[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-517158-105":59,"doc-detail-517158-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","playing-multi-action-adversarial-games-online-evolutionary-planning-versus-tree-search","Playing Multi-Action Adversarial Games - Online Evolutionary Planning versus Tree Search","","Playing turn-based multi-action adversarial games is difficult because players can take multiple actions each turn, producing extremely high branching factors that make standard tree search methods unable to reach sufficient depth. The paper introduces Online Evolutionary Planning (OEP), which searches action combinations for a single turn using a fitness function that scores the resulting state. Experiments compare OEP with multiple MCTS variants on Hero Academy and show OEP searches plans more efficiently as per-turn action counts increase.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/playing-multi-action-adversarial-games-online-evolutionary-planning-versus-tree-search/517158/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/playing-multi-action-adversarial-games-online-evolutionary-planning-versus-tree-search/517158.png","ImageObject",300,407,{"name":92,"@type":93},"A glass of water","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-16","2026-10-07",true,{"@type":102,"interactionType":103,"userInteractionCount":34},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"What makes tree search methods fail in multi-action adversarial games?","Question",{"text":112,"@type":113},"Players take multiple actions per turn, causing extremely high branching factors. This makes search trees very shallow and prevents methods like standard tree search from reaching enough depth.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"How does Online Evolutionary Planning (OEP) address the single-turn planning challenge?",{"text":117,"@type":113},"OEP searches for combinations of actions to take during one turn. It evaluates candidate turn endings with a fitness function based on the quality of the resulting state.",{"name":119,"@type":110,"acceptedAnswer":120},"How does OEP compare to constrained Monte Carlo Tree Search in the study?",{"text":121,"@type":113},"Constrained MCTS variants outperform vanilla MCTS by a large margin. However, OEP is more efficient at searching the plan space, especially when the number of actions per turn increases.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},517158,1791423920,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":34,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":52,"language":139,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":67,"update_tm":143,"read_time":144,"preview_page_count":19,"preview_urls":145},962090760832,"https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d","Playing Multi-Action Adversarial Games: Online Evolutionary Planning versus Tree Search  \nNiels Justesen, Tobias Mahlmann, Sebastian Risi and Julian Togelius  \nWe address the problem of playing turn-based multi-action adversarial games, which include many strategy games with extremely high branching factors as players take multiple actions each turn. This leads to the breakdown of standard tree search methods, including Monte Carlo Tree Search (MCTS), as they become unable to reach a sufﬁcient depth in the game tree. In this paper, we introduce Online Evolutionary Planning (OEP) to address this challenge, which searches for combinations of actions to perform during a single turn guided by a ﬁtness function that evaluates the quality of a particular state. We compare OEP to different MCTS variations that constrain the exploration to deal with the high branching factor in the turn-based multi-action game Hero Academy. While the constrained MCTS variations outperform the vanilla MCTS implementation by a large margin, OEP is able to search the space of plans more efﬁciently than any of the tested tree search methods as it has a relative advantage when the number of actions per turn increases.  \nI. INTRODUCTION  \nAdversarial games, in which one player's loss is the others gain, have a long tradition as testbeds in artiﬁcial intelligence (AI) . In this context, playing the game well can be viewed asa search from the current state of the game to desirable future states. In fact, many well performing game-playing programs rely on search algorithms that are guided by some heuristic function that evaluates the desirability of a given state. For adversarial two-player games with a relatively low branching factors, such as Checkers and Chess, search algorithms such as Minimax together with well-designed heuristic functions perform remarkably well [1] .  \nHowever, as the branching factor increases, the efﬁcacy of Minimax search is greatly reduced. In cases where it is hard to develop or learn informative heuristic functions this problem is further compounded. A classic example is Go, where it took several decades of research to come up with algorithms that play at world-class human level [2], [3] .  \nWhile Go has a branching factor of “only” 300 (a magnitude higher than the 30 for Chess), it is still far lower than most turn-based multi-action adversarial games, where each player takes multiple separate actions each turn, for example by moving multiple units or pieces. In this case, none of the mentioned methods currently perform well. Examples of multi-action turn-based adversarial games include strategy games such as Civilization, Warhammer 40k, XCOM or Heroes of Might and Magic but also card and board games such asthe board game Arimaa. This class of problems arguably also includes many real-world problems involving coordination of multiple agents.  \nBy including multiple actions and multiple units, the branching factor quickly reaches intractable dimensions. For example, a strategy game that allows the movement of six  \nunits every turn and each unit can perform one out of ten actions has a branching factor of a million (106 ) . Standard tree search methods tend to fail with such high branching factors because the trees become very shallow. To allow tree search methods to be applied in such circumstances, some authors resort to making strong assumptions to guide how to explore actions [4] . One naive assumption, which is wrong in the general case, is independence between units, which decreases the branching factor to only 60 rather than a million.  \nInstead of relying on a tree-based search method, Perez et al. introduced an evolutionary algorithm-based method for playing non-adversarial games called Rolling Horizon Evolution (RHE; [5]) . RHE evolves a sequence of game actions to perform in the near future. The agent then performs the ﬁrst action in the found sequence and evolves a new sequence from scratch. This process is continued a","cbCaid9ELO2mi99h","https://ap.wps.com/l/cbCaid9ELO2mi99h","pdf",1559353,"English","# Introduction\n## Game search and branching factor challenges\n## Evolutionary planning and rolling horizon evolution (RHE)\n## Online Evolutionary Planning (OEP)\n## Comparison with Monte Carlo Tree Search (MCTS) variants","[{\"question\":\"What makes tree search methods fail in multi-action adversarial games?\",\"answer\":\"Players take multiple actions per turn, causing extremely high branching factors. This makes search trees very shallow and prevents methods like standard tree search from reaching enough depth.\"},{\"question\":\"How does Online Evolutionary Planning (OEP) address the single-turn planning challenge?\",\"answer\":\"OEP searches for combinations of actions to take during one turn. It evaluates candidate turn endings with a fitness function based on the quality of the resulting state.\"},{\"question\":\"How does OEP compare to constrained Monte Carlo Tree Search in the study?\",\"answer\":\"Constrained MCTS variants outperform vanilla MCTS by a large margin. However, OEP is more efficient at searching the plan space, especially when the number of actions per turn increases.\"}]","Playing Multi-Action Adversarial Games - Online Evolutionary Planning versus Tree Search | PDF",1791336245,25,[146,150,154,158],{"page_no":8,"url":147,"width":148,"height":149},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/ff645ad293297de17a245c2ae03325b2.webp",983,1272,{"page_no":14,"url":151,"width":152,"height":153},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/444e038609e62fc68f4b60727e4dda3c.webp",982,1270,{"page_no":81,"url":155,"width":156,"height":157},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/1205c33500152136bd496b5590e1a567.webp",981,1269,{"page_no":19,"url":159,"width":160,"height":161},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/47c4d7b269533c27a9f92614c7a00b20.webp",980,1268]