DRAFT · Updated 11 October 2026

Change log

The public change record for the Scenario Library and each Edition.

Library versions follow semantic versioning: patch for wording and anchors, minor for parameter or loading changes, major for scenario additions, retirements or scale changes. Edition pages are frozen at release; corrections are published as errata, not silent edits.

Outcomes

Outcomes v1.0.0 · 10 October 2026

  • Outcomes become a body of work of their own, beside trends and scenarios, with a page and a download. An outcome is an isolation point: one named place in the capability framework crossing a threshold, whatever caused it.
  • Fifty outcomes replace the twelve the reverse stress test first ran on: twenty-five of failure and twenty-five of prosperity. Eighteen are set by rule, a fall and a rise for each of the nine capability domains. Thirty-two are named, proposed by the three-model panel in two rounds and settled in a final cut.
  • Every capability group now sits in an outcome; the named outcomes alone reach 72 of 73.
  • The reverse stress test is rerun on all fifty. Its upside half now mirrors the load test: an upside scenario lifts the capabilities it loads, scaled by the nation’s capture rate. Every outcome has a route for some nations.
  • The reverse stress test pages move under Outcomes. Old addresses redirect.

Site

Trends move to a site of their own · 11 October 2026

  • The trends, the trend network and the attention data now have a site of their own, GINC Trends. This site is about the scenarios: the Library, the outcomes, the tests and the Edition.
  • Each scenario page still names its key trends, and links to them there. The addresses the trend pages and trend downloads had here redirect.
  • This site keeps a copy of the trend list and its scores against the scenarios, taken from GINC Trends, and nothing else of the trends.
  • Earlier entries about the trends are in the GINC Trends change log.

Scenarios by evolution · 10 October 2026

  • A third overview joins the views by lens and by region: every scenario through six stages, from lead indicators and triggers, through its immediate, post-immediate and stabilisation phases, to the long term. The cells are written from each scenario’s own record.
  • The overview is no longer a page of its own. The Scenarios page now has four views: the list, by lens, by region and by evolution. Old overview addresses redirect.
  • The edition page, the matrix page and the results page are merged into one Analysis page with two views: ranking and matrix, and results by nation. Old addresses redirect.
  • The menu is now Scenarios, Outcomes, Analysis, Methodology, Data.

Small screens · 10 October 2026

  • No page scrolls sideways on a phone. Wide tables and charts scroll inside their own box, say when there is more to the side, and keep the column that names each row in view.
  • Charts hold a readable text size on small screens. Tables in the text pages keep a readable column width.
  • Buttons, toggles, filters and the menu are larger on touch screens.

Second pass · 9 October 2026

The whole build was put through a second pass with a panel of three AI models, each working first separately and then with sight of the others, and a fourth making the final cut. The methodology sets out how, and the record is on the data page.

  • Trends. The trend list was revised in the same pass: 119 trends carried forward unchanged, 106 under a more exact name, 15 merged into another, 9 dropped and 24 added. GINC Trends lists every change.
  • Library v0.5.0. Twenty scenarios, ten downside and ten upside, replace the 24. Four pairs of downside scenarios were folded together, mass displacement was retired and a cryptographic trust break was added. The upside scenarios were restated as events with thresholds. Scenarios are renumbered S01 to S20, and the addresses of first-pass pages redirect.
  • Scores. Every trend rescored against every scenario by all three models; the median is kept. Ratings for each scenario are the median of the models that proposed it. No scenario now reaches Tier I.
  • Recalculated. The load, capture and reverse stress tests; the vulnerability and opportunity indices; the seven archetypes, two of them renamed.
  • Anchors. Historical anchors re-pointed at the new scenarios; anchors no longer cited were removed and new ones added. Figures quoted for anchors were not re-checked against sources in this pass.

Method

A standard backbone of variables · 8 October 2026

  • The methodology now sets out every variable a scenario is built from, each with the question it answers and its standard values, so that custom scenarios can be specified on the same footing.
  • Onset becomes onset pace, on one scale: days, weeks, months, years. All 24 records moved to it.
  • Likelihood type is shown as event reach: global event or single-nation event.
  • ‘Negotiated’ added as an origin, for upside scenarios that come from agreement.

