Why It Is Hard
The twin prime problem sounds simple because the statement is short. The difficulty is that any proof has to control both the large-scale distribution of primes and the exact local conditions that produce gap-2 pairs.
This problem is hard because primes look partly irregular but obey strict arithmetic rules at the same time. A proof has to manage both sides at once, and it has to do so across infinitely many scales rather than just inside a large computed range.
It is one thing to prove that primes come close together infinitely often. It is a stronger and more delicate task to prove that the exact gap of 2 occurs infinitely often. That extra precision is where current methods still fall short. Bounded-gap theorems tell you that some fixed small distance returns forever. The twin prime conjecture asks for one exact distance and will not accept a near miss.
For a twin-prime pair (p, p+2), both numbers must avoid divisibility by many small primes at the same time. Those restrictions overlap and accumulate, which makes the local arithmetic much harder to control than the primality of a single number. A simple example is the usual 6k minus 1 and 6k plus 1 pattern: it filters out many impossible cases quickly, but it still leaves lots of composites behind. The real proof problem begins after the easy residue-class filtering has already been done.
A computer can find many twin-prime pairs in huge ranges, and those examples are genuinely interesting. But every computation ends at a finite cutoff. The conjecture asks what happens beyond every finite limit. That is why a million examples, a billion examples, or even vastly larger searches still do not by themselves produce a proof that infinitely many gap-2 pairs exist.
Modern methods often describe how primes behave on average or across large scales. The twin prime problem needs more than that. It asks for infinitely many exact local alignments, not just broad tendencies. Knowing that primes are well distributed on average is valuable, but it does not automatically force the exact local pairing needed for infinitely many twin primes.
Even though average distribution results are not enough by themselves, they are still part of the path forward. Progress on primes in arithmetic progressions, sieve methods, and bounded gaps all improves how much control mathematicians have over prime behavior. That is why the conjecture page, the Zhang page, and the Theory overview belong next to this page: they show which pieces of the larger puzzle have moved and which part remains unsolved.
Use this page when the conjecture seems easy to state but mysteriously hard to finish. Then move to Twin Prime Conjecture Explained for the clean statement of the problem, to the Zhang page for the biggest nearby theorem, or to Explorer and Analysis if you want to compare theorem language with concrete local patterns in a chosen range.
Use these links to keep reading or jump back into the live number views.
The full conjecture explainer gives the clearest statement of the exact gap-2 claim this page is discussing.
The Zhang page shows what modern mathematics did manage to prove without finishing the twin prime problem.
Theory keeps the full difficulty story connected to the rest of the conjecture context.
Explorer helps make divisibility and neighborhood structure concrete one row at a time.
Analysis helps connect the idea of local patterns and repeated small gaps to a real finite range.