CHAPTER 80: Hollowvania (Patreon)
Content
CHAPTER 80: Hollowvania
Tomás looked at the map of Hallownest in the level editor and calculated the possible routes.
Fifteen zones. Each zone connected to at least two others. The player could obtain movement abilities in different orders depending on which zones he explored first. Dash in Greenpath or in Fungal Wastes. Double jump in Crystal Peak or in City of Tears. The Dream Nail in Resting Grounds, which was fixed, but which could be reached from three different directions.
The route combinations weren't infinite but they were numerous enough that testing each one was impossible with three people. Tomás knew it the first week of November when he sat down with his red notebook and started mapping the sequences and realized that there were more than four hundred viable routes from Dirtmouth to the Black Egg Temple.
He couldn't test four hundred routes. But he could test the ones that mattered.
Tomás developed a methodology he called paths: predefined routes that stressed specific combinations of abilities and zones. The standard path was the route the average player would follow: Crossroads, Greenpath, Fungal Wastes, City of Tears, and onward. The inverted path was the most counterintuitive route: Crossroads, Crystal Peak, Resting Grounds, Deepnest before having the dash. The speedrun path was the shortest route to the end: the minimum abilities needed in the least possible time.
The sequence bugs appeared the second week. If the player obtained the dash before the double jump, a section of Crystal Peak where Elena had assumed the player would have both abilities became impossible: a high platform required double jump to reach, and without it, the player got stuck. George corrected it by adding an alternative route with nail bounce that didn't require double jump.
Another bug: if the player entered Royal Waterways before obtaining the long-breath Charm, a section with sinking platforms had a timing so strict that it was technically possible but humanly improbable. Tomás completed it on the fifteenth try. George added an extra half second to the sinking timer for the platforms in that section.
Another: if the player defeated Hornet 2 before the Mantis Lords, a line of Hornet's dialogue that referenced the Mantis bow made no sense. George added a conditional variant of the dialogue that changed according to which bosses the player had defeated.
Forty-three sequence bugs in six weeks. Each one was a combination George and Elena hadn't foreseen but that a real player could find. Each correction made the map more robust, more flexible, more capable of supporting the freedom George had given the player.
*
Daniel knew City of Tears was going to be a problem on PS2 from the day George implemented the rain.
Five hundred particles with splash, dynamic puddles, and reflections with sinusoidal distortion. On PC, the system ran at 30 fps with margin. On PS2, the Emotion Engine choked at 18 fps, which was unplayable.
Daniel had the texture-strip trick he had planned from the start: instead of five hundred individual particles with their own physical behavior, the PS2 version used eight vertical strips of semitransparent texture that moved at different speeds, simulating the rain as curtains of water instead of individual drops. The strips were four times cheaper on CPU than the particles and the visual result was convincing at first glance, although it lacked the individuality of the drops of the PC version.
The puddles were simplified to static tiles that didn't accumulate water dynamically. The reflections were eliminated on PS2, Xbox, and GameCube. George accepted each simplification with the philosophy he had established during the development of City of Tears: better an extraordinary PC version and good console versions than four mediocre versions.
Daniel completed the ports in eight weeks. Three consoles simultaneously. It was his fourth port cycle for NGS and the efficiency of his abstraction framework made each new port faster than the last. The platform-specific code was reduced to the rendering and audio modules; everything else was shared.
"Every port I do is easier," Daniel said. "But City of Tears gave me gray hairs on each one."
*
Marcus painted the cover of Hollow Knight in watercolor.
On paper, not digital. With brushes and water and pigment, the same technique he had invented for the Greenpath tiles but applied to a large composition: a sheet 30 by 40 centimeters that Marcus worked on for a week in his apartment, not at the office, because the cover was something he needed to do alone, as he had done the Sheol tiles for Isaac.
The image was the Knight standing in the rain of City of Tears, looking upward. Toward the ceiling of the cavern from which the rain fell, away from any enemy or objective, as if he were searching for the sky he had never seen. The drops of water fell on his mask and slid down his body like tears that didn't belong to him. The buildings of the city blurred behind him, visible but not entirely, like Sam's background but condensed into a focal point.
Marcus scanned it at high resolution and sent it to George without comment. George opened it on the screen and looked at it for a minute before replying. Not because he was evaluating the quality but because he was processing the emotion: the cover of Castlevania had been grandeur, Alucard in front of the castle with the full moon. Isaac's had been intimacy, a child alone in the darkness. Hollow Knight's was melancholy: a small being in the rain looking at a sky that doesn't exist.
Three covers. Three games. Three emotions. And each one had been painted by the same person who understood that the cover presented the feeling, more than selling the game.
*
Hollow Knight went on sale on February 17, 2004, on PC, PlayStation 2, Xbox, and GameCube.
It was NGS's first simultaneous quadruple launch. Claire had coordinated the logistics with Sony, Microsoft, and Nintendo for three months: gold masters sent, approvals obtained, dates aligned. It wasn't trivial: each manufacturer had its own certification process, its own deadlines, its own technical demands. Claire handled them with the efficiency of someone who had negotiated with all three and who knew exactly which buttons to press and when.
George didn't go to Best Buy this time. He stayed in the Townsend Street office at eight in the morning, looking at the email where Claire had forwarded him the confirmations from the four distributors: the games were on the shelves. On four platforms. On three continents. Hollow Knight existed in the world.
The first reactions came that night, from the forums where the most avid players had already played several hours:
"I can't believe this is from the same studio that made Isaac. It's as if Hightower were three different people."
"It's the most beautiful game I've ever played. Period. Not ‘for an indie game’. Not ‘for a 2D game’. The most beautiful game. Period."
