Mirror a peer
Each round, closes the largest gaps to a peer nation's 2026 profile, by the cheapest open mode.
How it decides
The peer is the nation five places above on the mean score. Each round the method measures every group's gap to the peer's 2026 score and funds the largest gaps first, each by the cheapest mode still open.
It measures gaps against current scores and does not count levels still in delivery, so it can keep funding a gap already paid for; the improvement loop has raised this as a harness finding.
- goal_gap
- the largest remaining gap to the peer
What it is for. A plausible imitation strategy, to test whether copying a stronger profile beats valuing the graph.
Its experiments
Every experiment that used the method: the capability index at the position and the median at the horizon across its simulations.
| Experiment | Cycle | Simulations | Index |
|---|---|---|---|
| United Kingdom, mirror a peer, 20 years cycle-001.e3-gbr-mirror-peer | 1 | 100 | 16.45 → 17.33 (+0.87) |
| South Korea, mirror a peer, 20 years cycle-001.e3-kor-mirror-peer | 1 | 100 | 16.00 → 17.20 (+1.20) |
| Estonia, mirror a peer, 20 years cycle-001.e3-est-mirror-peer | 1 | 100 | 11.49 → 12.52 (+1.03) |
| Brazil, mirror a peer, 20 years cycle-001.e3-bra-mirror-peer | 1 | 100 | 11.69 → 13.04 (+1.35) |
| Kenya, mirror a peer, 20 years cycle-001.e3-ken-mirror-peer | 1 | 100 | 7.18 → 6.97 (−0.21) |