the amplitude (matrix element)

When you flip a coin, you do not predict heads or tails outright; you predict a probability. Quantum physics works the same way, but with an extra twist: before you get a probability, you first compute a more fundamental quantity called the amplitude. Think of the amplitude as a kind of complex number — it has a size and a direction, like an arrow — that encodes everything about how a particular start turns into a particular finish. For a collision, this object is often called the matrix element, written with the symbol M.

Here is the crucial machinery. To find the probability that incoming particles produce a specific outgoing set, you compute the amplitude for that process and then take its magnitude squared. Squaring an arrow's length gives a positive number, and that number (combined with how much room the outgoing particles have to move, the phase space) is what turns into an observable rate or cross section. The reason amplitudes, not probabilities, are fundamental is interference: if two different histories can lead to the same outcome, you add their amplitudes first and then square, so the arrows can reinforce or cancel — a uniquely quantum effect with no everyday analogue.

Computing the amplitude is precisely what Feynman diagrams are for: each diagram is a recipe for one contribution to M, and you sum the contributions before squaring. The amplitude is therefore the central bridge of the whole field — it is what theory calculates and what, once squared and folded with phase space, experiment measures. Everything downstream, from cross sections to decay rates to the shape of a resonance, flows from getting the amplitude right.