Behind Every Great Game, There's Game Tech
Total Page:16
File Type:pdf, Size:1020Kb
Behind every great game, there’s game tech 2 GAME TECH Issue 02 ILLUSTRATIONS NOT ACTUAL GAME FOOTAGE FLAREGAMES MILESTONE FRONTIER Thorsten Grösch started his career as a high-school teacher, but is now Head of DevOps and IT at Flaregames. Gonçalo Antunes is Product Lead at Flaregames subsidiary Kopla Games, while Mikko Airaksinen is the Main Backend Developer and Programmer. Bringing the essence of action role-playing games to made-for-mobile, Nonstop Knight 2 sees players hacking and slashing their way through a series of dungeons, on a quest to win gold, armor, and weapons, as they take down hordes of formidable enemies. Players mix and match loot with skill sets and talents such as dancing flame and frostbite, join a guild, choose GAME TECH an AI-driven companion, and battle their way to the top of the leaderboards in weekly tournaments. Issue 02 3 FLAREGAMES MILESTONE FRONTIER Now part of mobile games publisher core mission. At that point, it was a case Flaregames, Kopla Games is the studio of: ‘Let’s go back to making approachable behind Nonstop Knight 2. It builds on the ARPGs, and see what we’ve learned.’ popularity of the first game in the series, which achieved more than two million Gonçalo: “It’s all about the loot” installs within three days of release in 2015, and went on to notch-up 15 million Nonstop Knight (NSK1) was very player downloads. approachable and had a really simple, It took one-and-a-half years and more incremental meta attached to it. NSK2 than 10 prototypes to develop Nonstop was the next step, so we wanted Knight 2. something closer to classic action RPG. That meant it had to be all about the loot you collect on your way through Gonçalo: “We wandered the game, and players would need to be too far off track” able to keep that loot, to make it more meaningful. We also wanted to make We spent a long time experimenting and GAME TECH other progression factors, in terms of Flaregames is a German-based mobile killed a lot of good, varied prototypes skills and talents, count, so we could before we came up with Nonstop Knight games publishing company, which works give players a much wider choice when 2. That was partly because we decided with subsidiaries Kopla Games and Keen it came to customizing the knights. We to put role-playing games aside and knew we could only do that if there were Flare, along with other games studios Issue 02 experiment with other genres for a bit. sustainable progression systems, rather around the world, to help developers While we were crossing the swamp, it than something that was just wiped away bring their creations to market. made us realize we’d wandered too far regularly, as in NSK1. off-track and away from the company’s 4 FLAREGAMES MILESTONE FRONTIER Mikko: “It’s gut-wrenching when that deconstruction, while making sure Thorsten: “Moving to the cloud you have to throw away your gear” the game didn’t lose its appeal. means competition can be more full-on” NSK1’s incremental metagame, where Mikko: “Robust base-line you lost your gear after resets, was a architecture was a boost” Enemy and boss hunts, where players go solution to a problem. Close to its release after the bad guys, are still among the date, there was a struggle around how to most popular features in NSK2, because Adding the three new gear slots for get the metagame live. The incremental there’s that strong competition element. NSK2 didn’t provide a massive technical metagame worked, so the team went We had a problem around dungeon boss challenge because we reused a lot of the with that. But many of us are long-term hunts in NSK1, because players can easily base-line architecture from NSK1, which ARPG fans, and hated the gut-wrenching manipulate data in an offline game. That has a very robust system. In NSK1 and feeling that comes from having to throw led to a lot of design restrictions and NSK2, there are many ways perks can be away all your gear. not being able to make it as competitive activated, so adding a couple more item as we’d have liked. To compensate, we sources wasn’t a huge problem. added layers of co-operation, but that Gonçalo: “The shackles came off” meant you’d be competing with people in Mikko: “We knew we needed the tournament, while also co-operating Once we decided to leave the incremental to move towards a server with them for a common goal. behind, the shackles were off. We went authoritative game” from NSK1, where you had three sets of In NSK2, there’s clear separation. Co- items, weapons, and armor that were operation happens only in the guilds, One thing we had to re-factor and re- mostly disposable, to six in NSK2. The where players can gain access to exclusive think was around the fact that NSK1 was player also has the option of unlocking items they can’t find anywhere else in the a client-authoritative game. We knew special perks, by using more of one set, game, and competition takes place in the GAME TECH we needed to move towards a server- going half-and-half or using the joker hat tournaments. That makes it thematically authoritative game for NSK2, mainly which bends the rules a bit. accurate and allows it to be full-on— because of how easy it is to cheat when you’re there to beat everyone else and The challenge was to create a game in it’s client authoritative. Because of there’s no layer of co-operation involved. classic ARPG format, where the player technical limitations, it’s not fully there Issue 02 That’s something we can do because has loads of choices and needs to think yet. Some parts are still client driven, but the result, the logic, and the rules of the hard about strategy, while staying true we’ve managed to isolate most of the tournament are determined by the server to our mission of presenting the genre in sensitive ones and have developed tools and not the client itself. an understandable and easier-to-manage to detect cheaters. way. The toughest part was achieving 5 FLAREGAMES Gonçalo: “We spent months Mikko: “The companion knight Mikko: “This feature is cursed” Gonçalo: “The grind is real” honing it” kept doing weird things” Players can choose other players as We made the game much harder this One of the most complicated aspects of With NSK2, players can build different companions and we wanted to make sure time around. We wanted the early part the game that looks simple but caused skill sets to customize the gameplay. that wasn’t placing unnecessary stress on of the experience to be as approachable a lot of pain, is the companion knight That’s really popular with players, but it the system. We also didn’t want to keep and rewarding as possible, but at some feature. As a player, you can recruit an AI- took a lot of time and consideration to the player waiting too long, so tried to point, ARPG canon says progression slows powered ally to take into the dungeon, get right. We spent months improving pre-warm and pre-cache our data, as in down and is balanced-out with skill and MILESTONE so you don’t have to face the monsters that feature, from prototyping right up preloading companion data, to make sure determination. One of our maxims was: alone. But although the game architecture until soft launch. We had the base skills the UX flow was smooth. Our aim was to “The grind is real.” We wanted depth in and engine support having multiple from NSK1, so asked ourselves: “Can we have the data ready in the client, so when the core gameplay itself, so players have characters on screen at the same time, it themetize the skill sets? Can we make a player clicks through to inspect, they to invest more time weighing-up decisions took a long time to make the companion them all in one element or one class? Will don’t have to wait. We had a lot of issues and strategy around the best combos, behave helpfully, instead of doing weird they be as good, and what’s our design with that, because it was really complex timing, and how to face different enemies. things. We had to iron-out a lot of details FRONTIER ceiling in terms of how many we can do?” to make the caching work properly. It’s complexity without complication and, in the AI to make it act more like a human of course, we made the system harder It took months to build the Fire skills set. player. There’s no point having this other We also faced multiple situations towards the endgame, because players We worked on the Frost skills next, and ‘person’ around, if they’re not doing when the server randomized a certain who stick with it, want the game to get finally the Void, which was a bit easier anything useful. combination of companion knights. If tougher, so they can get better. because we’d learned lessons from the somebody happened to choose three first two. specific knights, weird and unexpected stuff would happen. This feature has been Mikko: “We were pulling our the one that’s ‘surprised’ us time and time hair out” again—it almost feels like it’s cursed.