Multi-player and multi-round auctions with severely bounded communication. ESA
- 格式:pdf
- 大小:161.32 KB
- 文档页数:12
采购管理理论综述作者:敬辉蓉, 李传昭, Gou Hui-rong, Li Chuan-zhao作者单位:敬辉蓉,Gou Hui-rong(重庆大学经济与工商管理学院,重庆400044;重庆师范大学旅游学院,重庆400047), 李传昭,Li Chuan-zhao(重庆大学经济与工商管理学院,重庆,400044)刊名:工业工程英文刊名:INDUSTRIAL ENGINEERING JOURNAL年,卷(期):2008,11(2)被引用次数:3次1.Harris F W Operations and Cost-factory management series 19152.Inman R A Quality certification of suppliers by JIT manufacturers 1990(02)3.Jahnukainen J;Lahti M Efficient purchasing in make-to-order supply chains[外文期刊] 1999(59)4.Fiorito Susan S;Giunipero Larry C;He Yan Retail buyers' perceptions of quick response systems 1998(06)5.Wu Min;Low SOi Pheng Economic order quantity (EOQ)versus just-in-time (JIT) purchasing:an ahemative analysis in the ready-mixed concrete industry[外文期刊] 2005(23)6.Niall Waters-Fuller Just-in-time purchasing and supply:a review of the literature 1995(09)7.Ganesh lyer;J Miguel Villas-Boas A bargaining Theory of distribution channels 2003(XI)8.Weng Z Kevin Channel coordination and quantity discounts 1995(09)9.thing-Hua;Chen-Ritzo;Terry P Harrison Better,faster,cheaper:an experimental analysis of a multiattribute reverse auction mechanism with restricted infor mation feedback[外文期刊] 2005(12)10.Leandro Arezamena;Estelle Cantillon Investment ineenfives in procurement auctions 2004(71)11.David C Parkes;Jayant Kalagnanam Model for iterative multi-attribute procurement auctions[外文期刊] 2005(03)12.Yossi Sheffi Combinatorial auctions in the procurement of transportation services[外文期刊] 2004(04)13.Rachel R Chen;Robin O Roundy;Rachel Q Zhang Efficient auction mechanisms for supply chain procurement[外文期刊] 2005(03)14.Hariga M;Haouari M An EOQ lot sizing model with random supplier capacity[外文期刊] 1999(01)15.Silver E A;Pyke D F;Peterson R Inventory management and production planning and scheduling 199816.Khouja M The single period (news-vendor) inventory problem:a literature review and suggestions for future research 1999(27)17.Willis T;Hnston C Vendor requirements and evualation in a JIT environment 1990(04)18.Naumann E;Reck R A buyer's bases of power 1982(04)19.Alan Smart;Alan Harrison Reverse auctions as a support mechanism in flexible supply chain2002(03)20.John G Riley;William F Samuelson Optimal auction 1981(71)21.Roger B Myerson Optimal auction design[外文期刊] 1981(06)22.Vickrey,William Counterspeculation,auctions,and competitive sealed tenders 1961(16)23.Yen Benjamin P-C;Ng Elsie O S The impact of electronic commerce on procurement[外文期刊] 2003(3-4)24.Monaban J P A quantity discount pricing model to increase vendor profits 1984(30)25.Susan A Shaw;Juliette Gibbs Procurement strategies of small retailers faced with uncertainty:an analysis of channel choice and behavior[外文期刊] 1999(01)26.Robert A Novack;Stephen W Simco The industrial procurement process:a supply chain perspective 1991(01)27.Dr David R Rink;Dr Harold W Fox Using the product life cycle concept to formulate actionable purchasing strategies 2003(02)28.Hammer Michael;Champy James Reengineering the corporation,a manifesto for business revolution 199329.Sridhar K Moorthy Managing channel profits:comment 1987(04)30.Jeuland A;Shugan S Managing channel profits 1983(02)31.James M Wilson A simulation analysis of ordering policies under inflationary conditions:a critique 1995(08)32.Wirth Andrew Technical paper:inventory control and inflation:a review 1989(01)33.Chang P L;Lin C T On the effect of centralization of the expected costs in a multi-location newsboy problem 1991(42)u A;Lau H The newsstand problem:a capaeitated multiproduct single period inventory problem 1996(94)35.Gullu R;Onol E;Erkip N Analysis of an inventory system under supply uncertainty 1999(59)36.Burton T T JIT/repetitive sourcing strategies:tying the knot with your suppiers 1988(04)1.周虹基于Web Services的物资采购管理系统的研究与应用[期刊论文]-中国科技信息 2010(21)2.王正武.黄巍物资信息的流转模式浅析[期刊论文]-兵工自动化 2009(2)3.周永宏绿色供应链条件下的企业采购管理[期刊论文]-科技资讯 2008(20)本文链接:/Periodical_gygc200802001.aspx。
