Orcs Must Die: Blocking Strategies

This post was requested by a patron. To support writing like this and request your own post, sign up on my Patreon page.

I have a confession: I’m a turtler. And Orcs Must Die 2 doesn’t want me to be.

The game I think of when I read “tower defense” is “Desktop Tower Defense.” It’s my mental archetype of that form, which arose from mods for Starcraft and Warcraft III: a game about preventing waves of invading enemies, or “creeps,” from reaching the exit of a map using stationary towers that attack when the creeps come into range; these towers are built with a budget you earn by killing creeps.

The strategy for “Desktop TD” is primarily about crafting a path for the creeps, one which is circuitous as possible. The towers in “Desktop TD” are solid, so they block creeps, making your towers also serve as your maze. Mastering “Desktop TD,” therefore, requires you to craft a perfect maze, a gleaming labyrinth made from the cheapest towers with just enough addition of special tower types and more powerful, upgraded towers.

Some tower defense games, like “Desktop TD’s” contemporary “Flash Elements TD” have a simpler approach where the creep paths are static and unobstructable. Towers can only be placed in the spaces around the path. I find this approach less interesting, as it allows for less creativity and diversity of play. The most a player can do to affect the process of the creeps, beyond killing them, is by slowing them, often with a tower themed around ice or viscous fluid. Orcs Must Die and the other games I’ll discuss here owe more to the “Desktop TD” style.

The Orcs Must Die series by Robot Entertainment belongs to a subfamily of tower defense games, probably birthed by Sanctum. These tower defense hybrids add a mobile player character with weapons that can supplement the stationary towers. In the case of the Orcs series, the player character is a martial wizard defending a fantasy world against hordes of orcs and other creatures. But unlike Sanctum, its differences go beyond just letting you help your towers with their work.

Dominant and Degenerate

Games can contain overwhelming strategies that work drastically better than other play styles. Economic game theory calls these dominant strategies (Rollings and Adams), but in game design some folks call them degenerate strategies (Salen and Zimmerman). If you want to win, these strategies are great, but they’re “degenerate” to the designer because they can exclude other approaches.

In “Desktop TD,” building perfect, automated defenses is what you’re meant to do. This play style, called “turtling” in some games, is the point of the whole game. It’s not a degenerate strategy because other strategies aren’t the focus of the game.

When you add the ability to run around and help the towers, you suddenly have another avenue of play to worry about. If turtling remains a dominant strategy, then you can build a maze of towers so strong that it stops all the enemies without your help. If so, what’s the point of your PC? As a designer, if you care about that part of your game, you have to balance your rules to remove that strategy’s dominance, preventing it from being degenerate.

Sanctum does this by making the towers weak enough that they need your help to defeat the waves of creeps. That’s a precarious balancing act, though. If the towers are too strong, then an experienced player can turtle. If they’re too weak, they feel flimsy and unsatisfying and the game can become frustrating. I think Sanctum succeeds at walking this knife’s edge, but the Orcs series takes a different approach.

Traps and Tricks

The first two Orcs Must Die games replace towers with “traps.” Traps are distinct because, while they’re still placed in the creeps’ path like in “Desktop TD,” they don’t directly block creeps (with an exception I’ll discuss later). They can deal damage, slow creeps, and toss them around with physics, but they can’t form an elaborate maze.

At first this seems to shift Orcs into the “Elements” category, where the creep path is set and you just have to do enough damage along the way. That discounts the degree to which traps affect the creeps. Unlike with many tower defense games, the creeps collide with each other, damage can stagger creeps for a bit, and many traps toss creeps around. You can create bottlenecks with clever trap placement where waves are slowed and concentrated and killed en masse.

Additionally, traps are very position-based in their targeting. In other games, towers can attack anything within range, but traps in Orcs only attack when creeps pass over them or otherwise into their trigger area. You can buy “guardians,” characters that attack anything within range, but guardians are vulnerable to attack from the creeps if the creeps can reach them. Thus, while you can’t build a maze out of traps, you still must focus on the path creeps take. To help with this, many of your direct weapons have primary or secondary effects that stall enemies or push them around.

There is one trap that blocks creeps: the “barrier.” However, the barrier first becomes available late in each game, and they’re relatively expensive: one small barrier costs as much as a middle-tier trap even after you apply upgrades that make it cheaper. Additionally, barriers can be damaged. If a choke point becomes too choked, creeps may attack the barrier to make another path. There are also kobold sappers that will target barriers and explode to destroy them. Reliably preventing this requires more traps, further increasing the cost of barriers.

This all disrupts the turtling strategy, preventing it from becoming degenerate. Barriers are still tremendously useful; in co-op Orcs Must Die 2 my partner and I handily beat the final level with the help of four well-placed ones that delayed and redirected half of the waves. You just have to be judicious in their placement.

Layers of Balance

This design pattern pervades Orcs. The games add a new option, then prevent it from becoming dominant by adding a downside. Checks and balances, advantages and tradeoffs. This is a great balancing process. Sid Meier advocates viewing games as “a series of interesting decisions.” Orcs heads off boring decisions (“Should I make the winning play?”) not by defusing them but by making them interesting (“Is it worth it to make this play?”).

The potential dominance of traps that toss creeps into certain death is prevented by including large creeps that can’t be tossed. If the player hides behind barriers or places guardians out of the way, there are occasional gnolls that will jump barriers and seek out players and guardians, requiring vigilance and care to take out. These downsides make choices more fraught and, therefore, more interesting.

This layering approach to balancing isn’t new in the tower defense form. Some of the earliest tower defense games, including “Desktop,” send out occasional flying creeps that bypass all barriers and can’t be targeted by some towers. These creeps serve to disrupt any overly-dominant strategies that exploit normal creep movement. Likewise, the emergent strategy of juggling in “Desktop,” where players constantly destroy and rebuild towers to get creeps to backtrack, was made harder in later updates by increasing the amount of time it takes to sell or build towers.

These early counters to dominant strategies survive in Orcs. Flying creeps take a different path than walking ones, requiring player attention or guardians to reliably take out. Traps can only be sold between waves, often in a very brief time window, preventing a player from rearranging their gauntlet to focus on a single problem at a time. Just like Orcs builds layer upon layer of complications for its own strategies, so do the generations of tower defense games build layer upon layer of conventions to balance their shared rules and potential strategies.


What are your favorite examples of games layering on complications to disrupt degenerate strategies? Are there any games, especially tower defense games, that you feel do an especially good job of balancing themselves? Please share your thoughts and feedback in the comments.