Six alternative branches. These are subjective probabilities from the seed, not observed frequencies or investment recommendations.
Why this trend
Load-bearing assumption
That the platform commission is the sector's largest single cost line, so a durable four-to-ten point reduction in the effective take rate changes what games get made rather than only where the margin lands.
Preconditions
Not recorded.
Grinding single-digit growth with a re-priced toll layer. The market grows mid-single-digits from Newzoo's $213.9bn 2026 base, mobile leads, live-service concentration persists, and commission settles in a 15-26% band with meaningful link-out leakage. Employment stabilises at a structurally lower level with a hollowed-out junior tier, generative AI becomes standard at the asset and localisation layer without changing design, and consumer XR remains a sub-$5bn niche.
Mechanism
Each jurisdiction's rate is set by a different instrument — the DMA in the EU, the Mobile Software Competition Act in Japan, regulator negotiation in China, CADE in Brazil, the Epic settlement for Google in the US, UK and EEA, and a contempt ruling for Apple in the US — so the effective global take rate falls as a weighted average rather than as a single cut. The decelerating element is that the saved margin is absorbed by per-jurisdiction compliance: fee agreements, age assurance, AI disclosure and child-safety architecture are fixed costs that fall hardest on small developers, so the pro-competitive remedy at the top of the stack is partly anti-competitive at the bottom.
Preconditions, early indicators and assumptions
- Early Indicators
- Indicator
Take-Two FY2028 net bookings guidance between roughly $6bn and $7bn, indicating the premium tentpole model is neither healthier nor worse than assumed
- Source
S-15-04 SEC EDGAR filings
- Would Be Visible By
2027-06
- Indicator
Apple's EU fee schedule still specifying a 5% Core Technology Commission a year after the 2026-10-01 effective date
- Source
S-15-01 Apple Developer News and fee-schedule pages
- Would Be Visible By
2027-06
- Indicator
Steam annual cohort of titles clearing $1m gross staying near the 311 recorded in 2025
- Source
S-15-10 GameDiscoverCo
- Would Be Visible By
2028-02
- Affected Industries
- 15
- 17
- 02
- 24
- What Businesses Should Do
Rebuild net-revenue models jurisdiction by jurisdiction rather than on a single global take rate, and measure whether link-out routes are actually being used before assuming the saved commission reaches the P&L.
- Precedence Note
Distinguished from upside by supply, not by margin: relief that lands in publisher profit without raising the count of titles clearing $1m gross stays in base.
- Would Change Our Mind
The Steam cohort of titles above $1m gross rising above 400 in a single year.
The toll cut converts into supply. Sustained four-to-ten point margin relief combines with AI asset-cost reduction to make mid-budget development viable again; the number of titles clearing $1m gross rises materially above the 311 recorded in 2025, and new-title share of playtime recovers from live-service concentration.
Mechanism
The named constraint that releases is the gross margin on the marginal unit sold. At a 30% take a mid-budget title needed a hit to clear its cost of capital; at 15-20%, with asset and localisation costs cut by generative tooling, the same title clears at lower volume. The new demand does not appear from nowhere: it is existing player spend redirected from a small set of persistent titles toward a wider release slate as discovery improves, paid for by the same wallets.
Preconditions, early indicators and assumptions
- Early Indicators
- Indicator
The 2027 and 2028 Steam cohort counts of titles above $1m gross rising above 400
- Source
S-15-10 GameDiscoverCo newsletter and data service
- Would Be Visible By
2029-02
- Indicator
Median price of top-selling PC games stabilising or rising after falling 15% from $19.99 in 2021 to $16.99 in 2025
- Source
S-15-10 GameDiscoverCo
- Would Be Visible By
2028-02
- Indicator
A Western publisher other than Take-Two greenlighting a new $200m-plus single-player premium title
- Source
S-15-17 Game Developer; S-15-16 PocketGamer.biz
- Would Be Visible By
2028-06
- Affected Industries
- 15
- 17
- 24
- What Businesses Should Do
Model a mid-budget slate at the new effective take rate before competitors do, and build the per-jurisdiction billing and compliance stack once rather than per title, because that fixed cost is what determines whether the relief is usable at small scale.
- Precedence Note
Requires the title count to move. Margin expansion at existing publishers, however large, is base.
- Would Change Our Mind
Two consecutive years of Steam $1m-gross cohort counts stagnating near 300 while effective take rates fall.
The commission relief is clawed back and demand stalls. The US district court grants Apple a commission on link-outs, other jurisdictions soften under trade pressure, consumer spending flattens in a sticky-inflation environment, hardware prices rise further on memory and tariff costs, and UGC platform engagement keeps declining under safety architecture.
Mechanism
The parameter that deteriorates is the effective take rate, reversing. Apple filed its 15%/10%/5% external-link proposal on 2026-08-14 after the Supreme Court refused a delay on 2026-08-13, and Epic's reading of the 'necessary costs' standard is 0%; both are currently defensible. On the demand side, Sony raised PS5 Digital Edition prices by $50 and PS5 Pro by $150 on 2026-04-02 citing the global economic landscape, which is tariff and memory-cost pass-through reaching the consumer.
