10 min read
Featured Blog

Design for Theorycrafting

What if instead of discouraging players from communicating with others on discussion boards, or encouraging blind playthroughs, we can design games around the reality of gaming today?

Recently, Danny O’Dwyer and the team over at Noclip have done a documentary series, interviewing prominent independent developers (Derek Yu of Spelunky, Jim Crawford of Frog Fractions, and Jonathan Blow of The Witness), talking about the games that inspired them as children, discussing their own takes on the significance of mystery in games, and explaining how they weaved elements of it into their design fabrics. The documentary series is fittingly titled “Rediscovering Mystery,” and a solid must-watch.

It takes courage for a developer like From Software to even attempt to include such a vast, and totally missable areas such as the Painted World of Ariamis or Ash Lake in the original Dark Souls. And there’s no doubt, that the effort from Jonathan Blow and his team, to build an island full of mystery and surprising treats, elegantly placed on the fine line between obscurity and conspicuity is absolutely commendable. However, aside from being a big help when you need it, the Internet is also a land riddled with spoilers. People are always discussing, debating, and engaging in joint-efforts to unearth even the most obscure secrets of all in games. And often as a result, the easy and most logical answer from developers on how to retain the player’s interests in games is to keep expanding their worlds, and to include even more obscure secrets for them to find out.

I do believe that this approach is, not only costly for sure, but also fairly limited in its efficiency. Not to undermine developers’ effort and control over the experience, but there’s one thing that many designers I know will probably agree on: the most profound, interesting, valuable and effective drive of all stems from within people’s heads. And the most common manifestation of such drive that one can see, comes in the form of theorycrafting.

A spoiler-resistant design paradigm

According to Christopher A. Paul (2011), the word theorycraft originated from World of Warcraft, which describes “the search for the optimal set of strategies with which to play, by using statistical analysis and mathematical modeling.” In the broader context, theorycrafting refers to the practice of analyzing theoretical scenarios, speculating possibilities, performing statistical reports, planning strategies for unexpected events, or simply “connecting the dots.” All of these are done mostly off-game, or at least not within the context set up by the games they’re playing.

Take a look at almost any game, you’ll find a universal advice suggesting that starters do a “blind” playthrough first, then come back a second time with a strategy guide to “one-hundred-percent” the game. While there’s nothing wrong with avoiding spoilers before you can enjoy a good book or a good movie, I do think that the old method of adding depths and maintaining player’s interests no longer holds itself well. What if instead of discouraging players from communicating with others on discussion boards, or encouraging blind playthroughs, we can design games around the reality of gaming today?

Minimal handholding

While this is a popular mantra adopted by many modern designers and may seem ostensibly obvious, I’d still like to stress its importance. Instruction on basic controls and maneuvers is fine, but make it so that players will have to do their own homework in order to overcome bigger challenges (say, higher difficulties). They can take notes, read up, discuss with other people, or even come up with their own ideas and strategies. You may be told how to move around and perform various attacks in Dark Souls, but nowhere in the game will you find an explanation on how Poise works, or how sometimes having low-to-no Poise is actually beneficial. The user interface and functionality of Darkest Dungeon may imply that you should treat your units as precious and cure them of devastating mental afflictions, but you’ll have to learn for yourself that it’s really costly to maintain a sane army, and the path to victory involves treating your units as expendables.

I myself have spent about 1,500 hours on the Long War mod for XCOM 2012 by Firaxis, and tried to pinpoint exactly what about it that reeled me and thousands of others in, even during its development period, and how it had us all absolutely enthralled. XCOM 2012 was fine, but it certainly, most would agree, did not have much to keep people playing for hundreds of hours. And that’s where the mod steps in. In direct opposition to Sid Meier’s idea at GDC 2010, which is basically to occasionally “cheat” in favor of players so that they themselves won’t ironically feel cheated, Long War relieved XCOM of every last bit of its shackles. As a result, the game was significantly harder, but a majority of XCOM players found the honesty and transparency worth a lot more in the long run than the ocassional helping hand.

Ensure High Interactivity between Game Elements

The first thing that comes to mind when people start thinking about strategies is how different elements within a game will interact with one another. And good news is, you don’t have to make such complex network of interactions obvious. Human beings are good at noticing and distinguishing patterns, and we love playing the game of connecting dots. What would happen if I combine Element A with Element B, and then sacrifice Element C, which on its own is very beneficial, to go for Element D? Would I then have to change my playstyle now that this setup is in place? And how? These are amongst the questions players are going to ask themselves when they theorycraft. Try to think ahead of them, and leave footprints on the ground behind so that players will follow them.

Extra Credits’ episode on Depth vs. Complexity explores this particular topic further.

