Collatz conjecture

Does the Collatz rule always lead to $1$? Choose a positive whole number. If it is even, halve it. If it is odd, triple it and add one. Apply the same rule to each new number.

Try $6$: it is even, so it becomes $3$. That is odd, so it becomes $10$. Continuing gives $6, 3, 10, 5, 16, 8, 4, 2, 1$. This start reaches $1$. The conjecture says every positive whole-number start does.[1]

Does every positive integer eventually reach $1$ under the ordinary Collatz rule? Start at $6$. Halving the even value gives $3$; tripling that odd value and adding one gives $10$. Halving gives $5$, and the odd rule then gives $16$. Repeated halving completes the trajectory:

$$6, 3, 10, 5, 16, 8, 4, 2, 1.$$

This takes eight steps to reach $1$. Continuing the rule produces the cycle $1, 4, 2, 1,\ldots$, so “reaches $1$” means arrival there, not that the rule stops.

To state the question precisely, write $C(n)$ for one step:

$$C(n)=\begin{cases}n/2 & \text{if }n\text{ is even},\\3n+1 & \text{if }n\text{ is odd}.\end{cases}$$

Starting at $N$, let $n_0=N$ and $n_{k+1}=C(n_k)$. The conjecture says that for every positive integer $N$, some finite step $k\ge0$ has $n_k=1$. The start $N=1$ already satisfies this at step zero. The question excludes zero, negative integers and changes to the rule.[1]

The Collatz conjecture asks about every positive whole number. A calculation can show that a chosen start reaches $1$, but checking any finite collection leaves infinitely many other starts.

No proof establishes the claim for every start. Nor has a counterexample been established: a start whose trajectory never reaches $1$. One successful path is easy to check; a proof covering them all remains missing.[1]

For the ordinary Collatz rule, Tao's theorem gives a precise kind of widespread descent. Choose a threshold depending on the starting value that eventually exceeds every fixed number; it may grow as slowly as you like. For almost all starts, some value along the trajectory falls below that threshold. This concerns the smallest value reached, rather than arrival at $1$.

Write $m(N)$ for the smallest value in the trajectory starting at $N$. For any chosen function $f(N)$ tending to infinity, Tao proves

$$m(N)<f(N)$$

for almost all positive integers $N$, in logarithmic density. To understand this share, compare starts up to a growing cutoff and give start $N$ weight $1/N$. The weighted share of the exceptional set tends to zero.

Exceptions may still exist, and the threshold must grow without limit. Setting it permanently to $1$ is not allowed. The theorem therefore does not establish that every trajectory reaches $1$; the full conjecture remains open.[1]

Collatz belongs to a bigger mathematical problem: predicting what happens when a rule is repeated and growth competes with shrinkage. Researchers use it to investigate how an exact, predictable next step can produce complicated behavior over time.

The work has already connected number patterns with probability and even ideas from wave equations. Tao adapted one of those ideas to prove a partial result about descent. That exchange of methods is a concrete payoff: an apparently isolated number puzzle brings different parts of mathematics together.[1] [3]

Collatz is a meeting point for arithmetic and dynamical systems: whole-number structure shapes the long-term behavior of a repeated rule. Lagarias places it in a broader family of maps with competing expansion and contraction, where general methods would be valuable beyond this particular conjecture. Probabilistic models offer predictions for these deterministic processes; establishing when those predictions justify theorems is part of the mathematical interest.

An existing result shows the value of those connections. Tao adapted an idea from Bourgain's work on nonlinear Schrödinger equations: use probability distributions with suitable invariance properties to extend control from shorter to longer times. In the Collatz setting, the distributions have to respect the arithmetic of the iterated rule. This helped produce Tao's descent theorem for almost all starts in logarithmic density. It remains a partial result, but the transfer of a method from wave equations into number theory is an actual contribution, independent of whether the full conjecture is eventually proved.[1] [3]