欲善其事必先利其器拍卖行插件Auctioneer教程Autioneer Advanced 101Auctioneer完全教程——教你怎样用Auctioneer赚钱而不是去Farming目录第一章Auctioneer介绍第二章使用简单拍卖出售第三章使用估价师第四章高级估价师:批出售和价格匹配(即自动削价)设置第五章使用Search界面和Auctioneer工具第六章使用Search界面中的查询器和过滤器第七章使用BeanCounter来跟踪你的效率第八章关于Auctioneer工具深入过程的建议方案附录:反馈和进一步帮助第一章Auctioneer介绍Auctioneer是什么那?它是一个在Wow游戏中扩占拍卖厅功能的用户界面插件。
基本上来说,它所提供的功能正是你在拍卖行逐渐变富过程中所要求的,如果你正在规划如何在Wow中赚取金币,那这个插件应该作为你的入门课。
这个插件帮助人们在游戏中赚取金币的主要方法是为你收集数据信息并分析这些数据。
你可以使用Auctioneer去扫描目前拍卖行在列的所有拍卖条目,然后记录正在售出的物品以及其正在售的价格。
你可以每天扫描一次,假以时日之后,你可以对大多数物品销售的最佳价格得到一个非常清晰的图景。
Auctioneer能够对比拍卖物品的热销价格和NPC的销售价甚至能够对比一个物品附魔分解以后的潜在价格,这样你要在拍卖行买东西这方面得到更多的金币,分解它或者卖给NPC。
下载和安装Auctioneer现在让我们开始下载并安装这个插件,Auctioneer这个插件主要有两种版本。
第一:Auctineer套件完全版,这个版本包含全部有趣且高级的Auctioneer组件,比如定价的估价师组件和能发现可以购买并以一定利润转售的物品的SearchUI界面组件。
第二:基本的Auctioneer插件精简版。
它仅包含基本定价和出售功能,这个简单版对那种仅仅出售自己拾取、收集、制造的物品人很有用。
Exploring auction mechanisms for role assignment inteams of autonomous robotsVanessa Frias-Martinez1,Elizabeth Sklar1,and Simon Parsons21Department of Computer ScienceColumbia University1214Amsterdam Avenue,New York,NY10027,USAvf2001,sklar@2Department of Computer and Information ScienceBrooklyn College,City University of New York2900Bedford Avenue,Brooklyn,NY11210,USAparsons@Abstract.We are exploring the use of auction mechanisms to assign roles withina team of agents operating in a dynamic environment.Depending on the degreeof collaboration between the agents and the specific auction policies employed,we can obtain varying combinations of role assignments that can affect both thespeed and the quality of task execution.In order to examine this extremely largeset of combinations,we have developed a theoretical framework and an envi-ronment in which to experiment and evaluate the various options in policies andlevels of collaboration.This paper describes our framework and experimentalenvironment.We present results from examining a set of representative policieswithin our test domain—a high-level simulation of the RoboCup four-leggedleague soccer environment.1IntroductionMulti agent research has recently made significant progress in constructing teams of agents that act autonomously in the pursuit of common goals[12,15].In a multi agent team,each agent can function independently or can communicate and collaborate with its teammates.When collaborating,the notion of role assignment is used as a means of distributing tasks amongst team members by associating certain tasks with particular roles.The assignment of roles can be determined a priori or can change dynamically during the course of team operation.Collaboration enables a team of agents to work together to address problems of greater complexity than those addressed by agents operating independently.In general, using multiple robots is often suggested to have several advantages over using a single robot[4,7].For example,[11]describes how a group of robots can perform a set of tasks better than a single robot.Furthermore,a team of robots can localize themselves better when they share information about their environment[7].But collaboration in a team of robots may also add undesirable delays through the communication of information between the agents.We are exploring—within dynamic,multi-robot environments—the use of auction mechanisms to assign roles to agents dynamically and the effect of different approachesto collaboration within the team.In order to evaluate this set,we have developed a theoretical framework and a simulation environment.The theoretical framework helps us to identify the space of possibilities,and the simulation environment helps us to evaluate the various degrees of collaboration.This paper begins by highlighting some background material on auctions and the use of auction mechanisms in multi agent systems.Then we describe our theoretical framework.Next we detail our experimental environment—a high-level simulation of the RoboCup Four-Legged Soccer League.We then present results of