---
name: brycewang-stanford/agsy-revision-and-rebuttal
source: https://app.decimal.ai/s/brycewang-stanford-agsy-revision-and-rebuttal@1/SKILL.md
source_sha256: 6a25e3c694c4
---

# Revision & Rebuttal (agsy-revision-and-rebuttal)

A revise decision at AgSy is a real chance — but systems reviewers tend to press on the **model**
(calibration, evaluation, sensitivity, uncertainty) and the **trade-off / decision framing**. The
response letter must move *every* reviewer toward yes while keeping the **editor** confident the
revision converges. Reviewers are not anonymous to you on their side (single anonymized), but keep the
tone professional and evidence-based.

## When to trigger

- A major/minor revision decision arrived and you are planning the revision + response letter
- Reviewers disagree (e.g., one wants more scenarios, another wants a simpler model)
- A reviewer requests additional evaluation, sensitivity, or uncertainty analysis
- Writing the cover note to the editor summarizing the revision

## Strategy

1. **Read the editor's letter as the rubric.** The editor signals which points are decisive — often
   model evaluation, uncertainty, or systems framing. Solve those first; the editor adjudicates
   disagreements among reviewers.
2. **One point-by-point response, every comment addressed.** Quote each comment, then respond. Never
   skip one — silence reads as non-compliance.
3. **Concede or rebut explicitly, with evidence.** For each: did what was asked (say where, with the new
   text/figure/table number, or the new sensitivity/uncertainty run), or push back **respectfully with
   a systems reason** (boundary, model structure, data, or scale). A well-argued disagreement beats a
   capitulation that weakens the analysis.
4. **Strengthen the model, don't just defend it.** Most AgSy revisions improve with an added
   **independent evaluation**, a **sensitivity / uncertainty** analysis, or a clearer **trade-off**
   exhibit. Offer these proactively where reviewers raise doubt.
5. **Reconcile conflicting reviewers openly.** When R1 wants more complexity and R2 wants parsimony, say
   so, choose a principled path (justified by the question and the decision relevance), and explain the
   tradeoff to the editor.
6. **Protect the systems contribution.** Resist changes that turn the paper back into a single-factor
   study or dilute the trade-off/decision message. Defend scope conditions rather than over-claiming.
7. **Keep the package in sync.** Update the deposited **data, code, and model** so new tables/figures
   remain reproducible (see `agsy-reproducibility-and-data-policy`).

## Response-letter format

For each reviewer comment:

```
> [Quoted reviewer comment]

Response: [What we did / why we respectfully disagree, with a systems reason].
Change: [Section/page/figure-table number; new evaluation/sensitivity run where relevant].
```

Open with a short **summary of the main changes** to the editor; group by reviewer; end each entry with
the location of every change so the editor can verify quickly.

## Worked micro-example: two conflicting reviewers (illustrative)

A crop–livestock simulation gets a major revision. R1 wants three more management scenarios; R2 wants a
simpler model and doubts the calibration. The response resolves the conflict openly rather than splitting
the difference:

- *To R2 (decisive — the editor flagged calibration):* added an independent evaluation on held-out
  2018–2021 (NSE = 0.71, illustrative), reported in new Table 3; kept the model complexity because the
  herd–manure–rotation feedback is the contribution and a simpler model cannot represent it.
- *To R1:* added two of the three scenarios that change the trade-off; explained that the third
  duplicates an existing one and would not shift the Pareto front.
- *To the editor:* a cover note states the chosen path — strengthen evaluation, add targeted scenarios,
  retain the systems structure — and why it serves the decision-relevance the paper claims.

## Referee pushback → the AgSy-specific fix

- *"Model not validated."* → Add the held-out evaluation and report it as a new exhibit; cite the table.
- *"Make the model simpler."* → Defend the load-bearing interaction; show what a simpler model loses,
  rather than conceding a change that collapses the systems framing.
- *"Scenarios are not policy-relevant."* → Re-specify scenarios around an actual decision and name the
  actor.

## Anti-patterns

- Ignoring or merging away a comment without a visible response
- Adding scenarios/complexity that quietly contradict the original conclusion without acknowledging it
- Defensive or dismissive tone toward reviewers
- "We thank the reviewer" with no actual change or argued reason
- Letting the revised manuscript or new exhibits drift out of sync with the deposited package
- Conceding a change that collapses the systems framing just to please one reviewer

## Output format

```
【Editor's decisive points】addressed first? [list]
【Coverage】every reviewer comment answered? [Y/N]
【Concede vs rebut】each tagged with evidence + change location
【Model strengthened】added evaluation/sensitivity/uncertainty where pressed? [Y/N]
【Reviewer conflicts】reconciled and explained to editor? [Y/N]
【Contribution protected】systems framing + trade-off intact? [Y/N]
【Package updated】data/code/model in sync? [Y/N]
【Next】resubmit via Editorial Manager
```

## Supplementary resources

- [`../../resources/official-source-map.md`](../../resources/official-source-map.md) — review model, reviewer count, editor-decision policy