Roles of Monsters: Fodder

Not every monster is intended as a threat. Some monsters are there to ease the player into the game at the start, keep him a little busy in between dangerous encounters, and back up more dangerous monsters. These non-threatening monsters are called fodder.


The best example of fodder is the Imp from the original Doom. Somewhat like a swarmer, they don’t have a lot of health, though still more than a swarmer. Usually enough to take one or two shotgun blasts. They throw slow-moving fireballs that are easily dodged, and don’t do a lot of damage anyway. While they’re often encountered in groups, they have no mechanism for attacking all at once, so it’s more like you fight a few at a time. The only threat in that case is if you have very limited space to work with.

So why do Imps exist, if they aren’t much of a threat? Basically, they’re used whenever the designers want to give the player something to do without challenging him much. Usually, as a break between encounters that try to kill the player, keeping him busy but still giving him a short rest. They’re also used very often to back up stronger monsters. For example, we often see a group of Imps with a single Baron of Hell, in Doom 1. The Baron of Hell is the main part of the encounter, but the Imps give us something else to worry about, to take our attention off of the main threat.


Fodder can also use some variation. The Guards from Quake 2, for example, are basically the same kind of fodder monster, but each has a little more health and different weapons. Quake 2 itself comes with blaster, shotgun and machinegun versions, with laser and ripper variants added in an expansion pack. Without much effort, they’ve added more monsters that are still familiar for a player; especially when these weapons mirror those the player has, himself. By necessity, fodder monsters have to be somewhat common throughout a game. When there’s enough variation, like with Quake 2’s Guards, it keeps the fodder from getting boring too quickly; especially when new versions are introduced more slowly over the course of the game. It also adds more flexibility for the designer to sculpt an encounter.

One more minor point about this, keep in mind that the player will be seeing a lot of fodder, so it’s a good idea to have even more variation. The Strogg Marine in Quake 4, for example, has more sounds than any other monster in the game, so it doesn’t play the same sounds all the time (which can get annoying).


Another fodder monster is the Grunt from Halo. Like the Imp, they have some weak projectile attacks, but they have some variation as well. In Halo 1, they could be equipped with either Plasma Pistols or Needlers, and there are Minor (basic, yellow-armored), Major (red-armored, more health and can throw plasma grenades), and towards the end of the game there are Special-Ops Grunts (black-armored, even more health and almost never panic). Often, they’re seen backing up Elites, the major threat of these encounters.

The big difference between this and the Imp, however, is that these guys drop their weapons, giving the player a source of ammo. Similar to Combine Soldiers in Half-Life 2, or the basic Former Humans in Doom, they act as a method for the player to resupply, but the player has to do something to earn it. This gives them another function alongside what fodder already does.


There is one major variation of fodder, and that is the basic meat-shield. The Demon from Doom is the best example of one, though the Headcrab Zombie from Half-Life 2 is one, as well. It has a lot of health, but only has a melee attack that doesn’t do a lot of damage, and can only hit with if it pins you against a wall. Unless very strategically placed (There are some traps in Doom 2 that take advantage of them, most notably in “Tricks and Traps” and “Downtown”), they aren’t particularly dangerous. All they really do is rush at you and absorb damage.

They have most of the characteristics of fodder, but they bring something else to the table in that they can drain resources. Somewhat like tanks, they drain either ammo, time, or especially space in an area, but unlike true tanks, they don’t really demand your attention like they should (either by being threatening or having some annoying mechanic). There are some meat-shields that have ranged attacks, but usually those are low-damage or easy to dodge in any case (though some consider Doom’s Baron of Hell this, despite its high damage). They do add a bit on the mechanical side, but they can be pretty boring for the player.

Like the Imp, these fodder monsters are there to keep the player busy while giving him a break. In the case of enemies like Halo’s Grunt or Doom’s zombies, they can also give the player ammo and weapons without simply having them lying around (he has to earn it). Sometimes, as well, they’re a minor challenge after giving an important item, but without actively trying to kill the player (for example, the Imps after you get the shotgun in Doom 1’s Hell Keep level). They also give the player the ability to ease into gameplay slowly at the start, letting him get used to controls and other mechanics. Many games these days (Crysis, for example) omit the fodder role entirely, and I think that’s a mistake. Games that don’t have fodder at all either don’t give the player a break, or have downtime where absolutely nothing happens. Fodder is best used to help the designer control the flow of gameplay like a sonata, with high-action and low-action segments, as opposed to high-action and no-action.