simulation experiments evaluating both collaborative and non-collaborative models of information sharing as well as various auction policies.Finally,we close with a brief discussion and directions for future work.2AuctionsFollowing Friedman[9],we can consider an auction to be a mechanism that regulates how commodities are exchanged by agents operating in a multi agent environment.An auction mechanism defines how the exchange takes place.It does this by laying down rules about what the traders can do—what messages they can exchange in an interac-tion—and rules for how the allocation of commodities is made given the actions of the traders.Auctions have been used in different environments for resource allocation, such as electronic institutions[6],distributed planning of routes[13]or assignment of roles to a set of robots to complete a common task[10].3Theoretical frameworkIn our auction,there are two types of agents:the auctioneer and the trader—a player in the RoboCup soccer game.The player makes an offer and the auctioneer’s job is to coordinate the offers from all the players and perform role assignment.There arefive main components to our model.First,we define R to be the set of possible roles:R={P A,OS,DS},where P A is a primary attacker,OS is an offensive supporter,and DS is a defensive supporter. Note that the goalie is not considered a role to be assigned in this manner,since it cannot change during the course of the game.Next,we define P to be a set of player attributes:P={d ball,d goals,d mates,d opps} where d ball contains the distance from the player(who is making the offer)to the ball;d goals contains the distance from the player to each goal;d mates contains the distance from the player to each of its teammates;and d opps contains the distance from the player to each player on the opposing team.Third,we define F to be a set of functions which define the method for sharing per-ception information between agents.This information could be shared with teammates, the auctioneer,or both.Fourth,we define M to be a matching function,the method used by the auctioneer for clearing the auction,i.e.,matching the offers with roles.In other words,the matching function captures the coordination strategy.Finally,we define an auction,A,to be:A= P,R,M,f where P⊆P and P=∅;R⊆R and R=∅; M⊆M and M=∅;and f∈F.Our work is systematically exploring the space of all possible auctions P×R×M×F.B denotes the set of possible types of offers in a particular auction,A∈A: B={r,w}where:r⊆R is a set of roles for which the player bids;w is a set of real-valued weights,one weight corresponding to each of the roles in r(a weight of0 means that the player is not interested in making an offer for the corresponding role);and f(p),p⊆P,is the mechanism by which perceptual data is used to determine r and w.To date,we have defined two different types of auctions within this framework—a simple auction[5]and a combinatorial auction[3].We can define a simple auctionb t∈B as:b t={r,w},where the role r and w are singletons(unique offer).And a combinatorial auction,is defined as:b t={(r0,r1,r2),(w0,w1,w2)}where r i and w j are ing different combinations of weights allows the agent to bid for different combinations of roles,and this makes the auction combinatorial[1].4SimRob:our Simulated Approach to a RoboCup GameWe are using RePast[14]to implement our environment.RePast allows us to build a simulation as a state machine in which all the changes to the state machine occur through a schedule.In order to model a RoboCup soccer game in RePast,we need to define the agents,the environment and the state machine that RePast will execute at each scheduled tick,i.e.,simulated time step.4.1Agent parametersThe RoboCup Four-Legged League environment has four Sony AIBO robots per team and a bright orange ball.Each one of the robotic agents is associated with an array containing the values that define their perception and localization:(x,y,φ,d