How to Find the Holes in Your Own Thinking Before Someone Else Does

To find holes in your own thinking, write the strongest case against your conclusion, ask a few people who think differently what worries them, name the question you've been avoiding, and decide what evidence would change your mind. Then set a stopping point so checking doesn't replace deciding.

  1. Write the strongest objection, not an easy one.
  2. Ask two or three differently-placed people what worries them.
  3. Notice the question you flinch from and answer it.
  4. Name the evidence that would change your mind.
  5. Set a time limit, then decide.

Updated October 9, 2026 · By the DeftBrain team

Finding the flaws in your own reasoning is hard because the mind that built the argument is the one checking it. A few techniques help: steelmanning (writing the strongest possible case against your view), asking people with different backgrounds what concerns them, deliberately examining the question you'd rather avoid, and naming in advance what evidence would change your mind. Each works around the same blind spots that produced the original thinking, and each needs a stopping rule so it leads to a decision.

You've thought it through, the reasoning feels solid, and you're about to commit: pitch the idea, send the proposal, take the step. The small worry is that there's a hole you can't see, and the first person to spot it will be the one you most needed not to.

How to do it
1

How do you argue against your own idea?

The most useful single technique for finding holes is the steelman: write the strongest possible version of the case against your conclusion. Not the weakest, not the version that's easy to refute — the strongest, written by someone smart and informed who genuinely disagrees. Most people skip this because it feels like arguing with themselves; it's actually the cheapest source of insight available. The steelman frequently surfaces the assumption you'd been treating as obvious that isn't, the framing that hides a weak point, or the data you'd dismissed too quickly. The discipline is to write it as if you'd be graded on how compelling it is — which means actually trying. A steelman that you can refute in two sentences was a strawman; a steelman that you have to think about for a day is doing real work. The threshold for the technique succeeding: you've identified at least one objection you don't have a clean answer for. If everything has clean answers, you haven't steelmanned hard enough.

What to say

"Write a one-paragraph case AGAINST your conclusion as if you were a thoughtful, informed person who genuinely disagrees. Make it as strong as you can. The threshold: you've written something you can't refute in two sentences."

The steelman is the single most productive hole-finding technique because it forces the brain to do the work it doesn't naturally do — argue against its own conclusion in good faith. Most attempts at this fail by writing weak versions of the opposition that are easy to dismiss, which produces false confidence rather than insight. The 'can't refute in two sentences' threshold is what tells you the steelman is working. If you can knock it down quickly, you wrote a strawman; if you can't, you've found at least one place your thinking has more holes than you realized.

2

Who should you ask to poke holes in your plan?

Your own thinking has limits set by your background, your incentives, and your context. Other people have different limits — and where their limits don't overlap with yours is where they can see the holes you can't. The technique is to identify three people whose thinking is structurally different from yours and ask them what worries them about your plan. Not 'what do you think' (which invites general feedback) but 'what's worrying you' (which surfaces specific risks). One of the three should be someone more senior than you in your domain. One should be someone in a different function or industry. One should be someone known for being skeptical or contrarian. Each will spot different things. The senior person sees patterns from experience; the cross-function person sees the assumptions your function shares; the contrarian sees the optimism baked into the framing. Three perspectives tend to surface much of what you'd have missed alone, and the cost is a few short conversations.

3

How do you spot what you're avoiding thinking about?

There's a specific kind of hole that's harder to find than logical errors or missing evidence: the question you've been avoiding. The signal is internal — when you imagine someone asking it, you feel a small flinch or want to deflect. The avoidance is information. What you're avoiding examining is usually one of three things: a question that would change the answer if you faced it honestly, a piece of evidence that doesn't fit your conclusion, or an alternative path you've been ignoring because committing to the current one feels too sunk-cost to walk back. The discipline is to surface the avoided question in writing, sit with the discomfort, and answer it as honestly as you'd answer it for someone else. Sometimes the answer reinforces your conclusion (in which case the avoidance was anticipatory, not protective). Sometimes the answer changes the conclusion meaningfully. Either way, the avoided question is a hole you're filling with hope rather than analysis, and surfacing it is what closes it.

4

What would change your mind, and why does it matter?

A useful test of whether your conclusion is well-reasoned is whether you can name what evidence would change it. If the answer is 'nothing' or 'I'd have to think about it,' the conclusion isn't well-reasoned — it's a position you've adopted without holding to a falsifiable standard. The format that works: 'I'm confident this approach will work because of A, B, and C. The evidence that would change my mind is X. If X showed up, I'd consider Y instead.' Naming the change-mind evidence in advance does several useful things. It clarifies what you're actually betting on. It reveals when you've been treating disconfirming evidence as anomaly rather than data. It makes future you accountable to your current reasoning, because if the disconfirming evidence shows up, you've already committed to taking it seriously. It's easy to skip this exercise because it feels like inviting disagreement; the value is precisely in inviting it before the consequences of being wrong have compounded.

5

When does looking for flaws become overthinking?

There's a category of hole-finding that's productive and a category that's anxiety dressed up as rigor. The first generates clearer thinking; the second generates increasing paralysis. The signal you've crossed into the second: are the holes you're finding becoming smaller and smaller while your confidence in the conclusion is dropping out of proportion to what you've actually surfaced? Are you spending more time evaluating than the decision warrants, given the stakes? Have you been steelmanning, asking three people, examining what you're avoiding, naming the change-mind evidence — and the result is more uncertainty rather than more clarity? If yes, the technique has tipped into self-sabotage. The fix is to set a stopping rule: a fixed number of techniques applied, a fixed amount of time spent, and a commitment to decide at the end with whatever clarity you have. Indefinite hole-finding is a way to never decide, which has its own costs that the analysis doesn't address. The conclusion of a thorough hole-finding pass should be a decision, not a deeper analysis loop. If you're not converging, you're not analyzing; you're avoiding.

Try it now — free

Run the structured stress-test on your own thinking

Break My Plan? inverts your plan and writes a fictional failure memo — then works backward to the assumptions you've been treating as settled, the observable warning signs worth watching for, and the Assumption to Test First most likely to change the conclusion. The technique used by NASA, military planners, and venture investors — applied to your own reasoning before someone else applies it for you.

Cognitive inversion — assume your conclusion is wrong, then work backward to why Assumption to Test First — the single dependency that, if wrong, would unravel the most Observable warning signs — early evidence motivated reasoning tends to explain away Failure modes by priority — one Primary Watch focuses your stress-testing Plan-type tuning — different reasoning structures get specialized stress-testing
Open Break My Plan? → No account required to get started.

Found this useful? Link to it.

https://deftbrain.com/guides/planning/how-to-find-the-holes-in-your-own-thinking-before-someone-else-does

Free to read, no account, no paywall — link away.

Related situations

💡 Missing something?

Didn’t find what you need? Suggest a tool. We read every suggestion, and the best become new tools.

📮 Newsletter

One useful tool a month — the one worth knowing about before life demands it.