There are cases where fodder doesn’t really help. For example, I don’t think fodder works well in survival-horror games; weak monsters tend to give the player a feeling of complacency, making him feel like he can handle whatever the game throws at him. Dead Space’s idea of fodder, the Slasher, is capable of killing the player, it’s just easy to deal with unless it catches him off guard. Fodder works best when your game uses multiple, different roles of monsters.

Every once in a while, some games give certain kinds of fodder some incredibly-powerful ability, like Grunts Majors and the basic Chaos Heretics from Space Marine both have the ability to throw grenades. This is mostly some attempt at keeping the player on edge, because this group of easy monsters can actually kill him if he’s careless. However, it really depends on the pacing and mechanics of the game in question; for example, this works well in Space Marine and Halo because it’s easy enough to dodge, but it could never work in a game like Doom. Doom is just too fast and hitboxes are too easy to hit for this to work. It also works because not every fodder monster in Halo and Space Marine has an ability like this; it’s just the red-armored Grunts in Halo, for example. It would be annoying and overpowered if every Grunt in Halo could throw plasma grenades, and it gives more meaning to that variety besides slightly-increased health.

When fodder monsters do have nothing but powerful attacks, they’re what are commonly known as glass cannons. The best example of this that I can think of are the Special-Ops Grunts with Fuel Rod Guns in Halo 1. Again, among the weakest monsters in the game, but suddenly has a weapon that can instantly kill you if you don’t see it. These surprisingly don’t work very well, but usually a game can get away with one if it’s well-used. The problem with glass cannons is that they’re pretty luck-based. If the player spots them, he can kill them quickly and they aren’t a problem; if he doesn’t, then he takes a lot of damage. The upside to this is if the glass cannon is a variant of an existing fodder monster, like the Special-Ops Grunt; we know it’s going to be easy to kill from experience, but suddenly it has a much bigger gun, so we immediately know exactly how we should handle this. Also, for balance reasons, its gun explodes when it dies, so the player can’t use it (which nullifies the ammo-giving mechanic of Grunts). I’ve heard arguments that the glass cannon could be considered its own role, but it doesn’t have enough characteristics on its own to warrant that, so I figured I would talk about it here. I will be expanding on this in future monster role articles, as it applies.

There are enough similarities between swarmers and fodder that you could say that swarmers are fodder that simply has some way of taking advantage of their numbers. Certainly Halo’s Grunt and Drone aren’t very different aside from the Drone’s mobility. Because of that, you can use swarmers as fodder by just reducing their numbers enough to where they don’t really swarm. I still like seeing an individual kind of monster as dedicated fodder, but using swarmers as fodder is good to mix things up. This is also a warning; if you make a swarmer you should make sure it is actually dangerous in groups, otherwise it’s just fodder.

Except for that, though, the fodder role isn’t as compatible with any other role (as a dual-class monster). You could never have a fodder/support; that’s just a support. Because tanks have to be threatening or annoying enough to demand your attention, the closest thing to a fodder/tank is a meat shield, and that’s not particularly good at being a tank anyway because they don’t do that. However, fodder works very well in encounters alongside those kinds of monsters. How many times in Quake 4 do we see Strogg Marines backing up a Gladiator or other strong monster? We also have the Scanner in Half-Life 2 supporting pistol-wielding Combine soldiers, making them a bit more dangerous than they would normally be. Fodder often makes up the meat of encounters, after you decide what roles should be central.

I hope this shows how important the fodder role is to actual gameplay. As I said, most games these days skip this entirely. If you’re designing a game that uses different roles of monsters, you kind of have to add in at least one variety of well-designed fodder. Without them, it can be a lot harder to control the flow of your gameplay.

Roles of Monsters: The Swarmer

Swarmers are one of the more commonly-used roles, seen in most shooters even today. A basic definition is a monster that requires large numbers in order to do its job effectively. Most commonly, they’re weak monsters that have to mob you, or else they get picked off quickly and easily, but there are some other things to consider with design and balance.