ball,d goals,d opps,d mates,b ball,b goals,b opps,b mates)(1) where(x,y)are the2D coordinates of the robot on thefield3;φis the orientation of the robot4;d ball is the distance from the robot to the ball,d goals is the distance from the robot to each goal,d opps is an array containing the distance from the robot to each opponent,and d mates is an array containing the distance from the robot to each team-mate.The boolean values in the second half of equation(1)indicate if the ball has been detected by the player(b ball),if each goal has been detected by the player(b goals),if each opponent has been detected nearby(b opps)and if each teammate has been detected nearby(b mates).4.2Simulation skeletonWe use RePast in order to simulate the development of a game with the agents.At the beginning of the simulation,we define four agents(per team)and a ball in thefield.Each of the agents is defined as explained above,by means of an array as in equation (1).The simulation run in RePast can be divided into the following steps:(1a)Generation of the agent parameters In thisfirst step,we obtain the parameters of each of the agents in thefield.The localization of the robot is expressed with the coordinates(x,y)in a2Dfield.We also obtain the distances to the ball d ball,to the goal d goals and to the opponents d opps.(1b)Amount of information shared by the agents The information shared by the agents is:mingoal,a boolean variable that is true when the agent is the one closest to the goal. This variable can be defined when the agents share the variable d goals among them. maxopp is a boolean variable that is true when the agent is farthest away from the opponents in thefield.This variable can be defined when the agents share d opps.And maxball is a boolean variable that is true when the agent is farthest away from the ball. This value can be defined when the variable d ball is shared among the agents.(2a)Defining a bidding policy for the agents For each simulation tick of the game play,the agent’s bid will be the role associated by the policy being tested to the set of perceptions gathered by the agent at that simulation tick.(2b)Defining an auction policy for the auctioneer The auction is responsible for dis-tributing the roles between the agents on thefield.The auctioneer will go through the different roles in the bid until one of the roles in the array is assigned to the agent, meaning that the bid is won.(3)Game Play Once the agent-roles are defined,we have to actually simulate the joint task to be developed by the agents.As stated before,our aim is that of simulating a soccer game.The game model is very simple.Each role has a state graph that will output a certain behavior depending on the perceptions gathered by the agent:–PA B EHAVIOR:If the agent sees the goal and the ball,then it kicks the ball,other-wise it turns to look for the ball without losing track of the goal.–OS B EHAVIOR:If the ball is seen,the agent kicks it.–DS B EHAVIOR:If the ball is seen,the agent follows it in order to prevent an agent from the opposing team scoring.Finally,if a goal is scored,the robots are sent back to their initial positions and the ball randomly changes location.Then,the three step(parameter generation,auction execution and game play)simulation is run again.5ExperimentsThis section describes our experimental work to date.We have started to explore the range of possible auctions and their effect on the coordination of a team,as measured by their performance in simulated games.We have experimented with four very sim-ple types of coordination and describe policies that we have used for experimentation, chosen somewhat ad hoc.In current work,we are learning policies[8].Table1.Example non-collaborative simple(S)and combinatorial(C)auctionsBall seen Mate seen Role(C)0DS01[OS,.7,DS,.2,PA,.1]1DS01[OS,.7,DS,.2,PA,.1]0DS11[OS,.7,DS,.2,PA,.1]1OS11[OS,.7,DS,.2,PA,.1]5.1Non-collaborative simple auctionThis approach