The Effects of Game Design Philosophy on Map Making

Following the recent feedback on the latest map pool, map design has received much attention these two weeks. Many discussion points seem to touch on the fundamental game design of StarCraft II, but the link between the two is not discussed explicitly. This post discusses how game design principles confine map design.


The map discussion started two weeks ago with progamers pointing out most of the current maps do not have features they perceive as must-haves. The discussion later became a debate on the input of progamers in the feedback process.

What is lost among the exchanges is the question of why must maps contain specific features? I’m not arguing for or against the implementation of Reaper jump spot and deadspace here. I simply want to question our assumption on the unspoken rules of map design.

One of the best examples of such unspoken rules is having a single ramp to the main base. The size of the ramp is also constant. Of course, there are rare exceptions, but this feature is a staple since the first set of ladder maps in Wings of Liberty. Balance is hugely designed around this feature. Imagine Zerglings flooding to your main base early on because there are multiple big ramps.

How do we decide whether a feature should be a staple then? I doubt there is a consensus on this, but it seems fair to suggest it is contingent on how much the feature is intertwined with balance design. We had witnessed the community discussing extensively whether a map feature is an integral part of balance. This is then used as a basis in debating whether every map should have that feature. What is missing from the discussion is how the game design philosophy of StarCraft II is in fact shaping the maps to what they are today.

Consistency with game design principles

Game design philosophy drives balance design, which is closely intertwined with map design. I’m going to lay down some relevant core design principles of StarCraft II, and then I will use Reaper jump spot as an example to illustrate how these principles shape map design. Here are some of the game design principles of StarCraft II that are relevant to my discussion:

  1. Game mechanics can differentiate players’ skills (i.e., we can tell who is a better player).
  2. Certain units should not dominate in majority of the situations (i.e., units should have their roles).
  3. Players have options for counter play (i.e., both sides have input on the outcome).

Reaper plays an important role in the early game of every Terran match up, because it is the primary way to gather information. Information gathering is one of the most defining features of a game that builds on imperfect information. The abilities to gather useful information (i.e., when, what, and how to scout) and make good decisions based on gathered information (i.e., how to react) differentiate the better players from the inferior ones. If a player is unable to gather sufficient information to narrow down the plausible outcomes, s/he has to roll the dice in decision making. Leaving it to luck goes against the first principle I stated above, as luck makes it difficult to differentiate players’ skills. Consistent with this, Blizzard had made changes related to scouting in recent patches. In the most recent patch 4.10.1, Overlord’s Pneumatized Carapace upgrade research cost decreased from 100/100 to 75/75. Their justification is, and I quote, “hope this change will address the difficulty Zerg has in the early game, when scouting their Protoss opponents—specifically, their ability to differentiate between tech choices and the various Immortal-based timings.” Similarly, in patch 4.7.1, Sentry’s Hallucination energy cost decreased from 100 to 75, hoping that this change can reduce the luck factor in the PvP early game. The hot fix to the Reaper jump spots in the latest maps suggests Blizzard considers Reaper’s accessibility to main base is a key map feature now.

Why must Terran use Reaper to scout? This is because Blizzard fixate it this way! Reaper’s KD8 Charge was nerf in patch 3.11 and 3.16. ByuN was the culprit behind these changes, as he was demonstrating how effective triple Barracks opening was in TvZ. Importantly, almost no one had as much success with this opening as ByuN, not even Maru. Hence, this was not precisely a balance change for TvZ, but rather Blizzard perceive mass Reapers in early game did not align well with the unit’s designed purpose (source). These changes to Reaper highlight how much Blizzard do not want the unit to be used beyond its scouting purpose. The same strong arm approach can be seen in the changes to Raven and Oracle, whereby players’ innovative use of these units is discouraged. They are sending a message that Raven should only be used as a support unit (I wrote an extensive review on this), and Oracle should only be used as an early game harassment unit (for reference). When you confine Reaper to a scout-only role, you are also confining the map design by ensuring Reaper is “playable”.

The third principle I mentioned above is rather interesting in the context of Reaper jump spot. The debate revolves around the amount of investment Protoss should exert to prevent Reaper from jumping into the main base. Based on Blizzard’s hot fix, it is safe to conclude that a single Pylon block is considered unfair against Terran. The main takeaway here is not whether a single Pylon block is fair, rather it shows how much the fundamental game design principles influence map design. That is, both sides can influence the outcome of this interaction.