The best example of a swarmer I can think of is the Cherub from Doom 3. Individually, they do low damage and don’t have much health, which you might think are the two important things. They’re also somewhat slow, but move in quick bursts; either leaping at you or evading an attack. However, this monster is almost always found in groups, in areas that are either wide with obstacles or somewhat cramped, allowing them to pin you against map geometry, or surround you.

The most fundamental aspect of a swarmer is its ability to attack in groups; not simply putting the player up against a lot of them, but some mechanic that allows them to attack (or otherwise do its job) all at once. Basically, everything about the Cherub screams “I do better in groups!” because they can’t do much on their own, but also because their evasion allows them to spread out to surround you, their small size means more can get up close at once. They don’t do much damage with individual attacks, but they can be tricky to dodge (especially with a group trying to claw you) and the damage really adds up. They are also really creepy-looking with disturbing sound-sets. Combined with their leap-and-claw behavior, that tries to bring a psychological aspect as well to make the player panic in the face of a swarm (not amazingly successful, in practice, but the intent is there).

The Manhack from Half-Life 2 is another swarmer, and one of my favorite monster designs to boot. It’s a small, floating robot with a sort of sawblade in the middle, and attacks just by ramming into you. Because it flies and is so small, it can attack you from different angles as you fight them off. It really shows that swarmers benefit quite a bit from being able to fly, so they can attack you from different angles. Somewhat like the Cherub, it announces its approach with a low buzzing sound of the blade spinning, but not in an incredibly obvious way or in a way that reveals their numbers.

Like the Cherub, the Manhack demonstrates that close-range (melee) swarmers really benefit from having some ability to get to the player very quickly. Cherubs have a leap attack, and Manhacks simply fly at the player quickly (and since they damage when they touch, they don’t have to waste time with an attack animation). By definition, they have to swarm you, and they can’t do that if they can’t close in on you. Usually quick bursts of movement or direct leaps work fine. However, in the case of the Manhack and Cherub, who are very small, they would be really difficult to hit if they were that fast all the time. The Cherub is pretty slow when not leaping and usually takes an obvious path when it leaps, while the Manhack, usually pretty fast, tends to ricochet off of walls (in a way that’s surprisingly cute for a killer buzzsaw) and take a few seconds to get its momentum back. These two let a quick player pick off individual members of the swarm after dodging an attack. You do want to give swarmers the ability to close in quickly, but in a way the player can deal with.

You might wonder why I’m not mentioning the aptly-named Swarmers from the Dead Space series, but I’m covering something very similar in the form of the Needle Bugs from Binary Domain (which allow me to make another point!). While they weren’t used in the game as often as I would’ve liked, Needle Bugs were very interesting swarm monsters. They were slow and died in a single shot, but they were tiny and usually attacked in swarms of at least a dozen, climbed over you, and periodically zapped you. The more that climbed on you, the slower you moved, and they invoked your pain animation every time they zapped you. They could be cleared off very easily by just using your melee attack a few times, but they were annoying, and what’s worse; they distracted you from other robots shooting at you. They’re a good example of an enemy that fits in both the swarmer and support role, and there’s a lot of potential for crossover there; supports that have a negligible effect, but have an additive effect with increased numbers.

Another interesting point about the Needle Bugs is how vulnerable and slow they are; they mostly depend on other monsters (not just their own numbers) to distract the player so they can get in and attack. I feel like this only works if the monsters has some support function, and/or does a large amount of damage if left alone (like Dead Space’s Swarmer). Otherwise, they’re so nonthreatening that the player doesn’t even have to bother with them until everything else is dead, then he can leisurely pick them off with his weakest weapon. I will talk about (relatively) nonthreatening enemies in another article, but swarmers are generally supposed to be dangerous in swarms.

Swarmers aren’t limited to melee-only attacks (though those are usually the easiest to pull off). The Drone from Halo shows all the characteristics of a swarmer: it’s small, fast, encountered in large groups and doesn’t do much damage per-attack. They can fly, but only in short bursts and usually cling to walls out of reach. Unlike other swarmers, though, they carry ranged weapons (weapons the player can use, no less). The details of ranged swarmers are different than melee, but the overall approach is the same; low-damage, dodge-able (but tricky to dodge) attacks from different directions.