defines a team of agents that don’t share any perception data.Hence, each one relies on the information that it gathers independently of the others.The offers made by the agents follow the policy in Table1column Role(S).This shows that we have defined the agent to offer to be OS when both ball and opponent are seen.In any other case,our agent will offer to be DS.We have chosen a simple matching policy that just associates afixed role to each of the possible sets of perceptions.5.2Non-collaborative combinatorial auctionIn this case there is still no sharing of perception,but the bid now contains a vector defining the agent’s role preferences For our experiments,we have defined two differ-ent bidding policies.The offensive policy,defined in Table1,column Role(C),repre-sents a team with an attacking approach,always looking for the goal and aiming to score.The other policy is more defensive.The offensive policy assigns the array of roles[DS,.7,OS,.2,PA,.1]to each of the agents.The matching is the same as before.5.3Collaborative simple auctionIn this case,the agents share all the perception data.Hence,when defining the bids,we can also share the three variables related to the minimum and maximum distances to the ball,opponents and goal.The table defining the bidding policy is huge.In Table2,col-umn Role(S),we show a few lines to give the sense of it,but it is deliberately similar to the policy for the non-collaborative auction to give a reasonable comparison.When no elements are seen by any of the agents,the agent bids for the role DS.When everything is seen and the distances are minimum,the agents bid to be OS.The matching policy is also the same as for the non-collaborative examples.5.4Collaborative combinatorial auctionHere the bidding Table2,column Role(C),is similar to the previous one,but contains a vector of bids and weights instead of only one role,and this vector is like that for theTable2.Collaborative simple(S)and combinatorial(C)auctions Ball seen Mate seen MaxOpp Role(S) 000[DS]000[DS]......111[OS]defensive offensivebid bidnoncollab simple16–4330collab simple40–3778ticks (time)g o a l s s c o r e d(a)unique matching policy (b)non-unique matching policyFig.1.Goals scored over the course of a gamesociated with the acceptance of the bids made by an agent.The higher the ratio,the more times its bid has been accepted.In the not uniqueness experiments,we obtained very low ratios,meaning that the agents almost never won a bid,and so,the roles were distributed randomly.6Conclusions and Future workThis paper has described our preliminary work in exploring the use of auction mecha-nisms to coordinate players on a RoboCup team.While this work is only just beginning,we believe that the results demonstrate the potential of the approach to capture a wide range of types of coordination,and to be able to demonstrate their effectiveness through simulation.In addition,this approach makes it simple to explore more complex,and po-tentially more flexible,kinds of role allocation than have been previously used in the legged-league,for example [2,16].Our longterm work is to build on this foundation and explore a wide range of pos-sible auctions through simulation and on real (physical)robots.We are currently using learning techniques to automatically explore the space of auctions.We further intend to implement the most effective bidding and matching policies developed on our real Legged-League team.7AcknowlegementsThis work was made possible by funding from NSF #REC-02-19347and NSF #IIS 0329037.References1.C.Boutilier and H.H.Hoos.Bidding languages for combinatorial auctions.In Proceedingsof the17th International Joint Conference on Artificial Intelligence,pages1211–1217,San Francisco,CA,2001.Morgan Kaufmann.2.D.Cohen,Y.Hua,and P.Vernaza.The University of Pennsylvania Robocup2003LeggedSoccer Team.In Proceedings of the RoboCup Symposium,2003.3.S.de Vries and R.V binatorial auctions:A RMS Journal of