The peer-review system has developed a troubling reflex: the compulsion to find flaws where none exist. As one researcher recently shared, it has become rare to see a review that simply says, "I don't have major concerns." Instead, even papers that all reviewers accept are met with demands for five to ten additional experiments during the rebuttal phase. This is not rigorous review. It is performative criticism.
The intent behind cracking down on low-quality reviews was sound. The field needed higher standards, and it got them. But the pendulum has swung too far. Reviewers now feel obligated to find something wrong, anything wrong, to prove they are doing their job. The result is a flood of "what if" experiments: try a different backbone, a different dataset, a different setting. These requests often serve no real purpose. They are not designed to test core claims. They are designed to produce a "gotcha" moment when something fails under time pressure. The author who cannot run five extra experiments in two weeks is penalized for being insufficient, even when the paper already meets the bar.
This dynamic hurts everyone. Authors waste precious rebuttal time chasing irrelevant experiments instead of clarifying their contributions. Reviewers who push back against unnecessary requests have to step in during discussions, as the original poster described doing themselves. Even the area chairs often side with the reviewer who says, "I don't think that experiment is important." But the damage is already done: the process has become adversarial by design, not by necessity. The standard should be simple: experiments must be sufficient to support the core claims. Not exhaustive. Not a checklist of every barely applicable case. Sufficient.
What this means for researchers is straightforward. When you review, ask yourself whether a requested experiment would actually change your rating. If the answer is no, say so. Write it plainly: "The paper passes the bar, and I have curiosity questions that do not affect my rating." That sentence is not a sign of laziness. It is a sign of clarity. The field does not need more gotchas. It needs reviewers who can distinguish between a genuine weakness and a curiosity that can wait for future work. That is how you protect research quality, not by demanding more experiments, but by demanding the right ones.