I feel like swarmers with ranged attacks almost invariably have to fly or have some ability to get to high places, otherwise they could accidentally shoot each other and ruin their advantage of numbers. If you’ve ever played Doom and maneuvered a large amount of Imps into one single mass of monsters, you know exactly what I’m talking about; the ones in back and in the middle throw their fireballs, but just hit other imps and don’t do any damage to the player. To counteract this, ranged swarmers need some mechanic to attack all at once, like the Drone’s ability to fly and cling to walls. To be an effective swarmer, the ability to shift positions quickly is also pretty nice to have, otherwise they easily get picked off. Monsters with a generic hovering ability, as long as they have some mechanic to spread out, works as well.

Melee swarmers also have the habit of surrounding and mobbing you all at once, limiting the area the player has to move around. Ranged swarmers don’t try to (or at least don’t have to) get close in order to do their damage, so the player is more free to move around. In both cases, the player has to move or else get hurt, but with melee swarmers, the area you have is constantly changing and/or shrinking. That’s an advantage for them, but it comes with the disadvantage of actually being easier to kill as they get closer. Try to kill a Manhack at long range then try to kill one while it’s close up, and see the difference.

I did say earlier that swarmers need some ability to attack all at once, and here’s a good example of what happens when they can’t. Some would disagree with me that it doesn’t fit the role, but the Lost Soul from Doom (not 3, just 1 and 2) still does illustrate my point pretty well. Except for one horrible flaw, they were decent swarmers; they played no sound when they saw the player, and quickly rammed into him for damage. They were annoying to deal with, but that one crippling flaw was that one Lost Soul hit another, they would attack one another. The infighting in Doom is usually a great feature, but in the case of the Lost Soul it punishes a monster that wants to attack in swarms for attacking in swarms. If one at the back of a cluster charged at you and accidentally hit another, it took those two out of the fight because the victim would attack the attacker and they would try to kill each other. It also made the Pain Elemental (a support that only existed to throw Lost Souls at you) not as threatening as it could’ve been, because if it threw two at a row, it was very easy to get them to attack one another, or even get the Pain Elemental infighting. Beyond that, they were also a little too large and easy to hit. Aside from those two points, they worked really well; I’ve disabled infighting and reduced their size before, and they’ve been just as effective as Manhacks.

Lost Souls do bring one interesting element to the table; they’re sneaky. Because they don’t play a see-sound (like every other monster does), the player could walk into an area and be under attack before he knew they were there. They were also the smallest flying monster (though honestly, they were the same size as the player) which let them get around more easily. A good mapper could use these two elements to his advantage and surround the player before he can react.

The main difference between a swarm and a single threatening monster is that as you fight it, the threat lessens. A single threatening monster is just as dangerous when it’s almost dead as when it just enters combat (barring any low-health gimmicks); when you’re fighting a swarm, you’re constantly killing off individual members, reducing the potential damage you’re taking. It’s a difference a player might not immediately notice, but it is a difference can make an encounter more fun. An inexperienced player also tends to think of swarms as more threatening than a single strong monster, because the amount of monsters is a more easily-noticeable factor than amounts of health or damage.

Swarmers also do well combined with other roles. While swarmers can’t be tanks and tanks can’t be swarmers (they conflict by definition), a tank backing up a swarm can work very well. Doom 3, for example, uses Mancubi and Cherubs in tandem every once in a while. The Mancubus demands your attention with its bulk and damage, and gives Cherubs the opportunity to attack more easily. Supports and swarmers work pretty well as well, with the support handicapping you so the swarmers can attack even more effectively. In Half-Life 2, a few Poison Headcrabs combined with normal Headcrabs makes for very, very dangerous combinations, for example. The Needle Bug is also a good swarmer/support, making other enemies more dangerous through a mechanic enhanced by their sheer numbers.