Comput-ing,(to appear).4.G.Dudek,M.Jenkin,E.Emilios,and D.Wilkes.A taxonomy for multi-agent robotics.Autonomous Robots,3(4),1996.5.R.Engelbrecht-Wiggans.Auctions and bidding models:A survey.Management Science,26:119–142,1980.6.M.Esteva and J.Padget.Auctions without auctioneers:distributed auction protocols.InAgent-mediated Electronic Commerce II,LNAI1788,pages20–28.Springer-Verlag,2000.7.D.Fox,W.Burgard,H.Kruppa,and S.Thrun.Collaborative multi-robot localization.In Pro-ceedings of the23rd German Conference on Artificial Intelligence.Springer-Verlag,1999.8.V.Frias-Martinez and E.Sklar.A team-based co-evolutionary approach to multi agent learn-ing.In Proceedings of the2004AAMAS Workshop on Learning and Evolution in Agent Based Systems,2004.9.D.Friedman.The double auction institution:A survey.In D.Friedman and J.Rust,editors,The Double Auction Market:Institutions,Theories and Evidence,Santa Fe Institute Studies in the Sciences of Complexity,chapter1,pages3–25.Perseus Publishing,Cambridge,MA, 1993.10.B.Gerkey and M.Mataric.Sold!:Auction methods for multirobot coordination.IEEETransactions on Robotics and Automation,2000.11.D.Guzzoni,A.Cheyer,L.Juli,and K.Konolige.Many robots make short work.AI Maga-zine,18(1):55–64,1997.12.G.A.Kaminka,D.V.Pynadath,and M.Tambe.Monitoring deployed agent teams.In J¨o rg P.M¨u ller,Elisabeth Andre,Sandip Sen,and Claude Frasson,editors,Proceedings of the Fifth International Conference on Autonomous Agents,pages308–315.ACM Press,2001.13.T.L.Lenox,T.R.Payne,S.Hahn,M.Lewis,and K.Sycara.Agent-based aiding for indi-vidual and team planning tasks.In Proceedings of IEA2000/HFES2000Congress,2000.14.Repast..15.M.Tambe.Towardsflexible teamwork.Journal of Artificial Intelligence Research,7:83–124,1997.16.M.Veloso and S.Lenser.CMPAck-02:CMU’s Legged Robot Soccer Team.In Proceedingsof the RoboCup Symposium,2002.。
西泠印社拍卖规则英文Rules for Xiling Yinshe Auction1. Introduction:Xiling Yinshe Auction is a reputable auction organization based in China. The auction aims to provide a fair and transparent platform for buying and selling artworks, including calligraphy, paintings, and other collectibles.2. Participation:2.1 Anyone above the age of 18 and possessing legal capacity can participate in the auction.2.2 Participants must register and obtain a bidding number before the auction. Registration requires providing identification documents and completing necessary forms.3. Bidding Process:3.1 Bidding can be done in person, online, or through absentee bids.3.2 Bidders must clearly indicate the lot number, title, and their bid amount. Bids are accepted within the stated bidding intervals.3.3 The highest bidder at the conclusion of the bidding period will be the successful buyer.3.4 The auctioneer reserves the right to reject any bid deemed unreasonably low.4. Buyer's Premium:4.1 A buyer's premium is applicable to the hammer price for each lot. The premium is a percentage of the final bid price.4.2 The buyer's premium is subject to change and will be announced prior to the auction.5. Payment:5.1 Payment must be made in full within a specified timeframe after the auction.5.2 Accepted payment methods include cash, bank transfer, or other approved payment platforms.5.3 The auction house reserves the right to refuse personal checks.6. Ownership and Delivery:6.1 Ownership of the purchased lot transfers to the buyer upon successful payment.6.2 The auction house provides shipment services for purchased lots. Additional charges may apply for packaging, insurance, and transportation.6.3 The buyer is responsible for any taxes, customs duties, or other charges related to the import or export of the purchased lot.6.4 Delivery timescales are estimates and may vary depending on various factors.7. Liability and Dispute Resolution:7.1 The auction house arranges all auctions with due care and professionalism. However, it is not responsible for any errors, omissions, or inaccuracies in catalog descriptions