Hello feeble meatsacks and welcome to another of my musings. This week, one of the Codemasters Design Minions has agreed to scribble some words and pictures about how they made my Netherghūl appear in one of those videogame things. It’s really quite clever, you know!
He was a little reluctant at first but became quite communicative once we’d removed the poker from his unmentionables.
Love, Gnarl x
Designing the Netherghūl
As Overlord: Fellowship of Evil was conceived as a multiplayer title, our earliest design meetings concerned the creation of the player characters. We very quickly realised that we couldn’t have multiple “Overlords” (as that doesn’t make sense, given the Overlord universe’s lore). Our solution, as is often the case, lay with the minions – we realised that the player using the minions is one of the defining characteristics of Overlord, and it would only make sense that the players assume characters that the minions would follow, even root for.
The Overlord games tend to begin with the minions “seeking out” their next master. As we’re fans of those games, we wanted to do this too – but naturally, for a multiplayer game, we needed a situation where players start at the same time, in the same location. This led us onto the idea of the “Trials of Evil”, which is the overarching story of which the Netherghūl are part. The minions have “summoned” these characters back from the dead, as candidates for the next Overlord, so each of them has some element that the minions particularly like.
From this, we thought about how each character would have an “aspect” of what it means to be the Overlord, and this was probably the first thing to be finalised. We wanted one character to represent “fury” and “rage”, another character to represent “arrogance” and “callousness”, a third character to represent “avarice” and “treachery” and a final character to represent “impatience” and “death”. That being said, as we’re making a video game, we also wanted the four characters to be varied in terms of appearance and move-set, so they’re more interesting to use and easier to tell apart.
At this point, we briefly considered the idea that each character would be a specific minion colour, and only be able to use that minion type – for instance, the red character could only use red minions. We abandoned this quickly though, because we knew that one of the core aspects of Overlord’s gameplay is picking and choosing the correct minions for the right task, and removing this would have been a step too far. Even at this early stage, though we new Overlord: FOE would be different, we wanted to preserve those fundamental elements of the original games as much as possible within the game’s multiplayer framework.
It was at this time that the art team started to work on the characters, producing concepts based on these initial ideas. The Netherghūl characters went through many revisions in their early design, sticking closely to the classical fantasy archetypes.
The original names of the foursome were Infera the Red, Terranix the Brown, Cryos the Blue and Aria the Green. This dates back to the earliest art concepts, where we gave them names for convenience. Internally, we previously called them CHA01, CHA02 (for character-01 etc.) but we found having a name, even an early one, made them much easier to work with.
This could, almost, be thought as v1.0 of the character lineup. That being said, nothing existed of them in 3D – our early game prototype (which already had the ability to wield minions, swing a weapon etc.) just used generic Overlord-style characters we’d put together quickly.
Rhianna comes onto the project
Rhianna had worked with us remotely since the earliest parts of the project (the concept of the “Trials of Evil”, earlier, was one of her ideas), but it was around this point where she started coming into the office day-to-day. As designers, our approach to the characters was very mechanic-driven – for instance, we consider range, size, shape, abilities and such, whereas Rhianna was interested in trying to forge the character’s identities, and deriving ideas from those.
We also started to build the systems for minion control and combat that would underpin much of the game. Our ideas were initially much more procedural – i.e. we would have randomly generated challenges, and aspects of this would apply to the characters too. However, Rhianna’s involvement made clear that the story is one of Overlord’s more compelling features, and we decided instead to go down the route of very clear-cut, written characters and scenarios, to allow for tighter characterisation and story-telling.
This led to a short period of very fast iteration, where we created many different character ideas, backstories and concept art to go with them. I won’t try to list specifics, but there were at least a dozen iterations on every character, eventually settling on the four characters you see in the final game. These are quite different to the above lineup, but they kept a few of our original plans – for instance, the characters having a 50/50 gender split, and each representing a different “aspect” of evil.
This is a piece of the Inferna concept art that was made toward the end of this process:
For a long time during development, Inferna’s special move was a hammer, in a nod to the original Terranix design. We switched it to a scythe later in development because scythes are cooler than hammers
During the above process, we continued to try and keep the gameplay design in-step with the characters.
One of the more difficult aspects of making an ARPG with various enemies, characters and abilities is balancing. It helps, as a designer, to build systems that are inherently balanced from the off – that way, you don’t have to completely start from scratch once the game is playable. For the characters, our initial approach was very simple:
These are the “play silhouettes” that each of the Netherghūl had at this point. The moves of each character exert a directional force on an area, making a rough shape. Cryos changed a fair bit after this image, but that’s the point – these “paper dolls” allow you to try out different combinations of enemies and characters (and minions!), and if something doesn’t work, you just crumple it up and make a new one.
For instance, this is a ranged enemy overlaid with Inferna. It allows you to see, easily, that Inferna may struggle against ranged foes – but someone like Malady, with long-range attacks, will have no problem.
There’s also no way you can lay down the character shapes where, if the players stood together, they could be totally “safe” – i.e. they can’t cover every location – and even if you had four Inferna players, they would be vulnerable to AoE attacks.