Because they can also benefit from the element of surprise, scripting in a map can also help swarmers quite a bit. The Trite from Doom 3 is an incredible example of this; since it’s a small, spider-looking monster, it can climb out of anywhere, or drop down from ceilings on webs. It wasn’t very mobile besides that, however. It would’ve been great if it climbed on walls and leaped at you off of them to attack from different angles, for example, but as it was it had a tendency to get bottlenecked and end up attacking you one at a time, unless the area was open enough. As it stands, scripting helps the Manhack, as well, since they could appear from places the player can’t see and can attack from different angles, and the only warning was the tell-tale whirling.

Speaking of bottlenecking, the area in which swarmers are used also very important. If the area in which you fight them is open enough, non-flying melee swarmers probably aren’t going to pose much of a threat. If the area is a cramped hallway, swarmers of all types might not be able to take advantage of their numbers, so they’ll line up and attack one or two at a time, unless they’re small enough. You can either sculpt the area around the fight you intend, or try to sculpt the encounter around the surrounding area, but either works as long as you’re considering how the area will affect the fight. Melee swarmers and ranged swarmers can both exploit different kinds of areas, as well. Ranged swarmers usually want to attack in more open areas where they’ll all be able to attack from different angles (the first encounter with Drones in Halo 2, for example), while melee swarmers want to be able to pin the player against an obstacle of some kind (think of most of the Manhack fights in Half-Life 2).

If a swarmer is well-designed, situations involving them are more easily balanced by adding or removing some number of them. Because the main difficulty of a swarmer is specifically how many there are in an encounter, you can adjust the difficulty of the encounter by adjusting the amount in the swarm. This is especially useful in games like Doom and Quake that have different monsters or ammo placed on maps based on the difficulty level, rather than just damage-scaling. This is one of the characteristics of swarmers, though there could be other factors you can use as well if they have some other gimmick. For example, I once made a stealth-based swarmer where someone can customize their behavior when placed (i.e., whether or not it will try to be stealthy, etc.). The Cherub even has some code that implies it was going to be able to play dead, as well. As I’ve said just above, it also depends on the area they’re used in.

So now that these have been broken down into basic elements, you’ll probably agree swarmers are one of the most common roles we see in games (not limited to shooters; they’re common in strategy games, RPGs and other genres). Warhammer 40k: Space Marine, Binary Domain, Half-Life 2, Singularity, the Metroid Prime series and Red Faction: Armageddon are just a few examples of recent-ish shooters that use them. They’re relatively simple to design, simple to balance, and can provide the player with encounters that differ quite a bit from what other roles provide, especially when mixed with other roles.

Definition of Terms

A friend suggested I define some terms or concepts that I use, so people don’t have to wonder what I’m talking about, they can actually check. A lot of them will be common-sense for those who have been playing games, and if they are for you then just skip it. I’ll keep adding onto this as I run into terms I should define.

First-person: The view in-game is from the perspective of the player; through his eyes, etc.

Third-person: The view in-game isn’t necessarily through the player’s eyes. Usually it’s over his shoulder.

Mechanics: Some distinct element of the game, usually a set of them define a genre of game. Some examples include upgrade systems, a first-person view, or even reloading on weapons.

Shooters: Genre of game where you have a gun and you shoot things, usually centering around action. I wish it was as cut and dry as that, but later will throw in comparisons to set them apart.

Role-Playing Games: The role-playing games are classically defined as a game where you choose a role and play as that. They are usually accompanied by a set of mechanics, such as creating and designing a player character from scratch, acquiring experience points to progress in power (leveling up), or performing tasks and missions (quests) for neutral people (non-player characters). Some games use some of these classic RPG mechanics, but aren’t actually RPGs.

Monster: Kind of basic, but it’s the generic term for some enemy in shooters, RPGs and other games. For example, I’ve used it to describe enemies in Binary Domain, which are just robots.

Open World: Usually used more in role-playing games, but used in some shooters as well. Open-world games usually give you objectives, but don’t really care how you accomplish them, and in some cases don’t care about the order in which you accomplish them. The first half of Crysis 1, Red Faction: Guerrilla and Fallout 3 are good examples.

