Mental models · first principles · case-style problem solving
The sharpest people you know aren't smarter — they're trained. They size markets on napkins, structure messy problems in seconds, and spot the flaw before the room does. NoodleBrain drills that thinking, 5 minutes a day: mental models, first-principles reasoning, and consulting-style case problems.
No app. No account. Your first drill starts right on this page.
What you train
Every daily drill pulls from these disciplines — the same toolkit consultants, founders, and investors run on. None of it is taught in school. All of it is trainable.
"How many pizzas does America eat a day?" Decompose, estimate, sanity-check — the napkin math that makes rooms go quiet.
Inversion, second-order thinking, opportunity cost, base rates — Munger's latticework, one model at a time, applied not memorized.
Strip a problem to its atoms and rebuild from what's actually true — instead of reasoning by analogy like everyone else.
MECE splits, issue trees, profit drivers. Take a messy "profits are down" and carve it clean — the core skill of case interviews.
Sunk cost, survivorship, sampling bias, the top-down fallacy. Learn the traps so you can watch other people fall into them.
The brain teasers interviewers still love — not for the answer, but for watching how your reasoning moves under pressure.
Today's drill
Six lanes, five drawn each day — everyone on Earth gets the same drill, and tomorrow's is already different. Reason it out before you pick; the rep is in the reasoning, not the click. Score 80+ and your thinking is officially al dente.
Today's Thinking Score
A new drill drops daily. Streaks and your model library — coming soon.
The official scale
Every drill grades you on the NoodleBrain doneness scale. There is exactly one correct texture — for pasta and for reasoning.
Reasons by vibes, argues by anecdote, accepts the first framing it's handed. Nods along in meetings. Recoverable — barely.
Smart but unstructured. Has good instincts, can't show the work, folds when someone confident disagrees. Most brains live here. Yours doesn't have to.
Firm under pressure. Sizes the market on a napkin, splits the problem MECE, names the bias in the room. The texture that gets hired, funded, and quoted.
The main event
Daily drills keep you sharp. The Bootcamp rebuilds how you think — a structured month, one cohort, one framework at a time:
Week 1
Deconstruct problems to bedrock. Daily teardowns of real decisions — pricing, careers, products — rebuilt from atoms up.
Week 2
One mental model per day — inversion, second-order effects, base rates, opportunity cost — each applied to a live problem, never just defined.
Week 3
MECE trees, market sizing, profit diagnostics. By Friday you can carve up "revenue is down 20%" like a partner at the whiteboard.
Week 4
Full mini-cases combining everything, against the clock. Finish with a graded final case and your Al Dente Certificate.
Next cohort starts the 1st
Founding cohort capped at 100 thinkers.
The Playbook
The reference library behind every drill — open it whenever a problem stares back at you. Reading it is knowing; the daily drills are doing.
1. Clarify the question — units, geography, timeframe ("pizzas per day, US, all channels").
2. Decompose into things you can actually estimate: population → households → frequency.
3. Estimate each chunk with round numbers — ~130M US households, pizza night every ~2 weeks.
4. Multiply & sanity-check: 130M ÷ 14 ≈ 9M/day → ~3.3B/year. Does that feel like America? It does. Round numbers, stated assumptions, one sanity check — that's the whole magic trick.
Instead of "how do I succeed?", ask "what would guarantee failure?" — then systematically avoid each item. Planning a launch? List the five certain launch-killers (no distribution, fake problem, confusing pricing, no follow-up, wrong audience) and design against them. Avoiding stupidity is more reliable than seeking brilliance, which is why Munger never shut up about this.
A good split is Mutually Exclusive (no cause lives in two branches) and Collectively Exhaustive (no cause escapes). "Profits fell" → revenue and costs. Then recurse: revenue = price × volume; volume = customers × frequency. Three levels down, a vague panic has become a checklist. If your buckets overlap or something could hide between them, re-cut.
1. Name the assumption everyone's inheriting ("rent just costs this much").
2. Decompose the thing into what it's fundamentally made of — space, location, convenience; materials, labor, margin.
3. Price each part independently and ask which constraints are real physics vs. inherited habit.
4. Rebuild. Analogy asks "what do rockets cost?" First principles asks "what do rocket materials cost?" — and that gap built SpaceX.
"1% of a $50B market" is arithmetic cosplaying as strategy — any small percent of a huge number looks great and explains nothing. The cure is bottom-up: who exactly buys, why, at what price, and what does winning each one cost? Customers × price = revenue you can defend. If someone can't tell the bottom-up story, the top-down number is decoration.
Before reasoning about your case, ask what happens in most cases: most restaurants fail, most projects run late, most "this time is different" isn't. Anchor on the base rate, then adjust for genuine evidence you're the exception. The inside view is a story; the outside view is data. Start outside.
Fair questions
It trains exactly the muscles those interviews test — structuring, sizing, hypothesis-driven thinking — as a daily habit instead of a two-week cram. It's not a replacement for mock interviews with humans; it's the reason you'll walk into them already firm.
You can read about push-ups too. Knowing that "inversion" exists and instinctively reaching for it under pressure are different skills — the second one only comes from reps. AI is a forklift: brilliant for heavy lifting, ruinous if you use it for every grocery bag. This is the part of your day where you deliberately lift without it.
Students prepping for consulting and finance interviews, founders who have to make calls with incomplete data, PMs and analysts who want to be the clearest thinker in the room — and anyone who's noticed their reasoning getting outsourced and wants it back.
Your streak resets and you keep going — the 30 days don't stop for you. But finish every day of a cohort and your next one is free. We're incentivized for you to actually change, not to churn.
Clear thinking is the last unfair advantage that can't be downloaded. Start your reps.
Do today's drill