Ultimately, Goblin Cave was a dungeon. As far as it was aware, its mechanical purpose was to transduce mana from system-space into physical space, converting some measure of that into potential experience in the form of mobs, which was then harvested by adventurers. There was an interplay here, where experience could be reaped back and forth between dungeons and adventurers, with greater success for either leading to further experience gains. Adventurers, presumably, died in other ways, and their experience was further concentrated down. Was there a second-order form of experience, too, the way mana could be used to instantiate experience-bearing constructs?

Its own feelings on the matter did not appear to matter. This was the machinery of the world, cogwheels endlessly crushing and grinding, transforming resources from one to the other, slowly sieving out grains of mana, experience grit, and piling them up in endless heaps: an idiot machine, operating to unknown purposes.

The next day, the surveyors returned and said they would return again in a month, with a collection of books. Goblin Cave had another awkward, perfunctory conversation with them, made all the more awkward by the midnight visit warning it about them. This gave it a blissful 28 days when it could watch its experiments unfold and didn't have to think about them. If there was going to be a shipment? It was very unclear about the time scales involved— would it be betrayed immediately? Did it have a month? A year? Ten years? What would a betrayal against it look like, aside from adventurers showing up to shatter it? What could it do to avoid that, aside from continue to do everything it was already doing?

It was thrown all out of its habits. Even the brief conversations with adventurers were stressful and disorienting. It wanted...

Well, it still wanted to restructure its old floors. But that impulse was undercut by how the last time it visibly changed, it had immediately attracted what appeared to be an enormous amount of trouble that it still couldn't fully conceptualize. But as time went on and it dug further into the mana theory book, and further into conversation, the itchy frustration of not making progress came to outweigh its fear. Again.

The secret of its mana funnel was that it had never really been intended to capture adventurer's mana. There was the world's mana, constantly colliding with its own at its entrance, and it was very easy to construct a mana funnel exposed to the surface, and then —

watch it fail to function at all. It was something about pressure: ambient mana wasn't directed enough to be scooped up by the grooves of the funnel the way a directed spell effect was. It changed the curve of the funnel, reshaping the spacing of its grooves, and finally managed to get a paltry seeping of mana, something in the range of a single point of mana a day... if it could capture it, which it couldn't. If it could... then, it would just have to wait a month or two, somehow storing all the mana all that time, and then somehow energetically release it, which it also couldn't do. But if it could do all that, then maybe it would be able to spawn a single flame wisp 'naturally', on command. Once it understood the precise way to optimize the eddying vortices that spawned a flame wisp... well, then it would be on to figure out how to spawn, say wind shades.

That time scale seemed absurd now. Months for a single experiment? Years to refine things? But it had done the math.

It wasn't sure what it was getting out of doing it, but it continued to bribe adventurers with simple metal ingots in exchange for short conversations. It could manage a conversation with an adventurer daily, but that always left it drained and exhausted, unable to concentrate properly on its theories or experiments. If this was how all dungeons felt after talking, it understood why most dungeons apparently didn't speak.

But even knowing it was being impatient didn't stop it from writing a new question, in its talks: WHAT WOULD YOU CHANGE ABOUT THIS DUNGEON, IF YOU COULD? It wasn't interested, precisely, in revising itself to their specification, but it was attempting to understand the other, the point-of-view of something not-itself. They had a different context to itself, a different set of priorities. They were capable of drawing comparison between itself and its fellow dungeons in ways it had never considered on its own.

(It was comical that one of the more common answers was more monsters, spawning faster. That was something it couldn't do, even if it wanted to.)

Apparently goblins did not efficiently exchange mana for experience; that was one thing it learned. Not all mobs of the same tier gave the same experience payout per level. It had suspicions to that effect: some of its mana goblins ate fish, and they all ate fungi. Dungeon mobs all, and they killed the fish the same way they killed the mana puppets, yet the experience received was markedly different. Fungi was worth almost nothing, fish were worth a pittance. So despite creatures of the same tier having mostly-similar mana costs, the experience outflow was different. Goblins, it had presumed, would be worth more than fish; it couldn't say how they ranked versus the mana puppets. But apparently all the nearby dungeons (save Deepmine Delve) had mobs with better effort-to-experience ratios, even if they were more deadly at lower floors.