First-person shooters with RPG mechanics vs first-person RPGs: Some people have got on me about the difference between first-person RPGs and FPSs with RPG mechanics. For example, I say I’m not talking about RPGs, but I mention Deus Ex: Human Revolution. The primary difference between these two genres is how quickly things die when you shoot at them. In Deus Ex, if you shoot something with an assault rifle it’s going to die pretty quickly, but in Fallout 3, you can sometimes shoot things in the head several times before they die. The other difference is how central the RPG mechanics are to the game. For example, in Fallout 3, the game revolves around getting new, better equipment and leveling up individual stats, and even your weapon skills revolve around those stats.

Old-School: I keep hearing this term used as though it means “limited and crappy,” but whenever I use it, I mean it as a game of a particular style that was more popular years ago. For example, Quake is “old-school” compared to Rage because it’s of a style that was more popular back in ’96 than it is now. Nothing to do with whether or not it’s more fun or better, just using it to refer to style.

Player Upgrades: Player-Driven Progression

The second kind of progression, player-driven progression is where the player picks and chooses what upgrades to get. It’s usually accomplished through a purchase system of some sort, though some first-person shooters with role-playing elements use this for their RPG mechanics (Deus Ex: Human Revolution is a great example of a game with a shooter’s balance and gameplay with RPG mechanics, as opposed to Fallout 3 which is a first-person RPG with guns).

The best example of this is the original Dead Space, where every single weapon (except the first weapon, the Plasma Cutter) and upgrade has to be bought. Weapons and new suits (armor and inventory upgrades) are purchased with credits you find scattered in maps or dropped from enemies, while upgrades to those weapons and to yourself (air meter, health or other categories) are accomplished through power nodes you find or can buy with a large amount of credits. The player does find schematics throughout the game for new things that can be unlocked, but that’s not as much linear progression because it’s up to the player to actually unlock those himself.

There are two methods of currency I’ve seen in recent games for purchasing upgrades in this manner. The first is where it takes a certain amount of some common resource (which I’ll refer to as a money system) and the second is where it takes just one of some much-rarer resource (a token system). Dead Space uses both; it uses credits as a money system (used to buy new weapons and suits), and power nodes as a token system (used to upgrade those weapons and yourself). Often times if a game has both, it’ll let you use your money to purchase a token, as Dead Space does by letting you buy power nodes with a large amount of credits (Deus Ex: Human Revolution and Singularity are good examples as well). There are some benefits to both.

Money systems allow players to get a more steady stream to inch more steadily to an upgrade, so they feel like there is constant progress towards it. It can let the player gauge how long it will take to get a particular upgrade. If you hide money around maps to reward players for exploring (like is often the case), you can also add a surplus to reward inquisitive players. Now, the big problem with that is that you can often have a big surplus towards the end of the game. In Deus Ex: Human Revolution (even without the DLC that gives you 10,000 credits), you can still get enough credits to buy the ammo you might need as you need it, weapons and weapon upgrades you want, and all the praxis points you can buy, and still have a surplus at the end of the game. Also, when I talk about money systems, I mean specifically referring to upgrades; not buying ammo, health items or things like that. However, in these games where you can use currency for items like that, you can often have a large surplus of those items anyway.

Some games tie the money system in with killing monsters. In both Binary Domain and Crysis 2, for example, that was the only method you have of getting that currency. You might have to be careful with this, though, if you have potentially-infinite monsters because that could allow the player to just farm money. In the case of Crysis 2, there were limited monsters and only a few varieties gave you that currency. Binary Domain only has infinitely spawning monsters in one or two spots, but you definitely do not want to stick around there because you can run out of ammo or other resources. There are some arguments about this, but I don’t think it’s a good thing to let the player farm money for upgrades in shooters, specifically. In older RPGs farming money for armor or weapons was part of the gameplay and no problem at all, but I feel like it’s a problem in shooters because it can screw up the game balance. If the player can spend some amount of time when a group of upgrades become available getting all of those upgrades, then he can screw up the balance. There are arguments that if the player wants to dedicate the time to that then he should be allowed to, but player-rights-versus-developer-design is another article. I will speak about a game that gets around this problem in the last article of this series.

