Jul 30, 2019

Albion Online - iancurtismayfield

Michael Schwahn, AKA Retroman, has been a part of the Albion Online team since 2015. Prior to joining Sandbox, he worked on Spec Ops: The Line at YAGER (also based in Berlin) and created an indie shoot-em-up called Teslapunk for mobile, Xbox One & Wii-U. As Combat Designer for Albion Online, Michael is responsible for keeping a huge range of weapon and armor skills up-to-date, and ensuring that the combat meta remains fair and vital with each game update.

With the next Guild Season starting this Saturday, we took the opportunity to talk to him about his approach to game design, what makes combat in Albion unique, and the games that have influenced him.

You've worked at Sandbox since the very early days of Albion Online. How has your work on the game evolved over the years?

The first big task I had was to overhaul the combat balance for the first beta. That was a very exciting, but frankly also a bit intimidating time, since I had never worked on an MMO before and Albion Online’s game design is quite a complex beast.

So, how to best approach the balancing of a sandbox MMO? Well, the game has PvP and PvE. Each of these aspects requires quite a different balance approach. But since Albion's endgame is heavily focused on guild warfare and PvP, I really wanted to focus on the PvP balance first and foremost. Of course PvE can’t be ignored, but all items are first designed around PvP and then after their PvP role is settled we can look at their performance in PvE.

I also want to see how far we can push the skill cap, to make the game less about numbers and stats, and instead increase the importance of group coordination and individual performance. With these requirements, I was more interested in MOBAs as a reference than other MMOs. There are a lot of similarities: MOBAs share the same perspective, also have a few spells active at a time, have a very high skill cap, and are often 5v5 like our GvGs.

As my forum name suggests, I am a gamer of a different era, and one of my absolute favorite game genres is fighting games. I played them semi-competitively and even held small tournaments back in school. With Albion Online, I wanted to bring this sense of excitement and personal mastery that I felt with these games to a new generation of gamers. That is the reason we have no random elements in PvP, because I always appreciated the super-deterministic mechanics of fighting games with no RNG. And on an abstract level, some of the concepts found in fighting games also translate surprisingly well to MOBAs: things like trying to bait enemies, the importance of positioning, and counterplay options.

Okay, so now I could decide on the first set of abilities and roles for the items. The next thing was to get the actual numbers into the game. To adjust so many numbers, you create a mathematical model with formulas with certain assumptions about the gameplay in mind. Things like how damage and stun translate. So every spell gets a certain power which comes from cooldown, cast time, AoE radius, etc… and also a ‘situational’ factor on top (for instance, skillshots have increased power because they can miss). This spell power is then split into the spell’s effects. For example, you could split the spell power to have 70% damage and 30% slow. This is a very useful approach to create a baseline for all abilities. However, it is just a tool to help you guide the values.

In the early days I sometimes followed these values even when they felt a bit off, which is not the intention. Because there are spells that are really hard to put a number on, like Blink. The original balance approach for Blink was, "you save so much time moving X meters, how much more DPS can you get out from the time you save". Which is one way to approach it, but it really doesn’t translate well into the actual value. It doesn’t consider, for example, that you can dodge abilities, make up for a positioning error, or secure a kill. The truth is this approach doesn’t cover all utility or synergy effects, which are so crucial. So nowadays I still use the math model to get a basic idea of the spell’s worth, damage, heal, CC, or energy values. But is is really just a first baseline. Playtests and meta observations play a very important role in changing the actual values.

Balancing combat and introducing new skills is a multistep process that involves multiple playtests. Can you give an overview of this process?

So first I start with a rough idea and a concept of an ability. Then I produce a prototype with our internal spell scripting system, using existing effects and animations. The first iteration of the ability will also use numbers from the above-mentioned math model. Then I try to get as much early feedback from players as possible. For this process we have the NDA playtests, to get the feedback of experienced players, because I often trust their feedback on PvP even more than our own feedback. Simply put, this is because many of the high-level players are better players than the devs.

I have to stay aware that all players are human beings, and so their feedback might be biased towards a specific weapon or activity in the game. But if many players from competing guilds complain about the same issue, it is very likely that they are on to something. We also record the gameplay sessions and I sometimes also join the playtests to directly talk to the players for feedback. But recently, FuS has been doing a great job covering the playtest responsibilities, which leaves me with more time to work on the game. After each playtest I keep making adjustments and performing more playtests until we reach a patch deadline. And after one patch ends another begins, so I'm often already working on the next patch as soon as the current one goes live.

Is there anything you wish you had done differently during development?