Inefficiencies added up throughout the process: mana cost to spawn, mana cost to level, soul volume, tier-related leveling costs, respawn cost, experience ratio. Effort for adventurers to slay them, danger for adventurers to slay them, Goblin Cave grudgingly added to the list, considering the viewpoint of the adventurers. Time taken to kill a mob. Those were the factors that had lead to its pecular delvings: Goblin-types increased in power slowly, but they still increased faster in power than they increased in experience gain. There was a point where the lines intersected, where delving deeper into a goblin dungeon became both profoundly unrewarding and incredibly dangerous. That was interesting, because that seemed to be the ideal situation for it. It didn't want adventurers delving deeper. So why did any other dungeon bother to populate itself with attention-grabbing, high-reward monster types? To kill adventurers, certainly, and consume their experience the way the adventurer had consumed its mobs. To increase in power, and avoid destruction. But the threat of destruction came from adventurers delving it. So...

(Well, one answer might be the rankings. "Difficulty" was a category. As far as it had seen, the rankings were only information. Were there rewards for reaching certain tiers? Or punishments for falling? Its rankings had fallen fairly precipitously as it had continued to talk to adventurers, but there was still a ways to go. What were the bottom rankings? How did they manage to maintain such low scores? It could see the top ten per category by name, for all that helped it, but there was no similar ranking for the bottom ten. What chaos had they wrought? What metric was being ranked here? Interactions with adventurers certainly seemed to be an influence on it, but the adventurers themselves were not consciously adjusting the rankings. So what was?)

This was all interesting, partly as a curiosity, partly as a mystery, but it naturally lead to a question: if its mobs were inefficient experience conveyors, what was the most efficient conveyor? It would be another microcosm: the mechanical purpose of the world, stripped bare. The most potent confluence of energies, transduced through system layers with minimal losses. A hellish box, every flow blasted wide, every mechanism pushed to its limit. Mana howling as it twisted into shape. Gears rattling as they vibrated themselves to pieces. Machinery screaming, pushed to its breaking point. The optimal configuration. An experience grinder.

Obviously, it would take experimentation. The way adventurers talked, it sounded like they thought the experience ratio of a given mob was determined by its type: goblin type, avian type, elemental type. Goblins, categorically, had a poor experience ratio. Fungi must be near-nonexistant. Fish may be very low. But there was some form of tier influence, perhaps? The adventurers certainly didn't seem to know for sure. They didn't appear to want to talk specifics of their system screens.

There was only one way to really test things. Its new habit of extruding mana locks into its old dungeon somewhat prepared it for what it would have to do next: it dug out new hallways from its second and third floors. First, in manastone, out of habit, and then with a thought of how obnoxious it would be if these got stripped too, it restructured them to use flat off-white quartz, with craggy outgroppings of granite poking through at erratic intervals. These hallways led to a transverse hall, which had a collection of square doorways along the far side. Each lead to a short, right-angle hallway, which opened up into a cubic chamber.

Goblins. Beasts. Goblin+Beast. Fungi. Plants. Fish. Elementals. Elemental Constructs. Undead. Flying. Aberration.

It spawned one example of each, each into their own chamber: a stark cubic room with all surfaces made from hard quartz. The fish required a recessed pool, which it fed with a long, thin slit midway up the far wall. For plants... it had hardly touched the category. It had cavegrass (t0, light) and chokevine (t0, death); it chose chokevine, and gave it a rocky outcrop to anchor to. Aside from that... a fresh [Goblin], for its goblin type. [Wolf]. [Beastkin Goblin]. [Brown Mushroom]. [Chokevine]. [Shadow Darters]. [Flame Wisp]. [Lesser Mana Puppet]. [Biting Skull]. [Bat]. [Mana Goblin]. One of each from each category page it had unlocked, aside from Orc and Ogre, which it didn't have low-tier mobs unlocked. On a plaque protruding from the hallway edge, it wrote: KILL THE MOBS AND DETERMINE THE BEST EXPERIENCE RATIO. WRITE DOWN YOUR LEVEL AND EXPERIENCE GAIN FOR EACH. It provided a basket of graphite sticks to do the writing.

This was too many mobs for an early floor; it was paying a substantial spawning penalty for that. It was also too many souls; it hadn't linked these mobs into the churning soul whirlpool that the rest of its upper floors were linked into, so these would take a week, potentially longer to respawn, once killed.

If this was what a dungeon was designed to do, it would construct a mechanism to perform that designation perfectly, and hopefully that would give it... some kind of insight.


About the author



Log in to comment
Log In

Log in to comment
Log In