The Making Of…

How Halo and clas­sic anime in­spired a Rogue­like shooter with the bright­est of en­e­mies

EDGE - - SECTIONS - BY MATT CLAPHAM & BEN MAXWELL De­vel­oper/pub­lisher 17-Bit For­mat PC, PS4 Ori­gin Ja­pan/US Release 2015

How Halo bat­tles and clas­sic anime in­spired 17-Bit’s smart Rogue­like shooter Galak-Z

Like any am­bi­tious space mis­sion, Galak-Z’s launch was fraught with peril and the po­ten­tial for dis­as­ter. The game rep­re­sented 17-Bit’s dif­fi­cult sec­ond al­bum fol­low­ing the suc­cess of Skulls Of The Shogun, and one that packed re­mark­able and po­ten­tially prob­lem­atic enemy AI be­hind its at­mo­spheric 2D vi­su­als. It didn’t go smoothly. A litany of bugs and tech­ni­cal gl­itches blighted Galak-Z’s release, forc­ing 17-Bit’s small de­vel­op­ment team into three weeks of dam­age lim­i­ta­tion as they tried to get their new ves­sel back on course.

“There was a bunch of stuff com­ing in that we’d never seen be­fore and we were freak­ing out, try­ing to patch it im­me­di­ately,” 17-Bit founder Jake Kaz­dal tells us. “So it was kind of a bumpy start. But we fixed it, and the re­views were fan­tas­tic. We made some really dar­ing de­sign choices, so it was grat­i­fy­ing to see that, for the most part, I think peo­ple really liked it.”

Those dar­ing de­sign choices had in­aus­pi­cious, but for­tu­itous beginnings. Kaz­dal had fin­ished his work on Skulls and was now fo­cused on mar­ket­ing. He’d pre­vi­ously been in­vited to a game de­sign work­shop fol­low­ing a talk he de­liv­ered at the DigiPen In­sti­tute Of Tech­nol­ogy in Red­mond, Wash­ing­ton, where five of the at­ten­dees had ex­pressed in­ter­est in a sum­mer in­tern­ship. “I was like, ‘Sure, we’re close by – why don’t you guys come down and we’ll start noodling on the next project’s con­cept?” Kaz­dal says. “I’d been think­ing about it for quite a while, and I gave them tons of notes, sketches and stuff like that. I just laid out the whole project for them and they started bang­ing away on it, and we had a pro­to­type very quickly.”

A sec­ond chance en­counter saw Ed Fries, for­mer Mi­crosoft vice pres­i­dent of game pub­lish­ing, in­tro­duce Kaz­dal to Cyn­tient, an ar­ti­fi­cial in­tel­li­gence spe­cial­ist. A startup, it was look­ing for a part­ner project as it fin­ished work­ing on its mid­dle­ware while de­cid­ing on the di­rec­tion the com­pany would move in next.

“They were pretty fo­cused on more so­cial stuff, and strat­egy games,” Kaz­dal ex­plains. “I said, ‘Well, this is go­ing to be an ac­tion game, but I really love cut­ting-edge AI.’ My favourite game of all time is prob­a­bly the orig­i­nal Halo on Leg­endary, where the com­bat is just so fluid, or­ganic and dif­fer­ent. Even if you’re play­ing the same level over and over, those big­ger bat­tles are orig­i­nal ev­ery time. So we talked a lot about that stuff and I was like, ‘That’s what I want to do – I want to do Halo in 2D. I want that same level of con­trol, that same level of fidelity and great, or­ganic, evolv­ing com­bat.’”

With Cyn­tient hap­pily on­board, Kaz­dal worked with one of Skulls’ an­i­ma­tors, in­dus­try vet­eran and for­mer PC Gamer jour­nal­ist Colin Wil­liamson, to de­sign the look of the game. The aes­thetic they set­tled on was heav­ily in­spired by ’70s and ’80s anime, the fi­nal com­po­nent – along with deep physics and AI- based game­play – in what was rapidly shap­ing up to be Kaz­dal’s ideal ad­ven­ture game.

“I’d been do­ing sci-fi stuff for years as a con­cept artist, and I wanted to do some­thing that would stand out and have its own look,” Kaz­dal says. “I hate ren­dered ex­plo­sions and go­ing for real­ism in games. I re­mem­ber watch­ing the old Macrosses, and the really sexy ex­plo­sions in the old ani­mes. I started look­ing at some of that stuff, as a first ref­er­ence, think­ing about how we could stylise our ex­plo­sions. But then I re­alised that [style] has sort of been for­got­ten. I thought it would be fun to doo­dle on all th­ese things, sketch stuff and build this campy world per­fect for a space shoot­ing game.