My biggest mistake was the Grudge meta. The main problem wasn’t that the spell was so crazily good that it dictated the meta (which it did), but rather that I didn’t try to push a balance patch out earlier. I remember this was after release, and I was kinda hesitant to make meta adjustments on the fly. I originally intended to have fix update cycles, so all players could rely on their team compositions for a season. But I think this approach was a bit outdated and came from my old-timer gaming experience from an era without patches. Now when I have the impression the meta is getting too unhealthy, I try to fix it as soon as possible.

What do you think are the biggest challenges when working on Albion Online?

Like I mentioned earlier, this game is a design beast. It has a lot of complex systems which are interwoven, which really makes it unlike anything else I've worked on. The whole player-driven economy and the the sandbox nature of the game impacts all design decisions. This goes down even on the combat level: for instance, fights can be 1v1, 5v5 or 100v100, they can be in different situations in the game: gathering, ganking, in Hellgates or dungeons or castles.

So while I mentioned earlier I tried to get as much influence from MOBAs as possible, there are of course a few crucial differences in Albion Online. First off, this is an open-world game, so mobility is extremely powerful. The one who has more mobility in the open world dictates the engage or disengage. Which is why I needed to tone down a lot of the early mobility options like Haste or the kiting potential of Warbow.

Another difference is that the introduction of traditional healer roles for MMO PvE really changes the PvP dynamic. If the healer has a lot of impact in a 5v5 fight, he becomes potentially very strong in 1v1 fights. Another difference is, because of the full-loot nature, in general the combat should not as bursty as a MOBA, so one-shot builds are not encouraged, which made finding the right balance for assassin builds always a bit tricky. Or how in ZvZ fights the big Wombo Combos are actually also a bit against the general idea of preventing one shots. On the other hand, these combos enable a smaller group to win against a big group, which is pretty cool. And I also consider players engaging in big ZvZ fights aware of that risk.

At first I thought "balanced" meant every item should be used the same amount. While it is true that every item needs a purpose, and ideally something it excels in, I don’t think the balance question is so easily answered. Because what does "balanced" in a game even mean? If everything is equal or very similar you get a very balanced game, but it won’t be very exciting to play.

I think a good balance for a game means it offers a variety in playstyles and interesting decisions, where weapons excel at one thing but feel equally strong. You can also bring up the mathematical model for game balance here again. I think if your game can be easily broken down in numbers, then players can enter your game into Excel and find the strongest composition. But if it is more about situational plays, synergies and counters, it is much harder to "solve" for players. And it will actually keep them interested longer.

Of course, trying to give everything a unique role makes it quite difficult to keep all the options viable in high-level play, but I think in the long run this approach creates a way more interesting and dynamic game. Ideally all abilities feel OP when used with the right setup or synergy. This is at least the direction I want to push the game further into. Abilities with potential high impact and outplay potential, but with counterplay options.

What are some of your favorite parts of working on Albion Online?

Definitely the community feedback. It can sometimes be harsh, but seeing our players so engaged with the game, on streams, videos and discussion on the forums and reddit is really an immense payoff for the hard work. It reminds me of when I was on the player side and was totally engaged with a game. The other thing I really enjoy working at SBI is the sense of personal responsibility and freedom I have. Being a small team means we are not very corporate and an individual can have a lot of impact on the game. And although I have been working here for more than four years, I still learn new things all the time. Recently I got involved with new topics like the Randomized Dungeons and the Season 7 adjustments. It can be a lot of work and stress, but seeing the result in the end is really fulfilling.

What are some of your all-time favorite games? Which games have you learned the most from?

I have no single favorite game. But I can list some of my favorite franchises:

Fighting games: Street Fighter, Tekken, Garou (and currently Samurai Shodown). Shoot’Em Up games, which probably no one knows: Dodonpachi, Giga Wing, Gradius V, Einhänder.

Apart from that I love all the From Software games since Dark Souls (Bloodborne & DS3 are my favorites), lots of retro games like Super Metroid, Final Fantasy VI, Castlevania games, Contra games, Zelda games, Advanced Wars and Tetris Attack. And a lot of other games that I'm sure I forgot to mention.

Anything else to add?

Maybe a bit out of the blue, but if some of you readers are considering becoming Game Designers yourself, you can do it. Just be aware it is a lot of work, it is not the most secure job and it won’t be easy. But if you really want to do it, it is possible. I remember as a kid I wanted to work in the game industry and the dream seemed almost impossible, all the grownups around me tried to talk me out of it. So I didn’t talk about it anymore, but inside I just couldn’t give up the dream. And in the end I found my way in the industry, even if it wasn’t the most direct way, but passion and dedication made it possible. And even if it is not the most rational decision, if your passion really lies with making games then give it a try. Maybe build a small game in Unity, Game Maker or even a board game. If you don’t end up in the industry, I think trying and failing feels much better than not trying and wondering “what if” your whole life.

Stay tuned for more Dev Spotlights in the coming weeks.