Batch cycle-002. Model 0.1.0, graph 0.2.0, parameters 0.1.0. 30 simulations, seeds 1 to 30; probes on 10. Every figure is a distribution over simulations of a structural model with seed parameter values; it shows what the model implies under these settings, not what will happen. Other nations are frozen at their 2026 ratings. No reconstructed historical data is used in this experiment.
Headline
China played 30 rounds from 1996 under the scripted method, with scripted events, 30 times. Its capability index moved from 9.13 to a median of 5.49 (−3.63), with 90 per cent of simulations between 5.46 and 5.53 [experiment_objectives].
Starting position
In 1996 the nation ranks 55 on the index against the frozen field, holds 0 of nine domains at Tier 1 and has a tier rank of 36. Its base income is 240 credits a year, calibrated so that its 2026 position is affordable with headroom for investment.
What was decided
A simulation spent a median of 490 credits on moves. The five groups that took most were Transport Infrastructure (build, 107 credits, 21.7 per cent); Energy Infrastructure (build, 94 credits, 19.2 per cent); Trade Policy & Market Access (build, 65 credits, 13.4 per cent); Industrial & Sectoral Base (build, 52 credits, 10.5 per cent); Research Talent & Knowledge Creation (build, 29 credits, 5.9 per cent) [experiment_decisions_by_group]. By mode: build 79 per cent, partner 16 per cent, buy 4 per cent [experiment_decisions_by_group]. Concentration of spending (HHI) had a median of 0.136, and the most repeated action appeared in 37 per cent of rounds.
The leading domain by period: rounds 1 to 5, Strategic Infrastructure; rounds 6 to 10, Trade & Investment; rounds 11 to 15, Strategic Infrastructure; rounds 16 to 20, Strategic Infrastructure [experiment_spend_by_period].
What the world did
A median simulation saw 8 shocks and 1 upside events. The most damaging per firing were S10 (60 firings in 30 simulations, −40.5 points each); S07 (60 firings in 30 simulations, −39.9 points each); S01 (30 firings in 30 simulations, −33.6 points each) [experiment_shocks]. Upside events fired 30 times across all simulations [experiment_shocks].
The groups that lost most to shocks, per simulation: Fiscal Capacity & Public Finance −7.49; Trade Facilitation & Logistics −7.48; Financial System Depth & Access −6.25; Labour & Employment −5.76 [experiment_shock_damage_by_group].
The edges that carried most change through the graph, per simulation: Fiscal Capacity & Public Finance to Force Generation & Mobilisation (−1.48 points); Fiscal Capacity & Public Finance to Social Infrastructure (−1.40 points); Fiscal Capacity & Public Finance to Transport Infrastructure (−1.00 points) [experiment_propagation_by_edge].
Where it ended
At 2026 the median rank is 111 (5th to 95th percentile 110 to 111), the median count of Tier 1 domains 0 (0 to 0) and the median tier rank 56; the resilient score, the mean score sum over simulations, is 407 [experiment_objectives].
Why
Without shocks the index ends +4.647 higher on average, so the shocks cost about 4.647 index points [experiment_probe_attribution].
Holding instead of following this method changes the index by −0.181: that is the method's worth against doing nothing [experiment_probe_attribution].
Removing any single one of the five largest actions changes the index by between +0.003 and +0.007, for example 1997 build Transport Infrastructure (12 credits) [experiment_probe_attribution]. The largest actions by credits come late, when income is highest, and are rarely the most valuable.
Removing the largest shock of a simulation raises the index by +1.231 on average [experiment_probe_attribution].
The replay probe reproduces the simulations exactly (index change 0.000000), so the attributions are clean [experiment_probe_attribution].
Surprises
tier_jump (warn): largest gain in one domain within five years: 3 tiers (30 simulations above 2).
retired_words (warn): e4-chn/report.md: 'simulations' 1 time(s), first near line 38; use 'simulations'.
The simulation furthest from the median (seed 17) ends +0.04 from it, with 8 shocks and 1 upside events [experiment_outlier_simulations].
Comparison
This is the baseline (hold) experiment, or there is none for this nation in the batch.
Against history: from a reconstructed index of 9.13 the model ends at 5.50, where the 2026 ratings put it at 15.64. Group scores miss by 10.18 points on average, against 6.77 for assuming nothing changed (skill −0.50); the change in each domain has the right sign in 0 per cent of domains. Furthest below history: critical technology (−14.67); furthest above: governance and integrity (−4.59) [experiment_against_history]. The starting position is reconstructed, so part of the error is the reconstruction's.