The other useful thing about these patterns is that you can use them to derive damage. In the final game, character damage is calculated based on the area that the attacks cover – a wide, sweeping attack will inherently deal less damage than a small focused attack. When you add duration/speed to the mix, you get pretty good values for balancing purposes.
We came away from this knowing that we essentially had four character gameplay types:
- Inferna is a melee character with more focused attacks
- Hakon is a melee character with wider, more general attacks
- Cryos is a ranged character with wider, general attacks
- Malady is a ranged character with more focused attacks
In Overlord FoE, just like the older games, we knew players would have control of the minions, and we wanted them to tie into this system. Here’s a very early shot of how they were put together (this went through many iterations after this photo was taken):
We knew, then, that you could combine them with the player abilities to overcome that character’s inherent weaknesses. For instance, in the above example, Inferna lacks the range to fight distant enemies, but…
… with the help of a Red, she has the necessary reach.
This last part was a very important part of the game for us, and one of the reasons we felt that Overlord FoE would be as fun for solo players as it is for groups. Traditionally, ARPG characters have strengths and weaknesses, and players use each other to supplement this. In our game, this is still true, but players also have the minions for this purpose. This means that even when playing solo, you’re never “alone”.
Inferna – the first Netherghūl
Once we had a good idea of how the characters should look and play, we decided to start with Inferna, mainly because she is the closest, visually, to Overlord characters from the previous games (tall, armoured, imposing).
First, the art team produced Inferna’s initial rigged model from reference concept art, similar to the image below:
We iterated on this several times. The first version was built to very similar proportions to the main character from Overlord, which worked for that game, but in an Overhead view, a great deal of the gravitas was lost. We realised that we were going to have to exaggerate these features, so Inferna gained bigger shoulders, bigger boots, a more prominent helm and a larger sword. This is pretty common for top-down action games.
We produced the animations via a mixture of motion capture (with various wooden swords we have in the office) and hand-animated movements. Even the motion-capture had to undergo a great deal of work, mainly because none of us are eight feet tall! The “actors” will remain nameless, but suffice to say, a fair few members of the team spent some time prancing around in the motion capture suit. Inferna’s animations went through many revisions for timing and overall “feel”.
Once we had Inferna in the game, and were able to run around, attack etc. the team started working on the various features that would support the gameplay. It was around this time we had the first environment built (the sewers) and the first enemy, so we could sit down and actually play through something which felt representative of the game, crude though it was.
We used Inferna as a testbed for many features and ideas, some of which went into the final game and some of which were cut.
For instance, at one point, we looked into letting the player “sacrifice” minions, drinking their essence to gain more health – we even considered the idea that the Netherghūl were always on the verge of death, and had to constantly consume life force to survive. Whilst interesting on paper, ultimately this felt poor in-game (the signs and feedbacks for it were tricky to communicate) and had to be cut. The concept was re-used for the “minfluence” – the life-force ammunition that the players pick up in order to be able to summon minions, similar to the “life force” mechanic from the old Overlord games.
We kept iterating on Inferna for as long as we possibly could, but naturally in we had time constraints to consider. Inferna was probably about 75% complete when we started on Cryos. Malady came next, and Hakon was last. As is usual in videogame development, Inferna, being the first, took the longest to create; Hakon was comparatively very quick, but by that time, all of the questions had been resolved and challenges dealt with. Animation, in particular, required far fewer revisions for the later characters than the earlier ones.
Of course, having the four characters in-game was far from the end of the process. We balanced and rebalanced, in some cases changing their moves.
Cryos, in particular, originally had a two-handed projectile attack that was a bit like the John Woo “guns akimbo”, only with magic. Unfortunately, whilst it looked cool when it worked, this created a variety of problems for gameplay balance and had to be replaced with his final light attack, which summons a storm-wind and showers enemies with shards of ice. You can actually see this old attack for a fraction of a second in Overlord FoE’s earliest trailer.
We had to be careful, however, because whilst the damage of a move can be edited in a flash, other elements, such as sound and animation, take much longer. Pretty soon, you realise that three-months-to-launch gives time for only a set number of revisions, and there are many other parts of the game that need just as much attention.
Much of the work, at this point, came from applying multiple characters to different situations. When we have four different players, four different types of minions (with several of each running around) and various enemy types, there are many variables. We also have a range of weapons for each character that alters how they play. This can be chaotic, but fortunately, that chaos is part of what makes the game fun.
The “paper-doll” method described above worked very well, and is still very evident in the final game. The only problem lay with Malady’s attacks, which by the maths alone were so powerful as to be absurd (we had to modify the formulae to compensate).
Like with any creative project, what you’ll hopefully take away from this is that we considered many, many different (and often conflicting) ideas before we settled on the final Netherghūl. Each one was crafted by a large number of people over an extended period of time, and there are many discarded ideas could see future use in a follow-up.
Of course, one question is what the future holds for the Netherghūl characters themselves. This is very much Rhianna’s secret. That being said, although each of the Netherghūl hates the others (they work together purely out of convenience), none of them are powerful enough to take up the mantle of Overlord (yet).
Perhaps one of them will succeed? Or maybe their activities will rouse the attention of something older..? Gnarl, it seems, is keeping his cards close to his chest.