Research
The research behind decision governance
Decision governance, as Kettei uses the term, begins after approval. The published record on what happens to a decision after it is made is thin, and the sources below are the ones the essays rely on.
The comparison of methods
The essay 7 Systems Have a Decide. None Has a Lifecycle compares 7 established methods, which are OODA, PDSA, double-loop learning, the Flow System, Toyota Kata, Al Shalloway's FLEX and Amplio, and DecisionProtocol. Each one has a Decide. In the sources read for the essay, no method among the 7 follows the decision through its life.
That comparison is Kettei's reading, and a reader can check it against the same sources. Architecture decision records and the Federal Reserve's SR 11-7 guidance on model risk each hold parts of a lifecycle, for one class of decision. Across that reading, a validity boundary and a set of abort criteria did not appear together for an ordinary organisational commitment.
What the sources found, and how Kettei applies each
| Source | What the source found | How Kettei applies it |
|---|---|---|
| Schwaber and Sutherland (2020), The Scrum Guide | "A Sprint could be cancelled if the Sprint Goal becomes obsolete. Only the Product Owner has the authority to cancel the Sprint." | A precedent for one named person holding the authority to stop. In Kettei the Owner, and only the Owner, declares the abort, and others may propose abort criteria that the Owner must agree to. |
| Staw (1976) | In a laboratory case, students who had personally chosen a division and then saw it decline put more of a fresh allocation into it than any other group. Staw's explanation was self-justification. | Abort criteria are set when the decision is formed, before sunk cost builds up and before a reputation is attached to continuing. |
| Simonson and Staw (1992) | Participants who wrote down minimum targets straight after a first choice, before any results, gave the losing product a mean of US$3.9 million against a baseline of US$5.1 million. Nearly 40% of the new money still went to the product that had missed its targets. | A written stopping condition reduces over-commitment and leaves much of it in place. Kettei claims that smaller yield, and adds a named Owner who answers for the outcome. |
| Kristal and Dorison (2025), by the paper's abstract | Across 9 preregistered experiments, observers judged decision makers who de-escalated as more trustworthy when a precommitment had been made. The benefit weakened when the pledge named a vague action or no concrete condition, and precommitment undermined perceived confidence at a project's inception. | An abort criterion is written as binary and observable. The confidence cost argues for spending precommitment where a decision is hardest to undo, which is how reversibility sets the weight of governance. |
| Congressional Research Service (2016), on the Nunn-McCurdy Act | The report counts 24 critical cost breaches from 2007 to 2015 and states, "Generally, a Nunn-McCurdy breach does not result in a program being cancelled." Each breach required a root-cause analysis and a written certification by one official within 60 days. | A condition written in advance makes a conversation due on a clock, with a named person answering on the record. Kettei reads the certification as closer to a validity boundary trigger than to an abort. |
| US Food and Drug Administration (2006), guidance on data monitoring committees | The statistical boundaries planned for interim analyses are "useful guidelines". A committee "will usually recommend termination when these thresholds are crossed, but it is not obligated to do so", and "in almost all cases" the committee is advisory. | A boundary agreed in advance fixes when the conversation happens and what it is about. The verdict stays with people, which in Kettei means the Owner. |
| Duke (2018), Thinking in Bets | Resulting is the tendency to "equate the quality of a decision with the quality of its outcome". Hindsight bias is the companion error. | The Kettei Card records the rationale and the knowledge when the decision is made. The Reflect phase reads those declarations before it reads the result. |
What the research does not show
These sources did not test Kettei. Each shows a single mechanism or a single record, and the effect of the 7 properties together is untested in the research cited here. The experiments asked participants to decide hypothetical cases. The public records describe single programmes. Kettei's claim is narrower than a promise of better results, and the essays state it in those terms.
A decision governed to the letter can still be wrong. With governance in place, a decision has a named Owner, written conditions for ending it and someone with the authority to end it when those conditions are met.
Sources
- Schwaber, K. and Sutherland, J. (2020) The Scrum Guide: The Definitive Guide to Scrum: The Rules of the Game. November 2020.
- Staw, B. M. (1976) Knee-deep in the Big Muddy: A study of escalating commitment to a chosen course of action. Organizational Behavior and Human Performance 16(1), 27-44.
- Simonson, I. and Staw, B. M. (1992) Deescalation strategies: A comparison of techniques for reducing commitment to losing courses of action. Journal of Applied Psychology 77(4), 419-426. DOI 10.1037/0021-9010.77.4.419.
- Kristal, A. S. and Dorison, C. A. (2025) Precommitment can allow decision makers to maintain trust when de-escalating commitment. Journal of Applied Psychology 110(10), 1395-1411; online 12 December 2024. DOI 10.1037/apl0001243.
- Peters, H. M. and O'Connor, C. V. (2016) The Nunn-McCurdy Act: Background, Analysis, and Issues for Congress. Congressional Research Service report R41293, updated 12 May 2016.
- US Food and Drug Administration (2006) Guidance for Clinical Trial Sponsors: Establishment and Operation of Clinical Trial Data Monitoring Committees. March 2006.
- Duke, A. (2018) Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts. New York: Portfolio.
Read the originals
The comparison of methods was published on Kettei on Substack on 8 October 2026. The record on stopping rules is in The Honest Yield of 4 November 2026, and the experiments on escalation are in Escalation Is a Policy of 28 October 2026.
From the essay 7 Systems Have a Decide. None Has a Lifecycle on Kettei on Substack, 8 October 2026.

Kettei
The Discipline of Governed Intent
