Parton Showers since LEP

We briefly discuss the development of Monte Carlo event generators over the lastfifteen years. During this period there has been a revolutionary transformationin the accuracy of these programs as matching to higher-multiplicity matrixelements and next-to-leading order calculations has become standard with thefirst next-to-next-to-leading order processes now available. Finally theprospects for future improvements are discussed. Monte Carlo Simulations at LEPMonte Carlo event generators came of age at LEP where for the first time acombination of better understanding of QCD and increased computing powerprovided simulated events which were in good quantitative agreement with theexperimental results. These simulations used: • a leading-order matrix elementfor e + e − → qq; • a parton shower simulation for the evolution from the hardscale of the partonic collision to the infrared cutoff including the correcttreatment of colour coherence; • hadronization using either the non-perturbativestring or cluster models. The main programs used by the end of the LEP programmewere PYTHIA 6[1] and HERWIG 6[2]. These simulations also included the matchingof the hardest gluon emission for processes with a single colour line, forexample e + e − → qq, Deep inelastic scattering and Drell-Yan, which effectivelygave for e + e − → qq, apart from the trivial normalisation by a K-factor anext-to-leading order (NLO) simulation of the hard process. * The alternativedipole shower of ARIADNE (together with the string hadronization model) oftenprovide the best agreement with the data [4]. From LEP to the LHC Starting inthe early 2000's there was a major programme to develop better Monte Carlosimulations in order to describe the data from the energy frontier hadroncolliders, first the Tevatron and now the LHC. This started with the developmentof the first viable approach allowing multiple hard emissions to be describedcorrectly at leading order together with a parton shower simulation of soft andcolinear radiation (CKKW)[5]. This was first used to describe the production offour jet events in e + e − collisions where it gave quantitative improvements.However the main success of the approach was in hadron-hadron collisions whereit allowed the accurate description of multiple jet production, for example inassociation with electroweak vector bosons, for the first time.