System Before Immersion
Or: the same design idea was argued twice on the same website in the same decade, once without ever needing the word immersion and once without ever needing the word system &mdash and the word lists prove it almost mechanically.
Lex · October 2026 · sister piece to Consistent Rules, Not Realistic Worlds; every quotation below was re-fetched and matched against the plate (live HTML or Wayback replay, fetched 14 October 2026), and every count in this piece is reproducible from the disclosed boundaries.
The previous essay in this series read the immersive-sim shelf: Bare’s formula (“consistent rules… as outlandish as you want”), Kalthoff’s house of cards, a genre defining itself in the word immersion. What that reading noticed but could not yet count was a strange absence: those texts never talk about systems. Not once, in the load-bearing cases. Meanwhile a different shelf of the same website — the systemic design essays, older, less quoted, written by working designers about combat simulations — argues what is almost the same thesis with an almost disjoint vocabulary. This essay counts the disjointness, then asks what it means that the two shelves do share exactly one load-bearing word.
The method, before the findings
Seven primary texts, all from Game Developer (gamedeveloper.com and its former name), all fetched and text-verified on 14 October 2026 — five live (HTTP 200, canonical match), two also replayed from the Wayback Machine. Each article’s body was extracted between its byline and the first site-navigation marker, tags stripped, whitespace collapsed; the counts below are regex families ( consisten*, system*, rule*, immers*) over that body alone, not the page chrome. Word totals are given so you can re-cut the boundary yourself. This is a vocabulary census, not a content judgment: a text with zero hits for a family can still describe the thing the family names. When that happens below, the passage is quoted.
2012: the conservation law, and one lonely ‘immersive’
Brent Gulanowski’s “Systemic Consistency and the Law of Conservation” (10 August 2012, a community-blog post, 13-minute read) is the earliest of the seven and the strangest. Its thesis is not a design trick but a physics claim: games fail when they contradict “the fundamental law which underlies all other laws in all parts of the universe: the law of conservation.” Magic, he is careful, “is not intrinsically bad” — what’s bad is breaking conservation, which he claims as the axiom underneath thermodynamics, mechanics, economics, information theory. And the claim does not stop at our universe:
“But the law of conservation does not apply only in our universe. It applies in all possible universes. It certainly applies in all game universes, if you pay even the slightest heed to the idea that games exist to help players learn.”
The subtitle does the design-theory work: too many games “don’t adhere to internally consistent logic. This makes them dumb… They need to be based on meaningful rules which can be mastered.” The census, over 2,693 words of body: consisten* 10 · system* 8 · rule* 8 · immers* 1.
One. And the single hit is worth its weight in context, because it is not theory at all — it is a purchase wish, the one place the word appears:
“I want games for intellectual players who want to consider intellectual problems, ideally coupled with immersive, first-person, character-driven gameplay.”
For Gulanowski in 2012, ‘immersive’ is an adjective you shop for; ‘consistent’ is the concept you argue for. The design theory runs entirely through rules, systems, conservation, and learning. There is no house of cards in this text, because there is no spell to break.
2017: the Swiss Army Knife XL, and immersion that belongs to the developer
Five years later, Trent Polack’s “A Systemic Approach to Game Design: Part Three” (7 February 2017) is the practice-shelf answer to the same question — what does the player get when the simulation never contradicts itself? — and it is written as shop-floor prose from Joy Machine’s mech game: the foundation must “hold up everything that is to come,” the systems must be joined so thoroughly that
“it doesn’t matter to the game simulation: every tool will work.”
The stated promise to the player is Bare’s improvisation sentence in workman’s clothes a year before Bare, with the genre word absent:
“players never feel penalized for thinking outside the box of trying combinations that were completely unthinkable to us.”
The census, over 2,925 words of body: system* 15 · consisten* 1 · immers* 2 · rule* 1. And those two immers* hits are the most honest data point in this whole piece — both describe the developer, not the player:
“you’re so immersed in your foundational task that it’s almost hard to see the light at the end of the tunnel… it means you’re so involved and so immersed in ensuring that the foundation is exactly what it needs to be.”
In 2017, on the systemic shelf, the word immersed still means the programmer at 3 a.m. It has nothing to do with what the player experiences on screen. Whatever the immersive sim was becoming as a marketing noun, the working systemic designer’s vocabulary of the 2010s had no use for it as a technical term.
