Your Standard Should Never Be Paper Acceptance Borderline

Chapter 3 Episode 1

Getting a paper accepted at a top-tier conference like ACL or CHI is not as hard as it might seem. If we set aside the randomness of reviewers for a moment, clearing the acceptance borderline is, for the most part, an achievable goal for anyone willing to prepare thoroughly. Why? You may ask, how is that possible?

I will start with a concept from philosophy of science that is genuinely useful for understanding what we are all doing day-to-day.

Thomas Kuhn, a prominent philosopher who popularized the term “paradigm shift” and substantially influenced the philosophy of science, observed that the vast majority of scientific work is what he called “normal science”: researchers theorizing, observing, and experimenting within an already-established paradigm. In his influential work The Structure of Scientific Revolutions (I would highly recommend reading it), Kuhn described normal science as work that fills in the details, extends known findings to new domains, resolves ambiguities, and strengthens the evidence base within an existing framework, rather than overturning the paradigm itself. The paradigm-shifting, groundbreaking work that most people imagine when they hear “scientific discovery” is exceedingly rare, and it typically occurs only after a long accumulation of normal science has revealed anomalies that the existing paradigm cannot explain.

Why does this matter for you? The majority of research publications, including those at the most prestigious venues, are contributions to normal science. They are incremental by design (and I do not mean that as a criticism; it is how scientific knowledge actually accumulates). If most work is meant to fill in details rather than rewrite the rules, then the metric for whether a paper deserves to be published obviously cannot be “how groundbreaking is this idea?” Otherwise, almost no paper would ever be accepted.

The community needs a more objective and accessible way to assess quality, and that ultimately comes down to scientific rigor. The core of scientific rigor, when you strip it down, is avoiding critical mistakes, and I would categorize the requirements into four pillars.

First, you need a coherent motivation with no critical logical weak spots. Second, you need a study design without obvious threats to validity. Third, your findings should not merely repeat what other work has already reported, nor read as common sense (if anyone could have anticipated the result without your study, the contribution is questionable). And fourth, your claims should not overstate what the evidence actually supports.

Sounds very straightforward, and it should be. Avoiding critical mistakes across these four pillars is a basic skill that everyone can develop if they invest substantial effort in preparing the project. You do not need to make the work “impactful” or prepare it to be a paradigm-shifting contribution (and honestly, that is often not something you can engineer in advance anyway). You just need to make sure that the critical mistakes, the ones that can invalidate your entire project’s validity, are properly addressed.

To get there, you need to conduct a thorough literature review so you know what the field knows and what remains underexplored, iterate on the rationale until the logical chain holds up, carefully design the study to align with the research question, and frame the findings systematically. In the later chapters, I will teach you how to execute each of these. A surprising number of submissions fail on one or more of these fronts, not because the authors lack intelligence or effort, but because they did not invest sufficient care in the preparation. And that is precisely why the standard is achievable. It comes down to diligence and thoroughness, not to talent or luck.

So that is the borderline. For me and for my group, though, I always hold a standard much higher than just getting the paper accepted. Why?

Reading Your Own Work with Fresh Eyes

Avoiding critical mistakes keeps your work from being rejected, but it does not make your work genuinely strong. The four pillars are about preventing the fatal flaws that would invalidate a project. If a reviewer identifies such a flaw, they will stop reading and reject the paper outright. Strength is a different matter, and it lives in a gradient that the four pillars do not capture: how clearly a reader follows your argument, how readily they accept your claims, how few places there are where they pause, frown, and think “wait, why?” (Believe it or not, this could actually be the most frequent phrase I use in all meetings)

The method I rely on, and the one I ask my students to practice, is to read your own work as if you were a fresh reader who has never seen it before. After you have written a draft, you put aside everything you know about the project that is not on the page, and you read it the way someone encountering it for the first time would. Where would they get lost? Where would they misunderstand what you meant? Where would they question a claim that felt obvious to you because you have lived with the project for months? Every one of those moments is a place where the work can be strengthened before it ever reaches a reviewer.

The power of this method is that you do not need to submit your paper to anyone to find these problems. You become the test reader yourself. When I do this with my own drafts, I am often surprised by how much I find, because the gap between what I meant to say and what the page actually says is always larger than I expect. A sentence that felt perfectly clear when I wrote it turns out to assume three things the reader has no way of knowing. A transition that made sense in my head turns out to skip a step. None of these are critical mistakes in the four-pillar sense, but each one chips away at how convincing the work is.

I want to be honest about two limits of this method. First, you cannot fix everything. Every paper has limitations, including the ones being cited thousands of times, and there are always practical constraints that prevent you from producing a “perfect” piece of work. Rather than chasing perfection, you identify the issues you can address without overhauling the entire storyline, and there are usually more of those than you would think. Second, a fresh reader who knows nothing about your topic is excellent at catching problems of clarity and persuasion, but a naive reading will not catch every threat to validity, such as a subtle confound or a missing body of related work, because those require domain expertise rather than fresh eyes. Reading from different readers’ backgrounds and perspectives is a more advanced skill, and I will cover it in a later chapter. For now, the fresh-eyes reading is the single most useful habit you can build, and it will take you a long way.

Why Hold Yourself to This Standard

The deepest reason to read your own work this way connects directly to the shift from external to internal evaluation that I described in the previous post. When you practice fresh-eyes reading consistently, you are training yourself to evaluate your own work without waiting for someone else to do it for you. Over time, you internalize the reader’s perspective so thoroughly that you start catching the gaps as you write, not just when you revise. You are becoming your own first reader, and, trust me, the capacity of self-critiquing is one of the most valuable things a researcher can develop.

At the end of the day, the reward is a kind of confidence that no external feedback can give you. When you submit a paper that you have already read with fresh eyes, found the weak spots, and strengthened where you could, you submit it knowing it is solid. The review outcome still matters, but it no longer carries the same anxious weight, because your sense of the work’s quality is already in place before anyone else weighs in. The internal assurance you gain this way is the real reward of holding a high standard, and it is exactly the internal rewarding mechanism I talked about in the previous post.

There are accompanying benefits, too. A paper that has been strengthened this way is more resilient in review. A borderline paper is fragile, and one skeptical reviewer can sink it because there is no depth to absorb the challenge. A stronger paper gives a critical reviewer much less to grab onto. What most students do not expect is that the reviewer who has to engage seriously with your reasoning in order to critique it is often the one who comes to appreciate the contribution most. I have seen the harshest reviewer in the first round become the strongest advocate by the end, because the work earned their respect through its care. And beyond any single paper, producing consistently strong work over time opens doors to conversations and collaborations with people whose insight and judgment you can learn from, which is a reward I will say more about in a later post.

None of this requires more raw hours than aiming for the borderline. It requires a different habit of attention, one that treats your own draft as something to be tested rather than defended. If you build that habit early, it compounds, and within a few papers you will catch issues that would have slipped right past you at the start.