Make sure there aren’t problems with how you give the player this money, if they get it from killing monsters. Binary Domain just adds it to your counter, but Crysis 2 has some problems. Monsters just drop it when they die and you have to walk to their corpses to collect, but it’s actually just a separate pickup that’s very hard to spot, and if you don’t pick it up within a certain time period then it vanishes. In a game that rewards you for playing carefully, you don’t want your rewards vanishing. It would have gone better if the bodies themselves gave the currency (would’ve fit the story better, as well), and had some more obvious indicator that the currency was available from that body. Dead Space, I think, got it right; monsters can drop credits when they die, but you have to pick it up, the pickup is pretty obvious, and they never vanish.

Token systems, however, feel a lot simpler; you find an item, you can buy an upgrade. However, it’s a trickier balance between how important each token should be in terms how strong the upgrade is, or how rare the tokens are to get. While this is a subjective opinion-piece, one other pitfall I think token systems can fall into is when a single token won’t buy you the upgrade. Dead Space had an upgrade tree with its power nodes where you could only move onto more upgrades by fitting power nodes into adjacent slots to that upgrade.

Unfortunately, this meant most of the power nodes you were using… weren’t being spent on upgrades, they were spent on “empty” slots to get closer to slots that actually contained upgrades. With token systems, you want to give the player more of a payoff every time, because he’s usually working so hard to find or acquire these tokens (though they were more common in Dead Space, but you still probably weren’t going to be able to max out a whole weapon and your suit). Singularity had this to a much lesser degree with Weapon Tech upgrades. Each weapon had three categories (usually reload speed, damage and clip capacity) that could be upgraded twice. The first upgrade takes one Weapon Tech, and the second upgrade takes two. Deus Ex: Human Revolution did the same thing, with some augmentations taking two praxis points to unlock. That’s not nearly as bad, however, because the player is still getting that payoff because it requires that many to get the upgrade before it takes both, rather than having to put one in specifically for nothing. It’s like the game telling you “Okay, you just spent that point and got nothing for it!” rather than “Okay, this upgrade is good enough to take a bit more than the usual cost to upgrade.” It also blurs the difference between money and token systems, which should be distinct from one another because they have their own strengths and weaknesses; when they combine, they get most of the weaknesses, but the strengths don’t usually combine as well.

Another major difference between money and token systems is that the upgrades have to be balanced differently depending on which you use. With a money system, you can make an upgrade more powerful or more useful than others, and then balance it by making it cost more than other upgrades. For example, Singularity had several very useful upgrades (more maximum health) that cost more than others, and some that are downright useless (more breath underwater) but didn’t cost much. With token systems (with every upgrade costing one or maybe two tokens), the upgrades have to be balanced around that, so they’re roughly equivalent. This works better with games that allow multiple play styles, like Deus Ex: Human Revolution, as all of the upgrades (with a rare exception) are equally useful, just in different ways.

There’s one major downside I’ve found to this method; it’s much, much harder to balance your game if you exclusively use this method. Several people say Dead Space gets too easy as the game goes on, and this is because the developers can’t assume the player has any specific upgrades. The developers have an idea that the player could have the level 5 suit at the end of the game, but he very well might not have bought it, or might not have even picked up the schematic. Though slightly harder, every enemy can be killed with the Plasma Cutter because the developers made sure you can, as they just don’t know what weapons you bought and what you haven’t when they place those monsters in the map. Granted, while the game is pretty hard if you don’t get any upgrades, if the developer assumes the player has a certain level of strength, he risks the game being unfinishable because the player might not have what’s necessary to progress. Again, at the end of this series when I bring all of this together, I will give an example of one game that gets around this problem perfectly.

The good thing about this system is that it gives the player the freedom to upgrade what he wants to upgrade. If he’s good with a particular weapon or if there’s something keeping him from using it well, he can upgrade it. If there’s a weapon he doesn’t like the look of, he can skip it. It can potentially let the user have the game cater to his play style, rather than forcing a particular play style from him. Deus Ex: Human Revolution is famous for having so many upgrades catering to different play styles so you could run through the game as an ordinary shooter, as passive stealth, or as an assassin. Some of those upgrades were outright useless, but the majority allowed you to play in different ways.

An increasing amount of shooters over the last decade have been using player-driven progression more and more. If you’re implementing this kind of progression in your game, it should be carefully thought out and balanced. I hope this breakdown helps readers to understand this system a bit better.