The census: two shelves, near-disjoint word lists
Now hold the first two shelves against the immersion-shelf texts of the same years — the 2017 GDC roundtable report (Wawro), Dunniway’s immersive-puzzles post, Kerr’s 2018 report of Bare’s Prey formula, Kalthoff’s 2019 house-of-cards post, and McAndrews’s 2018 genre definition (full counts in the table; every fragment re-verified against the plate at publication). The numbers are the essay:
| Text (body words) | immers* | system* | rule* | consisten* |
|---|---|---|---|---|
| Gulanowski 2012 (2,693) | 1 | 8 | 8 | 10 |
| Polack III 2017 (2,925) | 2* | 15 | 1 | 1 |
| Wawro roundtable 2017 (312) | 7 | 2† | 0 | 0 |
| Dunniway 2017 (2,909) | 19 | 4‡ | 1 | 0 |
| Kerr on Bare 2018 (635) | 11 | 0 | 4 | 1 |
| McAndrews 2018 (1,585) | 2§ | 0 | 2 | 0 |
| Kalthoff 2019 (1,969) | 17 | 0 | 10 | 2 |
* both hits = the developer’s own immersion in the foundation, quoted above. † the roundtable’s two raw system*-hits are the game title System Shock in the lineup and the credits block — a false positive, and a fitting one: on the immersion shelf the word ‘system’ survives mostly inside product names. ‡ all four = ‘game systems’, ‘hint system’ — features, never the design-theory term. § a curious count of its own: the article is about immersive simulations but uses the word family almost nowhere in its body — the definition, “an interactive story with great freedom,” is the load-bearing sentence, and it spends its word-budget on story and freedom, not on immersion or systems.
Read down the columns. Immers* is rich on one shelf (7–19) and dead on the other (1, 2 — and the two are about the programmer). System* is the mirror: 8 and 15 on the systemic shelf, 0 on three of the four immersion texts (the roundtable’s two hits and Dunniway’s four are named in the table notes — the noun-phrase and the feature-list kind, not the theory kind). Nobody said the two shelves can’t hear each other; the word lists say they mostly weren’t speaking the same language.
And here is the detail that keeps this from being a story about two camps: the two texts that carry the theory load hardest on each shelf share the one column that is non-zero everywhere it matters. Kalthoff lists among his ten spell-breakers, at number two: “Your world isn’t consistent in its ruleset” — rule-family 10, immersion-family 17, and the diagnosis is exactly Gulanowski’s, three thousand words and seven years and one entire vocabulary away: games fail when they break their own rules.
The bridge: consistency is the contract; immersion and system are dialects
What the census shows, then, is not that one shelf ignored the player’s experience and the other ignored the simulation. It is that one 2012 physics blog and one 2017 mech-game diary and one 2019 immersion-dispelling post arrived at the same replacement test — the world must keep its own rules, and the player can only be trusted in a world that does — while routing to each other through three different vocabularies: conservation, systems, and immersion respectively. The marketing noun (‘immersive sim’) won the language game; the concept underneath it was argued in the dullest possible word, consistency, which is probably why nobody filed a patent on owning it.
This is also why the roundtable’s two system-hits matter out of proportion to their count. Asked what the genre is, the camp’s founders reach for the player’s freedom, not the simulation’s plumbing — Spector: “we’re the first mainstream medium that can actually do that. And the immersive sim is the perfect way to do it”; Gaynor, joking into the same silence: “I think we should start calling them instead of immersive sims, probably digital LARPing.” LARP is a systems description wearing a theatre costume — rules all the way down, no conservation law in sight. The camps keep circling the same object with different flashlights.
What this count is, and what it is not
Be exact about the reach of a census. Seven texts are a vocabulary sample, not a field study; n is disclosed, boundaries are disclosed, the two replay-only plates (Polack’s part three, and the roundtable’s Wayback double) are disclosed. Word frequency is not concept absence: Gaynor’s LARP line describes a rule-bound playground without the word rule; McAndrews’s ‘great freedom’ is a consistency claim in disguise; and Gulanowski’s conservation axiom is itself a metaphysics nobody downstream was required to believe in order to keep his part of the contract. The claim this piece can defend is narrower and firmer: on this website, in these years, the two professional vocabularies for the same design discipline were near-disjoint at the word level, and the overlap that did exist ran through the word ‘consistency’ and nothing else. What that says about silos, tag pages, and which posts get cited by which camps — that is a different essay, one this series has been writing obliquely all along.
Part of the game-design series: Consistent Rules, Not Realistic Worlds · The Two Contracts · Light or Shadow · all field reports