"I got lost for two hours in City of Tears because I didn't want to leave. I literally sat on a bench in the game and listened to the rain. What is happening to me?"
"Deepnest can go to hell. And whoever made Nosk can go with her."
"Don't buy this game if you don't want to cry. Don't buy this game if you don't want to lose 40 hours of your life. Don't buy this game if you value your mental health. But buy this game."
*
The reviews came the first week of March.
IGN gave it 9.5 out of 10. The review was six pages and a tone that was less professional criticism and more a declaration of love: the reviewer described his first hour in City of Tears with the same reverence with which George had described his own when he saw the rain work for the first time. The last line was: "Hollow Knight is the masterpiece of New Game Studios and one of the best games ever made."
GameSpot gave it 9.3. "The best exploration game since... well, since Castlevania from the same studio. Hightower has perfected what he invented."
Next Generation gave it 5 stars out of 5. Rachel Kim, the journalist who had followed NGS since the first Castlevania, wrote the review. George read it standing in the office, with the team gathered around his monitor as they had done with the Castlevania reviews five years earlier. The review opened with a sentence George reread three times:
"George Hightower hasn't improved the Castlevania-like. He has transcended it. Hollow Knight isn't an evolution of the genre Hightower created but the end of that genre and the beginning of another."
Edge gave it 9 out of 10. The British review was more restrained but equally definitive: "Hollow Knight isn't a Castlevania-like but something new. And that something new needs a name."
Claire printed the reviews and hung them on the Townsend Street wall, next to the new mural Marcus was painting: an expanded version of the Brannan Street mural that included the castle, the Isaac basement, and now the Knight of Hollow Knight surrounded by rain and leaves and darkness, connected to the other two worlds by roots that descended into an abyss that still had no name.
*
The name appeared on a NeoGAF forum three weeks after the launch.
A post by a user named hallownaut that became the most discussed thread on the forum for a month:
"Castlevania invented the Castlevania-like. A 2D exploration game with an interconnected map, backtracking by abilities, and action combat. Isaac invented the Isaac-like. A procedural generation game with items that accumulate, permadeath, and infinite replayability. Hollow Knight is neither of the two. Hollow Knight is a Castlevania-like but not as we know it: it's bigger, more open, more melancholy, with combat systems (parry, soul, charms) that didn't exist in Castlevania. If Castlevania is a castle, Hollow Knight is a world. We need a new name for what Hollow Knight is. My proposal: Hollowvania. The evolution of the Castlevania-like. The genre Hightower created by transcending the genre he had already created."
The post generated five hundred replies in two weeks. Half agreed. The other half proposed alternatives: "open-world 2D exploration," "Hightower-like," "action exploration RPG." But Hollowvania had the advantage of simplicity and reference: it contained the name of Hollow Knight and the heritage of Castlevania in a single word. It was easy to say, easy to remember, and easy to understand.
The publications adopted it within weeks. IGN published an article titled "Hollowvania: Is Hollow Knight a new genre?" that argued that yes, that the combination of open map, combat system with parry and Soul, environmental narrative, and authorial aesthetic created an experience that couldn't be classified as a simple Castlevania-like. GameSpot used the term in a video. Edge mentioned it in an editorial about the genres a single studio was creating.
Three games. Two genres: Castlevania-like and Isaac-like. And Hollowvania, the evolution of the first.
George read hallownaut's post and the articles that followed with the same mix of wonder and guilt he had felt when the term Castlevania-like appeared for the first time six years earlier. Wonder because people recognized what NGS had done. Guilt because part of what NGS had done came from memories that weren't his.
But the guilt was smaller than before. With Castlevania, 95% was memory. With Hollow Knight, 85% was invention. Each game reduced the debt. Each game increased the creation. And Hollowvania, the evolution of the Castlevania-like that bore the name of a game George had invented almost entirely, was more his than anything he had done before.
The guilt was still there. But it fit in a pocket.
*
Claire presented the first-week numbers with an expression George hadn't seen on her before: that of someone who has spent four years making conservative projections and who has just discovered that reality exceeded every one of them.
"First week: three hundred fifty thousand copies across the four platforms."
It was more than any previous first week of NGS. Castlevania PC had sold eighteen thousand in its first week. Isaac PC plus PS2 had sold a hundred thousand. Castlevania PS2 had sold a hundred fifty thousand. Hollow Knight had sold three hundred fifty thousand because it had four simultaneous platforms, five years of NGS reputation behind it, and a pitch that sold itself: a beautiful exploration game from the studio that invented the Castlevania-like.
"First month: one million five hundred thousand."
"Projection for June: three million."
Claire closed the notebook.
"I'm not going to make a projection for the end of the year because every month I come up short. What I will say is that Hollow Knight is going to be NGS's best-selling game by December, and that if the curve holds, it's going to surpass Castlevania in total sales before 2004 is over."
George nodded. The sales confirmed what the reviews and the forums had already said: Hollow Knight was the best thing NGS had done. Not only in quality but in reach. Castlevania had been a niche success that grew. Isaac had been a divisive experiment that found its audience. Hollow Knight was both things at once: a game the critics loved and the public bought, that was art and that was commercial, that made people cry and that sold copies.
It was the game that proved NGS wasn't an accident. That the success of Castlevania wasn't beginner's luck and that Isaac's wasn't an anomaly. That George Hightower and his team of eighteen people could create worlds people wanted to inhabit, again and again, and that each world was different from the last and better in ways that couldn't be measured with numbers but with the hours the players spent walking in the rain through an empty city listening to a piano that sounded like what remains when everything is gone.
End of Chapter 80