Chain and Weakest Link
also Weakest Link Principle · Chain Analogy
A chain's strength is fixed by its single weakest link, so improving any other link does nothing for the whole.
A chain holds only as much load as its weakest link can bear. If the weakest link breaks at 5000 newtons, raising a stronger link from 8000 to 8200 newtons leaves the chain’s limit at 5000. The only way to strengthen the whole is to find and strengthen the weakest link. This is Goldratt’s intuitive picture of a constraint: in any system whose parts are linked by dependencies, performance is set by one limiting factor, and effort spent anywhere else is wasted.
CF uses this image heavily. It explains why under 1% of elements in a tightly connected system drive over 99% of outcomes — when everything depends on a chain, only the weak link(s) matter. (Pareto’s 20/80 holds instead for independent causes, where there is no single chain.) It also explains why a balanced plant is a mistake: you cannot abolish the weakest link, and many links tied for weakest make a system more chaotic, not more efficient. Better to know where your weak link sits and plan around it.
The deeper payoff is epistemological. CF treats an idea like a chain: a single decisive criticism refutes it no matter how strong its other parts are, just as one weak link sets the whole chain’s strength. This grounds CF’s binary epistemology and its yes-or-no evaluation: ideas pass or fail, they are not scored by adding up strong points. Strengthening non-decisive features, like optimizing non-constraint links, changes nothing about whether the idea survives.