Preconditions, early indicators and assumptions
- Early Indicators
- Indicator
The US district court's ruling on Apple's external-link commission proposal
- Source
S-15-21 Epic v Apple docket via CourtListener
- Would Be Visible By
2027-12
- Indicator
A second console price increase or a further Game Pass restructuring
- Source
S-15-11 Circana US spending releases; S-15-17 Game Developer
- Would Be Visible By
2027-12
- Indicator
Roblox Q2 2027 bookings at or below the $1,557m recorded for Q2 2026
- Source
S-15-03 Roblox IR via SEC EDGAR
- Would Be Visible By
2027-08
- Affected Industries
- 15
- 17
- 03
- 02
- What Businesses Should Do
Do not build a 2027 plan on a commission rate that a single US ruling can reverse; keep in-app purchase infrastructure warm even where link-outs are currently cheaper.
- Precedence Note
Downside is the same mechanism at worse parameter values. A ruling that permits a commission moves the outcome here; a ruling that changes who may distribute, rather than what may be charged, belongs in regulatory.
- Would Change Our Mind
The court setting the permitted external-link rate at 0% and Apple implementing it without a further appeal.
Agentic production collapses the cost curve. Engine-embedded agentic assistants move AI from assets into systems and content generation, cutting production time for a mid-budget title by a large factor and inverting the discovery problem from too few games to catastrophically too many. Platform commission stops being the binding cost because production cost falls faster than the toll does.
Mechanism
The substitute is engine-level agentic tooling rather than standalone AI products, which reaches every user of the engine by default. Two independent parties are already visible: Unity shipped an official agentic coding plugin with 29 engine skills on 2026-09-10, and Valve exempted code and game logic from Steam's AI-disclosure requirement in January 2026, removing the main friction to shipping it. The incumbent asset that loses value is the production-capacity advantage of large studios, and the new binding constraint becomes storefront discovery.
Preconditions, early indicators and assumptions
- Early Indicators
- Indicator
A step-change in Steam annual release counts well above the current trajectory, combined with AI disclosure rates passing roughly 50% of releases
- Source
S-15-09 Steam AI disclosure fields and store metadata
- Would Be Visible By
2028-06
- Indicator
Unreal or a second major engine shipping a first-party agentic assistant with comparable engine-level integration
- Source
S-15-17 Game Developer; S-15-18 GameDev Reports
- Would Be Visible By
2027-12
- Indicator
A shipped commercially significant title publicly credited to a materially reduced team size
- Source
S-15-17 Game Developer
- Would Be Visible By
2028-06
- Affected Industries
- 15
- 01
- 24
- 17
- What Businesses Should Do
Invest in storefront positioning, wishlists and community rather than in production capacity, because discovery becomes the scarce asset; treat AI disclosure policy as a marketing decision with measurable player-sentiment cost, not only a compliance one.
- Precedence Note
Requires disclosure rates above roughly 50% together with a release-count step-change. Disclosure plateauing in the high twenties, which is the current trajectory, leaves the outcome in base.
- Would Change Our Mind
Steam AI disclosure rates flat or falling across two consecutive Next Fest events while release counts stay on trend.
Child safety becomes the binding constraint rather than commission. Age assurance generalises from Roblox to every youth-facing platform through the UK Online Safety Act, the EU Digital Services Act, Australian rules and US state statutes; identity verification becomes a precondition of play and youth monetisation is structurally impaired. Direction is two-sided, because a harmonised standard could as easily reduce per-jurisdiction compliance cost as raise it.
Mechanism
The instruments are named and dated: California AB 1709, enacted 2026-09-11, restricting engagement-driving features for under-16s; the UK Online Safety Act and the EU DSA; and Roblox's own architecture, mandatory in selected countries from December 2025 and globally from January 2026, with account tiers for under-9s, 9-15s and 16-plus and age assurance required to publish an experience. Roblox has quantified the cost that no other youth platform has priced publicly — age-check penetration at 57% alongside a 14-18% guided bookings decline — which makes it the reference case regulators and competitors both use.
Preconditions, early indicators and assumptions
- Early Indicators
- Indicator
A second major platform — Fortnite, Minecraft or a large mobile publisher — adopting mandatory age assurance at Roblox's level of rigour and disclosing a bookings impact
- Source
S-15-16 PocketGamer.biz; S-15-04 SEC EDGAR
- Would Be Visible By
2027-12
- Indicator
A further US state or national statute mandating age verification for game platforms
- Source
S-15-20 EU DMA register and national legislative trackers
- Would Be Visible By
2027-12
- Indicator
A platform ban or reinstatement in a further national market, following Kuwait, Indonesia, Bahrain and Russia's lifting on 2026-06-10
- Source
S-15-16 PocketGamer.biz
- Would Be Visible By
2027-12
- Affected Industries
- 15
- 17
- 22
- What Businesses Should Do
Price age assurance as a revenue variable with a measurable cost, not as a compliance line item, and build it once against the strictest jurisdiction; assume identity-gated architecture arrives before the statute that mandates it.
