Itisatributetoourprofessionthatatextbookthatwascurrentin1999isstartingtofeelold.TheworkforthefirsteditionofMonteCarloStatisticalMethods(MCSM1)wasfinishedinlate1998,andtheadvancesmadesincethen,aswellasourlevelofunderstandingofMonteCarlomethods,havegrownagreatdeal.Moreover,twootherthingshavehappened.TopicsthatjustmadeitintoMCSM1withthebriefesttreatment(forexample,perfectsampling)havenowattainedalevelofimportancethatnecessitatesamuchmorethoroughtreatment.Secondly,someothermethodshavenotwithstoodthetestoftimeor,perhaps,havenotyetbeenfullydeveloped,andnowreceiveamoreappropriatetreatment.
WhenweworkedonMCSM1inthemid-to-late90s,MCMCalgorithmswerealreadyheavilyused,andtheflowofpublicationsonthistopicwasatsuchahighlevelthatthepicturewasnotonlyrapidlychanging,butalsonecessarilyincomplete.Thus,theprocessthatwefollowedinMCSM1wasthatofsomeonewhowasthrownintotheoceanandwastryingtograbontothebiggestandmostseeminglyusefulobjectswhiletryingtoseparatetheflotsamfromthejetsam.Nonetheless,wealsofeltthatthefundamentalsofmanyofthesealgorithmswereclearenoughtobecoveredatthetextbookalevel,sowe"swamon.