“I didn’t really mean for it to be that era per se, but the more I ex­plored it, the more it made sense. Us­ing that era meant we could have all the stereo­typ­i­cal bad guys and all the ex­plo­sions and mechs, and it just sort of fell into place from there. There’s a whole gen­er­a­tion of play­ers who grew up watch­ing this stuff. The first time we showed it at PAX, peo­ple were walk­ing by and be­fore they’d even played the game were like, ‘Oh my god, this is amaz­ing, when can I buy it?’ just be­cause they were into that aes­thetic. I thought, ‘OK, this is cool, this is work­ing.’”

The PAX demo was a pro­to­type whipped up by Kaz­dal’s small team of in­terns. They worked on the game all through the sum­mer, two of them even­tu­ally tak­ing full-time roles. Kaz­dal also made a hand­ful of ad­di­tional hires as Galak-Z’s pro­to­type took shape. The build con­sisted of one gi­ant map pop­u­lated by ba­sic enemy ships, and Kaz­dal hand-placed ev­ery as­ter­oid and rock for­ma­tion, spend­ing time tweak­ing the twist­ing path through a labyrinthian cave net­work.

“We were a tiny team so we couldn’t do ev­ery­thing at once,” Kaz­dal says. “We were fo­cus­ing first on the feel­ing of the game – the over­all aes­thetic and what the com­bat and con­trols would feel like, and how the ba­sic flow of the game would work. That all came on­line really early, so we kept fine-tun­ing that and we kept tun­ing the level – you know, we would try dif­fer­ent enemy com­bi­na­tions, and be­cause they would be in dif­fer­ent spots and you might run into them in dif­fer­ent places, it really changed the feel­ing of the en­counter. It might be a really tight, con­fined space, it might be two squads that you’re fight­ing against in the same space, or it’s just one squad and it’s a wide-open area. Ev­ery playthrough felt com­pletely dif­fer­ent, even though it was the same level ev­ery time.”

Al­though the game was com­ing to­gether quickly, and show­ing ob­vi­ous prom­ise, 17-Bit was un­der ad­di­tional pres­sure be­cause it had to op­er­ate across an ocean. While Kaz­dal founded 17-Bit in Seat­tle, the draw of Ja­pan – where ear­lier in his ca­reer he had worked at Sega, as­sist­ing Tet­suya Mizuguchi on Rez and the Space Chan­nel 5 games – proved too great.

“I just really wanted to go back,” Kaz­dal tells us. “I thought, ‘What the hell? My wife is Ja­panese; we could just move there and have our com­pany, but we could go to lunch in Ky­oto ev­ery day. Wouldn’t that be awesome?’ So we did it. Ky­oto is an un­be­liev­able city to live in.”

Not that Kaz­dal is a fan of dis­trib­uted de­vel­op­ment. “I do pre­fer hav­ing ev­ery­one in

“THERE WAS A BUNCH OF STUFF COM­ING IN THAT WE’D NEVER SEEN BE­FORE AND WE WERE FREAK­ING OUT, TRY­ING TO PATCH IT”

the same room, in the same time zone. It makes stuff so much eas­ier, and be­ing the art di­rec­tor and cre­ative di­rec­tor it makes a lot of sense for me to be able to check in reg­u­larly as op­posed to once a day. It was dif­fi­cult, but we made it work.”

With Galak-Z’s ba­sics in place, and transcon­ti­nen­tal de­vel­op­ment headaches over­come, the team started to fo­cus on build­ing some­thing more sub­stan­tial. Kaz­dal had a high­level story in mind, and plenty of ideas for lev­els, but con­tin­u­ing to hand-de­sign and place be­spoke com­po­nents wasn’t sus­tain­able. Kaz­dal was also aware of the ex­tent to which the game was likely to evolve, mean­ing hand-de­signed lev­els cre­ated early could be re­jected – and the work that went into them wasted – down the line. This need to stream­line was a cat­a­lyst for the bur­geon­ing space shooter to take a new di­rec­tion.

“Go­ing in and mov­ing lit­er­ally ev­ery sin­gle rock in ev­ery sin­gle one of th­ese dun­geons was just in­san­ity; it was never go­ing to work,” Kaz­dal says. “So I talked to the lead en­gi­neer, Zach [Aik­man], about cre­at­ing a tool for me so I could hand-draw the rooms, then put them to­gether so we could re­use them and put them in dif­fer­ent spots. So we built this ed­i­tor and I cre­ated a hand­ful of rooms. By that time I was really con­scious of the fact the com­bat was the core of the game, and that ex­plo­ration and not know­ing where you were go­ing was a key com­po­nent of the whole ex­pe­ri­ence. So I was like, ‘Can we just try shuf­fling th­ese things up?’ “I was play­ing a lot of Spelunky and Rogue