or related information.7.2 Any disputes arising from the auction will be handled according to the laws of the jurisdiction where the auction is conducted.7.3 Participants are encouraged to clarify any doubts or seek expert advice before participating in the auction.8. Miscellaneous:8.1 The auction house reserves the right to withdraw any lot from the auction for any reason.8.2 These rules may be subject to modification or amendment, which will be announced prior to the auction.Note: The above rules are a general outline and may be subject to specific terms and conditions applicable to each individual auction conducted by Xiling Yinshe Auction. Participants are advised to review the official auction catalog and consult the auction house for the most up-to-date rules and information.。
JEL分类号JEL分类号JEL分类系统,是美国经济学会《经济文献杂志》(Journal of Economic Literature)所创立的对经济学文献的主题分类系统,并被现代西方经济学界广泛采用。
该分类方法主要采用开头的一个英文字母与随后的两位阿拉伯数字一起对经济学各部类进行“辞书式”编码分类。
目录A-经济学总论和教学B-经济学思想流派和方法论C-数理和数量方法D-微观经济学E-宏观经济学和货币经济学F- 国际经济学G-金融经济学H-公共经济学I-卫生经济学,教育经济学和福利经济学J-劳动经济学和人口经济学K-法律和经济学L-产业组织M-企业管理和商务经济学;市场营销学;会计学N-经济史O- 经济发展,技术变迁和增长P-经济体制Q-农业经济学和自然资源经济学R-城市经济学,农村经济学和区域经济学Z-其他专题A-经济学总论和教学( A- General Economics and Teaching )A1A1 - General EconomicsA10 - GeneralA11 - Role of Economics; Role of Economists; Market for Economists A12 - Relation of Economics to Other Disciplines A13 - Relation of Economics to Social ValuesA14 - Sociology of EconomicsA19 - OtherA2A2 - Economics Education and Teaching of EconomicsA20 - GeneralA21 - Pre-collegeA22 - UndergraduateA23 - GraduateA29 - OtherA3A3 - Multisubject Collective WorksA30 - GeneralA31 - Multisubject Collected Writings of IndividualsA32 - Multisubject VolumesA39 - OtherB-经济学思想流派和方法论( B-Schools of Economic Thought and Methodology )B0B0 - GeneralB00 - GeneralB1B1 - History of Economic Thought through 1925B10 - GeneralB11 - PreclassicalB12 - ClassicalB13 - Neoclassical through 1925B14 - Socialist; MarxistB15 - Historical; InstitutionalB16 - QuantitativeB19 - OtherB2B2 - History of Economic Thought since 1925B20 - GeneralB21 - MicroeconomicsB22 - MacroeconomicsB23 - Econometrics; Quantitative StudiesB24 - Socialist; MarxistB25 - Historical; Institutional; Evolutionary; Austrian B29 - OtherB3B3 - History of Thought: IndividualsB30 - GeneralB31 - IndividualsB4B4 - Economic MethodologyB40 - GeneralB41 - Economic MethodologyB49 - OtherB5B5 - Current Heterodox ApproachesB50 - GeneralB51 - Socialist; Marxian; SraffianB52 - Institutional; EvolutionaryB53 - AustrianB59 - OtherC-数理和数量方法(C-Mathematical and Quantitative Methods) C0C0 - GeneralC00 - GeneralC1C1 - Econometric and Statistical Methods: GeneralC10 - GeneralC11 - Bayesian AnalysisC12 - Hypothesis TestingC13 - EstimationC14 - Semiparametric and Nonparametric MethodsC15 - Statistical Simulation Methods; Monte Carlo Methods; Bootstrap MethodsC16 - Specific DistributionsC19 - OtherC2C2 - Econometric Methods: Single Equation Models; Single VariablesC20 - GeneralC21 - Cross-Sectional Models; Spatial Models; Treatment Effect ModelsC22 - Time-Series ModelsC23 - Models with Panel DataC24 - Truncated and Censored ModelsC25 - Discrete Regression and Qualitative Choice ModelsC29 - OtherC3C3 - Econometric Methods: Multiple/Simultaneous Equation Models; Multiple VariablesC30 - GeneralC31 - Cross-Sectional Models; Spatial Models; Treatment Effect ModelsC32 - Time-Series ModelsC33 - Models with Panel DataC34 - Truncated and Censored ModelsC35 - Discrete Regression and Qualitative Choice ModelsC39 - OtherC4C4 - Econometric and Statistical Methods: Special TopicsC40 - GeneralC41 - Duration AnalysisC42 - Survey MethodsC43 - Index Numbers and AggregationC44 - Statistical Decision Theory; Operations ResearchC45 - Neural Networks and Related T opicsC49 - OtherC5C5 - Econometric ModelingC50 - GeneralC51 - Model Construction and EstimationC52 - Model Evaluation and TestingC53 - Forecasting and Other Model ApplicationsC59 - OtherC6C6 - Mathematical Methods and