The Collatz Conjecture: A Simple Rule, A Stubborn Mystery
The Collatz conjecture defines a deterministic sequence by halving evens and mapping odds to 3n+1, and its apparent convergence to 1 for every tested integer remains unproven because the orbits resist the analytic tools that normally tame number-theoretic statements.
A complete interactive classroom, not just a preview.
Start when you are ready to enter this Stage's 9 scenes and explore, respond, and learn as you go.
Why does the Collatz sequence always seem to reach 1, and what makes that observation so hard to prove?
Take any whole number and apply two simple rules: if even, halve it; if odd, triple it and add one. Try 27. You will be amazed — and then unsettled.
Something this easy to state has fooled mathematicians for nearly a century. Nobody can prove that it always works, yet no one has found a counterexample.
An interactive simulation that lets you pick starting numbers and watch the trajectory unfold, comparing the behavior of small, medium, and large inputs side by side.
You will see exactly what the rule does, why it seems to always reach 1, and why that visible pattern is so hard to turn into a proof.
Because halving shrinks numbers faster than tripling grows them, repeated application should eventually drag any starting value down to 1.
- Full formal proof attempts (none exist)
- Connection to the Riemann hypothesis or other open problems
- Stochastic models and probabilistic heuristics in depth
- Computational verification campaigns beyond a brief mention
- 01Test Your First NumbersinteractivePrediction
Choose a starting integer and predict whether its Collatz sequence will reach 1, enter a short cycle, or grow without bound.
- Pick a small starting number like 6 or 27
- Commit to a prediction before stepping through the iteration
- Notice how quickly your intuition is challenged
- 02Before We Explain It: One PredictionquizPrediction
Commit to one answer about what the Collatz rule does for every positive integer.
- Decide whether the conjecture is proved or unproved
- Decide whether a counterexample is known
- 03A Rule You Can State in One SentenceslideQuestion
State the Collatz rule precisely, show a few worked steps, and frame the question that has haunted mathematicians since 1937.
- Even step: n becomes n/2
- Odd step: n becomes 3n+1
- Question: does every positive integer eventually reach 1?
- 04Watch Orbits in Real TimeinteractiveEvidence
Run several starting numbers simultaneously and compare total stopping times, peak heights, and the shape of each trajectory.
- Compare small, medium, and large starting values
- Observe total stopping time versus starting size
- Watch trajectories spike upward before collapsing
- 05What the Numbers ShowslideEvidence
Survey the empirical record: every integer tested up to roughly 10^20 reaches 1, and some famous cases like 27 take an unexpectedly long detour.
- Verification covers about 10^20 starting integers
- 27 takes 111 steps and peaks at 9232
- No counterexample has ever been found
- 06Why Halving and Tripling Should Cancel OutslideExplanation
Build the intuitive argument: even steps divide by 2, odd steps roughly multiply by 3, and over long runs the shrinking should win — but only on average, and that is exactly where trouble begins.
- Each odd step triples then adds 1
- Each even step halves
- Two odd steps and three even steps roughly multiply by 27/8
- Average drift is downward, but worst cases are not controlled
- 07Try a Modified RuleinteractiveTransfer
Change the constants in the rule (for example, 5n+1 instead of 3n+1) and see whether the modified sequence still converges, enters a cycle, or escapes to infinity.
- Edit the multiplier and additive constant
- Observe whether trajectories still reach 1
- Discover that small rule changes break the conjecture
- 08Where the Proof Breaks DownslideBoundary
Explain why the conjecture resists standard tools: there is no known invariant, no useful Lyapunov function, and no way to rule out a hidden long cycle or escape orbit.
- No conserved quantity has been found
- Stochastic models suggest convergence but are not proofs
- A counterexample would only need to exist somewhere above 10^20
- The problem sits at the edge of computability and number theory
- 09Answering the Driving QuestionslideResolution
Resolve the opening tension: the Collatz sequence reaches 1 for every integer we can test, and that visible regularity is precisely what makes its unproven status so unsettling.
- The rule itself is fully defined
- Empirical evidence is overwhelming but not a proof
- The conjecture remains famously open
- Its simplicity is what makes it deep
Discussion threads for a Stage aren't available yet.