- Precedence Note
Regulatory outranks base when a statute or a second platform's mandatory adoption lands, even if commission rates and market growth look base-like.
- Would Change Our Mind
Roblox reporting bookings recovery to growth with age-check penetration above 80%, which would show the architecture is not structurally impairing youth monetisation.
The tentpole model breaks in public. A generational release slips again or disappoints, console hardware economics deteriorate further, one or more platform holders retreat from first-party development, and the sector consolidates into live-service annuities plus a very long tail with nothing in between.
Mechanism
The thesis that fails is that a premium single-player tentpole can still carry a sector aggregate. Take-Two's FY2027 guidance of $8.0-8.2bn and a large part of the console growth forecast rest substantially on one release on 2026-11-19; recurrent consumer spending is already 84% of Take-Two's net bookings. The holders of the impaired asset are the publishers and platform holders carrying mid-budget production capacity, and the historical analogue is the 2005-2010 plastic-instrument music-game cycle, which went from zero to $1.7bn annually and back to near-zero in five years, taking a category of studios with it; the surviving structure there was live-service, which is the same answer this branch reaches.
Preconditions, early indicators and assumptions
- Early Indicators
- Indicator
No Western publisher other than Take-Two greenlighting a new $200m-plus single-player premium title during 2027
- Source
S-15-17 Game Developer; S-15-16 PocketGamer.biz
- Would Be Visible By
2028-01
- Indicator
Take-Two FY2028 net bookings guidance below roughly $6bn
- Source
S-15-04 SEC EDGAR filings
- Would Be Visible By
2027-06
- Indicator
A platform holder closing or divesting first-party studios beyond the 2026 pattern of Bluepoint and TiMi Montreal
- Source
S-15-08 videogamelayoffs tracker; S-15-17 Game Developer
- Would Be Visible By
2028-06
- Affected Industries
- 15
- 24
- 03
- What Businesses Should Do
Assume the middle of the market is not a viable position without a live-service tail, and structure any premium project with a recurrent-spending component from the design stage rather than bolting one on.
- Precedence Note
Failure outranks downside on occurrence. Downside is commission relief being clawed back with the release model intact; failure is the premium tentpole ceasing to be greenlit.
- Would Change Our Mind
Two or more Western publishers greenlighting new $200m-plus single-player premium titles during 2027.
Additional scenario notes
- Disclosure Note
Electronic Arts completed its take-private by PIF, Silver Lake and Affinity Partners on 2026-08-04, removing a top-five publisher's quarterly reporting from the resolution set. The observable universe for this sector is shrinking, and indicators below are chosen from sources that do not depend on EA's disclosure.
- Probabilities Sum
1
- What Must Be True To Grow
- The statutory commission regimes — the EU DMA and Japan's Mobile Software Competition Act — hold, since the corpus records these as durable while the US judicial remedies are contingent
- Developers actually use the link-out and alternative-distribution routes, which only Apple and Google currently know
- Per-jurisdiction compliance cost does not consume the saved commission at small scale
- Consumer spend holds against rising hardware prices driven by memory costs and tariffs
- Age assurance does not generalise in a form that removes youth monetisation across the category
- What Could Stop It
- Regulatory reversal: a US court restoring a commission on external links, which is pending now
- Capital cost: the EA leveraged buyout servicing debt at a 3.50-3.75% policy rate, unlike the 2021 LBO vintage
- Physical constraint: memory pricing and tariffs raising console hardware costs, already visible in Sony's 2026-04-02 increase
- Political licence: child-safety statutes and national platform bans arriving before the economics deteriorate
- Measurement revision: app-store estimate providers lose visibility precisely as alternative distribution grows, so mobile spend becomes less measurable exactly where it is changing fastest
- Uncertain Assumptions
- Text
Alternative distribution and link-out routes are being used at material volume, which no participant discloses
- Confidence
low
- Load Bearing
false
- If Wrong
The commission changes are nominal rather than economic, and every branch collapses toward base with a lower effective relief than assumed
- Text
Sensor Tower and Appfigures estimates remain a usable proxy for mobile spend as alternative distribution grows
- Confidence
low
- Load Bearing
false
- If Wrong
The sector loses its main third-party measurement just as the trend being measured accelerates, and the upside branch's supply indicators become the only observable ones
- Text
Platform commission is the sector's largest single cost line, so a durable reduction changes what gets made
- Confidence
medium
- Load Bearing
true
- If Wrong
Relief lands entirely in publisher margin, the upside branch cannot occur by its own mechanism, and its probability mass moves to base
- Text
Age-verified older cohorts monetise differently from the younger ones displaced, in a direction nobody can currently predict
- Confidence
low
- Load Bearing
false
- If Wrong
The regulatory branch's cost estimate is wrong in an unknown direction; this is a genuine natural experiment currently running at Roblox with no precedent
- Authored
2026-09-15