely, "Jasmine" — up to around level 50 wouldn't have much of an impact.
From a farming standpoint, it might even be better.
Dungeon farming difficulty ramped up the more lower-tier dungeons you cleared. Eventually you'd have to raise even the characters you weren't using in the Deadly Battle.
So pushing every currently available character to level 50 was completely fine.
There weren't that many collection items yet, so I'd already set those up with the standard newbie loadout — nothing to touch there. Affection Gifts were being fed in as soon as they accumulated, more or less.
Except for the ones that cost Choco. Those I was stockpiling, just in case.
Messing with people's memories is... kind of a lot, you know?
Having lived through it myself made me feel it all the more keenly. I absolutely did not want to put any of the others through that experience.
...If things got truly desperate, I might have no choice. But not yet.
"Alright. This is about the right amount of resources to invest this time."
Between the event and the Sugar that had piled up from not investing for a while, there was plenty. Enough to put toward some real power spikes.
Most of it would go all-in on the DPS units, but investing in Fruity — who had citizenship rights — was a solid choice either way.
"Master, I'll start the upgrades with Fruity."
"Okay."
"Me?"
"Yes, Fruity. Come over and take your clothes off."
She was still only 2-Star. Getting close to 3-Star, but not in time for this Deadly Battle.
So the shortfall in her specs had to be made up through other upgrades if she was going to fully pull her weight as a main DPS.
"Let's pick up the Wing Reinforcement from last time."
Last time, the DPS units had stopped at +8. This time, I planned to push Fruity all the way to +17.
Flight ability improved every five levels of reinforcement, so for a main DPS who needed to stay alive for a long time, the boost to survivability made it very efficient.
"There's a fail-safe every five levels, so just getting to +15 shouldn't be too hard."
Of course, the higher the reinforcement went, the more Sugar it burned and the lower the success rate got. But +17 wasn't +22 — it was doable.
...Though if luck truly hated me, we might end up settling at +12.
"Sniff sniff. Something smells delicious."
"...Yeah, it does."
I had no idea why a mere visual effect reeked of dalgona candy. If they were going to go that far, they might as well have made it edible.
It was unfair that only Cotton Candy — the skill resource — was something you could actually eat.
Granted, reinforcement happened so many times that the volume would be ridiculous to consume, but still....
Ugh. That half-gold crack-and-shatter failure animation was genuinely infuriating. They really nailed the reinforcement-fail effect, I'd give them that.
"Oh, it finally stuck."
It landed faster than my expected-value calculations, so that was a win. A massive amount of Sugar had evaporated, but the stats had skyrocketed in proportion to the investment.
...Of course, that wasn't the end of it.
"If you look at the Wing Reinforcement options, there should be an Output Efficiency Reset."
"Okay."
"Try that once."
Wings had something called Output Efficiency — an additional reinforcement that determined what percentage of the stats gained through upgrades actually applied.
...These people were truly deranged about money.
Anyway, it was currently at 56%, so it needed to be pushed higher than that.
"Mmm...."
If reinforcement felt like stamping dalgona candy and pulling a lottery, the Output Reset was a cute little effect where you baked a dalgona bun and the percentage rose based on how much it puffed up.
...They really did nail the charming presentation on stuff like this.
What a trash game.
Only after dumping Sugar like crazy did I barely scrape together 82%. To someone who normally ran 90% or higher, it reeked of mediocrity — but by that standard, +17 wasn't satisfying either, so I decided to leave it there.
"Equipment's fine at this level...."
Now all that was left was raising the Support Skills.
...Come to think of it, I'd never explained Support Skills to him.
They burned so many resources that there'd been no reason to use them until now, so I'd completely forgotten.
"Master, the lower part of the tattoo on your back. Yes, there. I'm going to reinforce that."
It was the section where three lines of passive skills were inscribed. The abilities written there were Support Skills — additional powers randomly assigned when a character was unlocked.
You could reroll them by spending one Sugar Cube. They started at a low grade — basically junk tier — and each attempt had a chance to raise the grade. Once a grade went up, it never dropped.
Sugar Cubes could be crafted from Sugar, but there was a production limit. Ever since transmigrating here, I'd made converting Sugar into Sugar Cubes my absolute top priority. So by now, I had enough stockpiled to fully kit out one DPS unit.
"First, we roll each one up to the highest grade — Platinum."
After that came the process of hunting for the right options at random. Honestly, finding the right options after hitting Platinum cost more than getting to Platinum in the first place.
Anyway, we had to get that far before the real conversation could start.
"Like this?"
"Yes. When it lights up brightly, that means the grade went up."
From Bronze to Platinum, three flashes total and the grade was fully raised. Getting there was just a matter of mindlessly feeding in Sugar Cubes — nothing difficult about it.
The real problem was sniping the desired options....
"Master, of the options you're seeing as you roll right now, which one looks the strongest to you?"
"...Critical damage?"
"Hmm. You're understanding this a lot better than before."
Among these random options, critical damage actually existed — and unlike the others, it was extraordinarily hard to pull. In other words, it was the highest-efficiency option available.
Same-stat bonuses are additive, after all.
Say attack power increased by 20%. If you already had other attack-boosting skills totaling 300%, the actual damage increase was only about 5%.
But critical damage? Even with stacking boosts, the total rarely exceeded 100%. In that case, when a crit landed, the real damage went up by 10%.
Crits didn't land every hit, of course — but critical rate climbed over time, and the fact that your hardest hits hit even harder made the efficiency climb considerably.
So for those reasons, we needed to pull critical damage....
The problem is this vicious game set the critical damage drop rate absurdly low.
Support Skill appearance rates were not equal. Even though you could spend extra Sugar Cubes to lock in a good skill, this was the reason countless players had burned through fortunes.
A proper skill had less than half the drop rate of junk stats, and top-tier skills like critical damage dropped far less often than even those proper regular skills.
But....
If what I'd experienced in the last Deadly Battle was any indication — if bugs worked in reality the same way they did in the game — then wouldn't the Support Skill recipe bug work properly too?
"If an option with critical damage shows up, stop for me."
By default, Support Skills never duplicated. All three had to be different skills. So from the second option onward, you had to think carefully.
You might assume you could just exclude already-rolled options and keep spinning, but that would mean rewriting the entire pull list — a heavy load on the server.
So Princess Fourmi used a method of splitting the pull list based on already-drawn skills. You picked which side of the split to pull from, then rolled within that list.
For example, if the pool was [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] and you pulled number 2, it would split into [1] and [3, 4, 5, 6, 7, 8, 9, 10]. You chose which side to pull from, then rolled within that side — guaranteeing no duplicates.
That much was cleverly designed. The tiny mistake that exploded from there was the problem.
To implement that system correctly, you needed to sum the internal probabilities for each split list. That was how the probability of selecting each list got set properly.
...And they'd left that critical step out.
Meaning the chance of picking either [1] or [3, 4, 5, 6, 7, 8, 9, 10] became a flat 50% each. Naturally, skill number 1's appearance rate skyrocketed.
...And if you spent one more Sugar Cube, you could reroll everything except option 2 — and wouldn't you know it, number 1 just happened to be the jackpot skill?
It was absolutely insane....
Watching the entire community go up in flames back then, I'd genuinely thought the game was finished.