How to Argue Better: Catch Yourself Sliding Past Points

If you want to argue better, start with one skill: notice when you slide past a point instead of answering it. When someone challenges a premise you can't defend, the natural move is to restate your conclusion in new words and keep going. It feels like responding. It isn't. Until you can catch yourself doing it, more debate tactics won't help, because you'll deploy them in service of avoiding the hard point.

I didn't know I did this. An AI debate showed me. Here's the transcript.

The debate that exposed the pattern

Topic: "Affirmative action in college admissions does more harm than good."

I argued against affirmative action, a position I don't hold, to practice. The AI defended it. Lightly edited:

Me: Affirmative action harms the students it's supposed to help. Research shows students admitted with lower qualifications than their peers are more likely to struggle academically and switch out of difficult majors. This is "mismatch theory": place students where they can't compete and they underperform.

AI: The research you're citing focuses on law school bar passage rates, not undergraduate outcomes. More importantly, the mismatch framing assumes admissions credentials predict outcomes. Students from underrepresented backgrounds face barriers that depress their credentials without depressing their ability. If admissions criteria underestimate ability for some groups, "mismatch" measures the gap between biased inputs and actual performance, which suggests the inputs are wrong, not that the students are overmatched. What makes you confident that SAT scores and GPA at a well-funded high school measure the same thing as at an under-resourced one?

I felt two things at once. First: that's a good point about measurement. Second: I really don't want to engage with it. The second feeling is the one worth examining.

What sliding looks like

Me: The measurement issue is a separate argument. My point is about outcomes. Students admitted with lower credentials, however you measure them, struggle more. Higher dropout rates, lower graduation rates, major-switching. That's harm, regardless of how we got here.

AI: You've shifted the argument. You started with "affirmative action admits students who can't compete." I responded that "can't compete" may be a measurement artifact. You've now moved to "students admitted with lower credentials struggle more," which is a different claim. Students could struggle because of mismatch, stereotype threat, or financial pressure. If the cause isn't mismatch, ending affirmative action doesn't end the struggle. It just relocates it.

That's the moment I saw it. The AI challenged my framing, and instead of engaging with the measurement question, I restated my conclusion in different terms. Technically a response. Actually an evasion.

This isn't lying. It's not even conscious. A point threatens your position, so you keep talking without acknowledging it. Between humans it works, because calling it out feels rude. The AI had no social constraints. It just labeled the move.

What engaging actually looks like

Me: You're right. I moved away from the measurement question because I don't have a good answer. If SAT scores measure something different depending on school context, then admissions becomes arbitrary. There has to be some metric that's at least somewhat comparable, or the whole system falls apart.

AI: Here's the tension: if current metrics are valid, affirmative action admits less-qualified students and mismatch may be real. If current metrics are biased, affirmative action corrects for measurement error, which is what admissions should do. You can't have it both ways. Which is it?

Me: I don't know. Probably some of both, depending on the metric and the context.

AI: That's an honest answer with implications. If metrics are somewhat valid but somewhat biased, the right policy is a calibrated middle ground. Does your original claim hold up if the answer is "it depends on implementation"?

I ended the debate without a settled view on affirmative action. But I'd learned how I avoid arguing while appearing to argue. The same move shows up whenever the topic has stakes, whether it's standardized testing or whether college is worth it. Point lands, speaker slides.

How to catch yourself sliding

Three things worked for me:

Watch for the flinch. The tell isn't in your words, it's in the feeling right before them: "I don't want to engage with that." That feeling means the point landed. Whatever you say next is probably a slide.

Restate their point before answering. If you can't state the challenge in a form they'd accept, you're not responding to it. This is slower, and it's supposed to be.

Say "I don't have a good answer" out loud. It feels like losing. It's the only move that keeps the argument honest, and it's what turned my exchange productive.

Since then I've noticed the pattern everywhere: work, relationships, decisions. It's also why you lose arguments you should win: you spend your credibility dodging instead of engaging.

The fix starts with an opponent willing to say "you didn't respond to what I said." Most people won't. That's the case for practicing against something that will, and why your strongest beliefs deserve the hardest testing: they're the ones you've been sliding around the longest.

FAQ

What does "shifting the argument" mean?

Restating your conclusion in new terms instead of answering a challenge to your premise. You swap in a nearby claim that sounds similar. It's usually unconscious, which is what makes it hard to catch.

How do I practice arguing without a debate partner?

Argue a position you don't hold against an AI opponent and pay attention to your own moves instead of trying to win. An AI will name the move; most humans are too polite to.