Separate logarithmic shower evolution from hadron formation

In a collinear approximation a splitting probability contains alpha s, a splitting kernel, a logarithmic interval of an ordering variable, and a momentum-sharing interval. A cutoff excludes unresolved singular regions. Integrating the allowed rate over a finite interval gives a positive resolved-emission intensity. Probability conservation then pairs this rate with an exponentially decreasing survival probability: the Sudakov form factor. Expanding that exponential supplies the virtual and unresolved terms needed to avoid interpreting every real-emission integral as an independent added probability.

The upper and lower scales are resolution boundaries, not a universal physical clock. Different showers order in virtuality, angle or transverse momentum, with coherence constraints determining which logarithms are correctly reproduced. Running-coupling arguments and allowed momentum fractions depend on the implementation. The elementary exponential below assumes a positive leading branching rate in an ordered probabilistic approximation. General fixed-order event weights and some improved evolution terms can be negative, so the entire precision calculation is not always a simple positive classical branching process.

After a branching, daughters have less energy and inherit color and recoil information. Subsequent rates and phase-space limits therefore differ from those of the original parent. A shower is not a binomial sequence of identical independent emissions. Wide-angle soft radiation sees coherent combinations of color charges, requiring angular-ordering or dipole-based treatment rather than an incoherent sum of isolated emitters. Leading-color approximations simplify color evolution but omit interference terms suppressed at large N, which are not necessarily negligible in every observable.

A fixed-order hard calculation accurately describes selected resolved multiplicities, while a shower resums classes of enhanced logarithms. Matching and merging prescriptions combine them while avoiding double counting the same emission. Their accuracy claims depend on both the fixed-order multiplicity and the logarithmic content of the shower. Changing a shower tune does not upgrade an approximate matrix element to a higher perturbative order. Momentum conservation, recoil choices and cutoff sensitivity should be tested alongside formal accuracy when comparing exclusive distributions.

Below a cutoff of order a hadronic scale, increasing alpha s in a splitting formula no longer gives controlled evolution. String models represent color-connected energy as a confining string that breaks into hadrons. Cluster models group color-connected partons into color-singlet systems that decay into hadrons, using preconfinement-inspired structure. Both require phenomenological parameters and prescriptions. They are useful and constrained by many measurements, but neither constitutes a general analytic solution of real-time confinement from the QCD action.

Hadron decays, underlying events and color reconnection introduce additional modeling choices that can affect observed particles. A tune optimizes agreement for selected data and should be tested on independent observables and energies. Its uncertainty is not exhausted by one renormalization-scale variation. Lattice QCD offers controlled nonperturbative calculations for many equilibrium and matrix-element quantities, yet ordinary Euclidean calculations do not simply output a fully exclusive collider event. The exact binomial lab below only illustrates a supplied survival complement across independent trials, not correlated shower evolution.