Rubber Duck Narration vs Shared-Editor Timed Mocks
Engineers have explained code to rubber ducks for decades. Rubber duck narration for interview prep borrows the same idea: put a duck (or a mug, or a plant) on the desk and explain the problem, your plan, and every line of code out loud to it, as if it were the interviewer. The other habit is shared-editor timed mocks: a plain browser editor like many remote loops use, a timer, narration while typing, and a forced freeze-and-recover before you say “done.”
Neither is wrong. The useful question is narrower: when does rubber duck narration beat shared-editor timed mocks—and when does the reverse happen?
This is a workflow comparison, not a review of any editor or tool. Interview formats vary by company and change over time. What stays stable is the difference between building the habit of thinking out loud with no pressure and keeping that habit alive while the clock runs on the surface your interviewer will watch.
At a glance
| Category | Rubber duck narration | Shared-editor timed mocks (Luna side panel optional) |
|---|---|---|
| Session shape | Explain problem, plan, and code out loud to an object | Timer + plain editor + narration + forced recovery |
| Structure | Freeform, no clock unless you add one | One prompt, live rules, fixed timebox |
| Feedback | You hear your own gaps as you explain | Self-postmortem: plan, freeze, dry-run, typed bugs |
| Environment fidelity | Low pressure; any surface | High for remote coding rounds in a browser editor |
| Best for | Breaking the silent-coding habit, finding gaps in your reasoning | Typing under time, audible plans, mute recovery |
| Main tradeoff | No clock and no stakes, so it can stay comfortable | Fewer reps per hour; can feel stressful early on |
What rubber duck narration looks like
A typical duck session:
- You pick a problem you have solved or half-solved and put the duck where an interviewer’s face would be on a call.
- You restate the problem to the duck in plain words, including the inputs, outputs, and one edge case.
- You explain your approach before writing anything, then narrate each line as you write it: what it does and why.
- When you hit a sentence you cannot finish—“and then this works because…”—you stop. That is the gap.
- You fix the gap and explain it again. Closer to a remote loop, you still need timed, typed mocks—explaining calmly to a duck is not the same as explaining while a clock runs and someone watches your cursor.
Honest strengths. It is the lowest-friction way to make thinking out loud normal. Many candidates go silent in interviews simply because they have never narrated their code; a few weeks with a duck makes talking while coding feel natural. Explaining forces precision: you discover that you do not actually know why the loop bound is n - 1 the moment you try to say it. It works anywhere, costs nothing, and pairs with any problem source. It also turns review sessions into active recall instead of rereading.
Honest limits. There is no pressure. The duck never looks at the clock, never asks a follow-up, and never makes you nervous, so a comfortable duck session can overstate how ready you are. Without a timer, it is easy to explain beautifully for forty minutes on a problem you would need to finish in twenty-five. Narrating already-solved problems trains explanation, not the first two minutes on a fresh prompt. And many people drift into narrating only in their heads after a while.
When rubber duck narration wins: your bottleneck is that you code in total silence; you are early in prep and want to build the habit with zero stress; you need to find gaps in your own reasoning; or you are reviewing problems you have already solved.
What shared-editor timed mocks look like
Shared-editor mocks flip the polarity: same narration habit, interview conditions.
- Pick one fresh problem from a short pattern list tied to your loop.
- Open a plain browser editor—no linting, no autocomplete, no run button if your target loop will not have one. Timer on, often 25–35 minutes.
- Restate out loud. Type a signature and a two-line plan as comments before real code.
- Narrate while typing. Force one freeze—hands off the keyboard, three breaths—then restart: narrate, restate, smallest step, signature.
- Dry-run three cases against the typed code, out loud.
- One-sentence postmortem: pattern gap, freeze, or bug.
Optional companion: Luna Interview is a Chrome extension framed as a practice / calm side panel—not a substitute for knowing the material, and never an “undetectable” live cheat. Keep it beside the editor while you own the plan, the code, and the narration.
Honest strengths. Fidelity for remote loops. The timer adds the pressure that makes silence creep back, so you practice narrating when it is hard, not just when it is easy. Fresh prompts train the first two minutes. Typed code exposes bugs that a spoken explanation can gloss over. Forced freeze recovery rehearses the moment that sinks strong candidates.
Honest tradeoffs. Mocks are slower and more tiring—one or two per evening. If you have never narrated before, the timer can make you freeze harder at first, and you may drop the narration entirely to save time. Without a shortlist, you thrash on unrelated prompts.
When shared-editor timed mocks win: your interview is close; you already narrate comfortably in calm practice but go quiet under time; your interviews are remote and typed; or your bottleneck is a weak first two minutes on unseen problems.
Decision table for candidate situations
| Your situation | Prefer rubber duck narration | Prefer shared-editor timed mocks |
|---|---|---|
| You code in complete silence | Strong fit—build the habit first | After a week of duck sessions |
| You narrate fine until the clock starts | Light warm-up | Strong fit—timer + narration |
| Reviewing problems you already solved | Strong fit—explain each line | Optional |
| Loop booked in a week | Five-minute warm-up before each mock | Primary lane |
| Explanations feel vague even to you | Strong fit—find the unfinishable sentence | Add the dry-run rule |
| Freeze on fresh prompts at minute one | Light | Strong fit—forced freeze and restart drills |
| Remote, typed interview loop | Supporting habit | Strong fit—practice on that surface |
A fair practical method (try both in one week)
Do not pick by which one feels better. Build the habit with the duck, then test it under the clock.
Duck days (two short sessions): Pick three problems you have already solved from one pattern family—say, sliding window. Explain each to the duck from restating to the last line, no timer. Every time you cannot finish a sentence, write it down. Fix those gaps.
Shared-editor mock days (next three or four sessions): One fresh problem from that same family per session. Plain editor, 25-minute timer. Two-line plan as comments, narration while typing, a forced freeze at minute three, three dry-run cases out loud. Keep a calm practice side panel beside the editor if you use Luna. Score only: audible plan within two minutes, narration kept up past minute ten, dry-run completed before saying done.
Compare: did the duck habit survive the timer? Many candidates keep both—a few minutes of duck narration as a warm-up, shared-editor timed mocks for most prep minutes once a loop is near.
Common mistakes / when NOT to use each
Skip duck-as-progress when you only ever explain problems you already know, with no clock, for weeks. Comfortable narration with zero timed sessions is a common false summit. Also skip narrating in your head—if the duck cannot hear it, it does not count.
Skip editor-mock theater when you drop narration the moment the timer starts, or when you keep running code until tests pass—that trains debugging by trial, not interview performance. Also skip any practice-tool framing that promises stealth help in a real employer interview—that is the wrong goal and the wrong integrity bar.
For both: solutions and AI are for after the timebox, not mid-mock rescue if you are training transfer. The skill is your plan and your code, explained out loud, on the surface you will actually face.
Soft close
If your bottleneck is calm practice beside the editor you will actually use—timed prompts, narration, freeze recovery—Luna Interview is a Chrome side panel built for that practice loop, not for cheating or replacing study. Soft trial path: https://lunainterview.xyz/
Use rubber duck narration when building the habit of thinking out loud is the bottleneck. Use shared-editor timed mocks when keeping that habit under time is the bottleneck. The offer goes to the candidate who can keep talking when the clock is running—not the one who explains best to a duck.
Rehearse with an AI interview copilot, then browse commonly reported interview questions.