Encourage micro-and-meta strategies

To reiterate a point previously made: the higher the difficulty level (by which I mean the amount of depth the game can and may provide to the player) the more the game should require the player to think outside of the immediate in-context scenario being laid out in front of them and plan ahead for future possibilities. The player should have to not only care about major factors and decisions, but also minor details that at first may seem trivial, but will make a difference in specific scenarios. For example, a 5% hit chance boost may seem negligible—not a lot of difference between an 80% hit and an 85% hit. But in the specific scenario where the player’s hit chance is penaltied down to 0%, and the 5% boost will be applied after the penalties, then it becomes the difference between chance and no chance.

To do that, a game needs to require players to acquire more knowledge than it can offer explicitly inside itself (unless it decides to constantly break the 4th wall and rain down paragraphs after paragraphs on the player, which is a bad idea by the way.) Obviously, this requires the game to contain a lot more depth. People won’t have a lot to read up on or talk about if the game they’re playing is Flappy Bird!

Micromanagement in gaming is often referred to as tedious. However, I do think it can be interesting and sometimes deeply profound if the player can see that:

  1. It matters. It’s not going to matter when your SPD stat gets a 25% boost if you’re already the fastest moving unit in a turn-based combat and others have no way to catch up, which is what a lot of old RPGs were guilty of.
  2. Everybody else is doing it.
  3. The game is designed around the fact that everybody is doing it.

I found immense joy in discussing strategies with other players in the XCOM Long War community, and to play around with various ideas and theoretical scenarios in my head whenever I was on break at work. That made me feel like I was improving even when I was away from the game, and the game also helped me learn to think and to plan things carefully, and deliberately.

Reduce Linearity

Just like minimal handholding, even though reducing linearity isn’t anything new to experienced designers, it is amongst the key qualities of games designed for theorycrafting. Naturally, if every scenario in a game is scripted, every puzzle has but one single solution, and nothing that the player can do will make any significant impact to its outcome, then that game’s resistance to spoilers is clearly low. Someone nonchalantly tells a player that maybe they can turn left at this juncture, and they can’t unlearn that! You want to make sure that problems almost always have more than one answer, and there is a moderate space in which the player’s theorized solution may go off-track.

Maintain control over gameplay

With that said, in certain games, even if they’re non-linear and designed to encourage theorycrafting, there is still the risk of the player figuring out the optimal play and sticking to it rigorously, ultimately ruining their own experience. Soren Johnson, one of the Civilization designers, wrote “Given the opportunity, players will optimize the fun out of a game.” (2011)

The current most efficient ways to tackle this problem that we know of, is to introduce random elements, or in some context, procedural generation. The idea is to allow the computer to systematically and algorithmically generate environments or outcomes based on the player’s input and a dice roll. Randomness is not necessarily “lazy design,” but rather meant to serve as insurance against optimal play always producing the same results, making sure that no strategy is without its flaws. Allow people the freedom to metagaming and micromanagement, but at the same time try to mitigate certain-win strategies as much as possible. Work closely with testers and even the community.

Final Words

Designing a game that is challenging, complex, with a lot of depth, and which encourages strategizing, planning for possible scenarios, coming up with creative ways to solve problems, might seem to be an overwhelmingly difficult task. But I do think it is a good ideal to push towards. If people are going to read up on games before diving in and perhaps spoil it for themselves, let them do it as much as they want. Make it so that the challenges in your game can take that fact on the chin and still give them a hard time. And a good time.

Lastly, I’d like to leave you with a few words from Professor Brian Moriarty’s lecture back in GDC 2002, The Secret of Psalm 46.

“If super power is what people really want, why not just give it to them? Is our imagination so impoverished that we have to resort to marketing gimmicks to keep players interested in our games? Awesome things don’t hold anything back. Awesome things are rich and generous.

The treasure is right there.”


Christopher A. Paul (2011) Optimizing Play: How Theorycraft Changes Gameplay and Design. Retrieved from

Sid Meier (2010) The Psychology of Game Design (Everything You Know is Wrong). Available at

TangledAxile (2016) Yes, XCOM2’s RNG cheats - in your favor. Here’s how. Message posted on Reddit. Available at

Soren Johnson (2011), GD Column 17: Water Finds a Crack. Retrieved from

Brian Moriarty (2002) The Secret of Psalm 46. Available at

Extra Credits (2013) Depth vs. Complexity - Why More Features Don’t Make a Better Game. Available at

Latest Jobs


Playa Vista, California
Audio Engineer

Digital Extremes

London, Ontario, Canada
Communications Director

High Moon Studios

Carlsbad, California
Senior Producer

Build a Rocket Boy Games

Edinburgh, Scotland
Lead UI Programmer
More Jobs   


Register for a
Subscribe to
Follow us

Game Developer Account

Game Developer Newsletter


Register for a

Game Developer Account

Gain full access to resources (events, white paper, webinars, reports, etc)
Single sign-on to all Informa products

Subscribe to

Game Developer Newsletter

Get daily Game Developer top stories every morning straight into your inbox

Follow us


Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more