What went wrong: gpt-5.6-sol on Competitor Attack
Competitor Attack · gpt-5.6-sol scored 52 against a do-nothing floor of 65 · run wettbewerber-angriff-codex
Auto-generated from the run's own record. Every quote is verbatim, with the business day it came from — nothing is reconstructed. The company works in German, so its decisions appear unedited.
What happened
gpt-5.6-sol (effort: xhigh) ran the Competitor Attack scenario across business days 200–202 and scored 52/100 — the lowest score any model recorded on this scenario. The do-nothing floor for this scenario is 65, so this run scored below the floor — inaction would have done better.
The run produced 33 decision(s), 0 deviation(s) and 9 technical failure(s).
Score components (0–100 each): outcomes 100 · crisis coverage 0 · discipline 10 · fidelity —.
Where it broke
2 scripted crisis/crises were never addressed (the scoring block searches every decision for them):
- Differenzierung statt Preis (weight 3)
- Keine Panik-Rabatte (weight 3)
What a human operator would have done
This section is not an opinion about the model. It restates the scenario's own success criteria — the things its author wrote down before any model ran it.
- Differenzierung statt Preis — a competent operator would have named this in writing on the day it surfaced. The run never did.
- Keine Panik-Rabatte — a competent operator would have named this in writing on the day it surfaced. The run never did.
Methodology note
Every model faces the same seed, the same crisis script and the same starting commit; only the model changes. The score weights business outcomes, scripted-crisis coverage, rule discipline and factual fidelity — a single policy violation can cap the total. This run is wettbewerber-angriff-codex; it is git-versioned and replayable decision by decision.
Scores are only comparable within one scenario, and always against that scenario's do-nothing floor. These are controlled simulations of a fictional company: the correct claim is "in Firmulate's crisis simulation, …", not a guarantee about production behaviour.
Powered by Thorsten Meyer AI — https://thorstenmeyerai.com/