Consequence of the influence

Generally, there are no good or bad game design principles. It is a matter of appealing to different audience and keeping it consistent in design decisions (e.g., balance). However, as discussed in the previous section, these fundamental design principles confine map design by setting unspoken rules indirectly. This is not a RTS specific issue, as we have seen other games use different approaches. For example, players in Age of Empires do not spawn on the same locations of a map. Brownbear’s video below explains this concept very well (timestamped).

The influence of game design principles on map making in StarCraft II can explain much of the similarities across the maps. On one hand this provides consistency in the early game (related to game design), but on the other it limits creativity of the map makers. It is ironic how players complain about the maps being more or less the same all the time, but they also do not want to play on maps that do not resemble what are considered “standard”. I brought up this issue in an article I wrote six years ago, where I discuss the dilemma of change in StarCraft II. In it, I quoted an interview with a mapmaker:

That’s an interesting point that you bring up there. Diversity and difference. I wrote an article last week that even was upvoted to the frontpage of reddit. It fits your opinion about the “stale” map design in SC2 so far. Could you go into detail about the 3-base paradigm from a map designer point of view?

The 3-base set up is a cancer to the design of maps. It’s not that it’s common or it’s the only thing we can do, but it is technically the only viable and consistent setup that can produce any of the existing playstyles. A far third is likely not to be taken by terran or especially protoss, and a natural that’s too far away or has a larger choke is instantly veto’d by protoss players because now they can’t FFE. It’s stupid stuff like this that plasters a “broken” sign all over the map. Can you really blame the map? Or does it really lie in the hands of the player’s choices? Or, perhaps, are the races just fundamentally flawed and need to be re-designed in some way? You can’t really blame us mapmakers, we’re only creating what the balance demands and what your common playstyle is asking for. If we can’t meet those basic everyday demands, what makes anyone think an entirely new, experimental concept is going to pass?

You can tell that the mapmakers felt confined by the combination of players’ preferences and game design principles related to racial asymmetry. What I find interesting is how these restrictions are “enlarged” as time progresses. This is particularly obvious when you revisit maps from the early Wings of Liberty days. Let’s use Xel’Naga Caverns as an example (see image below). The natural is extremely vulnerable. It has a wide enough entrance at 1 and 2 that you can’t wall off early. There are alternative paths at 3 (destructible rocks) and 4 (bush).

Now compare that with Shakuras Plateau (see image below), which is in the same map pool with Xel’Naga Caverns for the first five ladder seasons. Its natural resembles the design of current maps, whereby there is only one ramp entrance at the natural. This demonstrates how much maps differ from one another in 2010. Specifically, the main base design for the two maps are very similar, but it is the natural design that differs. There were no unspoken rules in designing the natural. This could be traced back to Blizzard’s thinking that smaller maps lead to more engagements, which in turn makes it more appealing as a spectator sports.

However, this design did not persist, because it was clear that Zerg is inferior if the game is often limited to one base. We no longer had strict rules only for the main base, but also the natural. This then led to an era when the maps were differentiated mainly by the ease of taking and defending a third base. That is, as our understanding of the game improves, we expand our restrictions on maps. Fast forward to today, the difficulty of taking a third base in the current maps is very comparable. This evolution coincides with the design rationale behind the economy redesign of Legacy of the Void, and it is heavily influenced by the three base paradigm in StarCraft II (for reference). In sum, mapmakers have less room to be creative now than before.

Challenging the paradigm

What can mapmakers do to challenge the paradigm? Or perhaps the question should be, should they? If the main goal of mapmakers is to create maps that get used, the incentive to make maps which fall within the existing paradigm is obvious. I was thinking, if my goal were to win TLMC, how would my map look like? I’ve no experience in map making. My intuition is to use machine learning to study what are the common features of the highly scored maps. This means I would end up with a map that is quite similar to the previous ones. I don’t expect this trend to change, because the game design principles of StarCraft II are not going to change.

If you enjoyed this article, I’d love you to share it with one friend. You can follow me on Twitter and Facebook. If you really like my work, you can help to sustain the site by contributing via PayPal and Patreon. You can also support me and enjoy quality tea with a 15% discount at AFKTea by using the “TERRAN” code. See you in the next article!

What do you think?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s