Results

Reverse stress test · 8 October 2026

  • The reverse stress test is run and published: twelve outcomes, seven of failure and five of prosperity, for 197 nations.
  • Each result gives the likelihood of the outcome, the most likely route to it, headroom and the lever.
  • It replaces the break-point test in the methodology; headroom carries what that test was to report.
  • Nation-level results are public. The disclosure policy now keeps confidential only a reverse stress test commissioned by a nation on its own data.

Second run · 3 October 2026

  • Exposure now set nation by nation, from 25 indicators drawn from the 2025 World Factbook blended with each scenario’s regional exposure. The first run applied one exposure to every nation in a region.
  • Capture test extended to the five new upside scenarios; gain and loss exposure taken from indicators where one fits.
  • Two composite indices added: a vulnerability index (expected loss, likelihood-weighted) with worst case and tail beside it, and an opportunity index. Both are computed at three and ten years.
  • Further measures added for each nation: breadth, shock absorbed, policy dividend and weakest link, and a placing in one of four groups.
  • Seven archetypes added: nations grouped by clustering on their results in all 24 scenarios, each named and described.

First run · 3 October 2026

  • Load test run on the fourteen downside scenarios and capture test on the five upside scenarios, for 197 nations, against capability scores from the GINC National Capability Index 2026.
  • Results published by scenario and by nation, with settings and downloads. The run is illustrative: scores are candidate ratings and exposure is set by region.
  • The break-point test has not been run.

Library

v0.4.0 · 3 October 2026

  • Five further upside scenarios added, one in each family: emerging-market growth breakout (S20), Middle East regional settlement (S21), agricultural yield revolution (S22), digital public infrastructure leapfrog (S23), hundred-day pandemic defence (S24).
  • The Library now holds 24 scenarios: 14 downside and 10 upside. Every trend in the GINC 250 is scored against the five new scenarios.

v0.3.0 · 3 October 2026

  • Five upside scenarios added, one in each family: superintelligence takeoff (S15), energy abundance breakthrough (S16), biomedical breakthrough wave (S17), great-power détente and peace dividend (S18), coordinated sovereign debt reset (S19).
  • Every scenario now carries an orientation, downside or upside. Upside scenarios also state who gains and who loses.
  • Upside scenarios are rated on the same scales and are not ranked in Edition 2027.
  • GINC Trends list added: 183 trends (150 from the Barclays 2035 Thematic Roadmap, 33 added from the scenarios), each scored against every scenario on a seven-point rubric, with derived trend similarity, sensitivity and network measures.

v0.2.0 · 2 October 2026

  • Fourteen scenarios in five families entered with provisional GINC desk scores.
  • Added: global financial crisis (S03), hybrid campaign by a hostile state (S06), mass displacement shock (S08).
  • Widened: single-supplier cut-off (S02) from minerals and chips to any input with a supplier share above 40 per cent; catastrophic disaster in the core (S10) from climate to seismic hazards.
  • Reworked: political legitimacy crisis (S07) to include the internal-order dimension; critical infrastructure collapse (S12) reframed around power.
  • Split: food and water system shock (S11) from the climate scenario.
  • Recovery horizon split from duration as a tenth common slider.
  • Added an overview block to every scenario: six lenses (political, economic, social, technological, legal, environmental) and nine regions, as GINC desk summaries of the record.
  • Parked, not in the Library: extreme space weather; chemical, biological, radiological or nuclear incident.

Edition 2027

Development build · 2 October 2026

  • Provisional ranking by priority score with the published tie-break rules.
  • Judgement recorded: S04 takes the tenth place over S10 on the current debt-distress and aid-withdrawal context.
  • Judgement recorded: S13 is ranked seventh as drafted; the published arithmetic gives a priority of 12 rather than 16. Referred to the review committee.
  • Headline findings are placeholders.