Legacy at the time and I was really get­ting into that for­mat, just jump­ing in and hav­ing fun and not hav­ing to mem­o­rise stages. And I’ll never forget the night that, hav­ing played through our dun­geon a thou­sand times, all of a sud­den we had this shuf­fling thing that would give you a new ex­pe­ri­ence ev­ery time. I mean, I was

flat­tened by how awesome it was.” Kaz­dal hes­i­tantly told the rest of the team he thought the best struc­ture for their form­less pro­to­type was that of a Rogue­like. Ex­pect­ing re­sis­tance, he was sur­prised to find they all felt the same. Now, as a game about pro­ce­du­rally gen­er­ated en­coun­ters and ex­plo­ration, Galak-Z had real di­rec­tion. But while Kaz­dal had found plenty to ap­pre­ci­ate in the work of Derek Yu and Cel­lar Door Games, he still had mis­giv­ings about the use of ran­domised lev­els over hand-tuned creations. Con­cerned, and wrestling with tech­nol­ogy and al­go­rithms that were new to them, the team de­voured as many ar­ti­cles and videos on the making of Rogue Legacy, Spelunky and other pro­ce­du­rally gen­er­ated con­tem­po­raries as they could. It quickly be­came ap­par­ent that the num­bers could be con­trolled, re­fined and di­rected with pur­pose; chance didn’t have to fac­tor into it.

“We had that one hand-de­signed dun­geon that we played over and over, and had tuned the hell out of to be sort of a per­fect stereo­typ­i­cal Galak-Z level. So we thought, ‘OK, let’s re­verseengi­neer this thing.’ It took a while, but we were spend­ing a lot of time on how to recre­ate that al­go­rith­mi­cally so it would keep a lot of th­ese key pac­ing de­ci­sions we’d come across nat­u­rally.

“Some­times cer­tain en­e­mies are way too hard, so you go in and tune that al­go­rithm. The same for when the lev­els are too short or too long. And so we started to go down the line and nip all the prob­lems in the bud, keep­ing an ar­ti­fi­cially higher level of con­trol be­hind the scenes that play­ers don’t nec­es­sar­ily need to know about. So now I’m the big­gest fan in the world of this stuff, be­cause even as the de­sign­ers of the game, we get to play some­thing new ev­ery time.” That’s all very well for hu­man ex­plor­ers, but it pre­sented a unique chal­lenge for Cyn­tient’s AI. Af­ter a brain­storm­ing ses­sion in which 17-Bit talked with Cyn­tient about what it would mean to have enemy craft that could think for them­selves, ad­here to or­ders, re­treat when threat­ened and even make pre­dic­tions on what the player might do next, a ba­sic sys­tem was plugged in. It proved im­me­di­ately amus­ing and made for com­bat­ants that were ca­pa­ble of sur­pris­ing – ex­actly what Kaz­dal had hoped for when he de­scribed Halo’s en­coun­ters. It proved so ef­fec­tive, in fact, that thehe team quickly re­alised they would have to dial down the fe­roc­ity of the op­po­si­tion’s in­tel­li­gence. e. Enemy pi­lots would dodge in­com­ing shots and place their own with un­nerv­ing tac­ti­cal nous.

“We were like, ‘All right, we’ve got to take th­ese guys se­ri­ously,’” Kaz­dal ex­plains. “They’re e not nec­es­sar­ily faster than you, they’re just re­al­lyy good at their jobs. They’re th­ese lit­tle ter­mi­na­tors rs try­ing to kill you, and they’re creep­ily in­tel­li­gent.. I don’t want that AI go­ing into drones or any­thing ng any­time soon be­cause it’s ter­ri­fy­ing.”

Get­ting Galak-Z into or­bit may have been a fraught ex­pe­ri­ence, but Kaz­dal is thor­oughly sat­is­fied with the re­sults. 17-Bit cre­ated a con­glom­er­ate of clas­sic gen­res that in­tro­duced new ideas to each com­po­nent part, while at the same time offering up some­thing that feels re­fresh­ingly un­fa­mil­iar to mod­ern play­ers.

“Some of the sys­tems and tech that are go­ingng into th­ese clas­sic gen­res is so ex­cit­ing to me, andnd that’s what I want to keep fo­cus­ing on,” Kaz­dal l says. “The next couple of games I’m think­ing about are the same sort of idea – com­bin­ing clas­sic gen­res with mod­ern tech and see­ing how w that shakes up the whole ex­pe­ri­ence. And I thinknk it really does. I mean, Galak-Z is a shoot­ing game me that feels un­like any other ar­cade cab­i­net, just be­cause the AI is making such a dif­fer­ence. In that sense, each en­counter is much more of a com­bat sim­u­la­tion than it is a videogame.”

Aside from for­mi­da­ble pi­lots and plenty of mechs, there’s also in­dige­nous life to con­tend with in­side as­ter­oids

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.