ProgrammingC60 - GeneralC61 - Optimization Techniques; Programming Models; Dynamic AnalysisC62 - Existence and Stability Conditions of EquilibriumC63 - Computational TechniquesC65 - Miscellaneous Mathematical T oolsC67 - Input–Output ModelsC68 - Computable General Equilibrium ModelsC69 - OtherC7C7 - Game Theory and Bargaining TheoryC70 - GeneralC71 - Cooperative GamesC72 - Noncooperative GamesC73 - Stochastic and Dynamic Games; Evolutionary Games C78 - Bargaining Theory; Matching TheoryC79 - OtherC8C8 - Data Collection and Data Estimation Methodology; Computer ProgramsC80 - GeneralC81 - Methodology for Collecting, Estimating, and Organizing Microeconomic DataC82 - Methodology for Collecting, Estimating, and Organizing Macroeconomic DataC87 - Econometric SoftwareC88 - Other Computer SoftwareC89 - OtherC9C9 - Design of ExperimentsC90 - GeneralC91 - Laboratory, Individual BehaviorC92 - Laboratory, Group BehaviorC93 - Field ExperimentsC99 - OtherD-微观经济学(D-Microeconomics)D0D0 - GeneralD00 - GeneralD01 - Microeconomic Behavior: Underlying PrinciplesD02 - Institutions: Design, Formation, and OperationsD1D1 - Household Behavior and Family EconomicsD10 - GeneralD11 - Consumer Economics: TheoryD12 - Consumer Economics: Empirical AnalysisD13 - Household Production and Intrahousehold Allocation D14 - Personal FinanceD18 - Consumer ProtectionD19 - OtherD2D2 - Production and OrganizationsD20 - GeneralD21 - Firm BehaviorD23 - Organizational Behavior; Transaction Costs; Property Rights D24 - Production; Capital and Total Factor Productivity; Capacity D29 - OtherD3D3 - DistributionD30 - GeneralD31 - Personal Income, Wealth, and Their DistributionsD33 - Factor Income DistributionD39 - OtherD4D4 - Market Structure and PricingD40 - GeneralD41 - Perfect CompetitionD42 - MonopolyD43 - Oligopoly and Other Forms of Market ImperfectionD44 - AuctionsD45 - Rationing; LicensingD46 - Value TheoryD49 - OtherD5D5 - General Equilibrium and DisequilibriumD50 - GeneralD51 - Exchange and Production EconomiesD52 - Incomplete MarketsD57 - Input–Output Tables and AnalysisD58 - Computable and Other Applied General Equilibrium Models D59 - OtherD6D6 - Welfare EconomicsD60 - GeneralD61 - Allocative Efficiency; Cost–Benefit AnalysisD62 - ExternalitiesD63 - Equity, Justice, Inequality, and Other Normative Criteria and MeasurementD64 - AltruismD69 - OtherD7D7 - Analysis of Collective Decision-MakingD70 - GeneralD71 - Social Choice; Clubs; Committees; AssociationsD72 - Economic Models of Political Processes: Rent-Seeking,Elections, Legislatures, and Voting BehaviorD73 - Bureaucracy; Administrative Processes in Public Organizations; CorruptionD74 - Conflict; Conflict Resolution; AlliancesD78 - Positive Analysis of Policy-Making and Implementation D79 - OtherD8D8 - Information, Knowledge, and UncertaintyD80 - GeneralD81 - Criteria for Decision-Making under Risk and Uncertainty D82 - Asymmetric and Private Information D83 - Search; Learning; Information and KnowledgeD84 - Expectations; SpeculationsD85 - Network FormationD86 - Economics of Contract LawD89 - OtherD9D9 - Intertemporal Choice and GrowthD90 - GeneralD91 - Intertemporal Consumer Choice; Life Cycle Models and Saving D92 - Intertemporal Firm Choice and Growth, Investment, or FinancingD99 - OtherE-宏观经济学和货币经济学(E-Macroeconomics and Monetary Economics)E0E0 - GeneralE00 - GeneralE01 - Measurement and Data on National Income and Product Accounts and WealthE1E1 - General Aggregative ModelsE10 - GeneralE11 - Marxian; Sraffian; Institutional; EvolutionaryE12 - Keynes; Keynesian; Post-KeynesianE13 - NeoclassicalE17 - Forecasting and SimulationE19 - OtherE2E2 - Consumption, Saving, Production, Employment, and Investment E20 - GeneralE21 - Consumption; SavingE22 - Capital; Investment (including Inventories); CapacityE23 - ProductionE24 - Employment; Unemployment; WagesE25 - Aggregate Factor Income DistributionE26 - Informal Economy; Underground EconomyE27 - Forecasting and SimulationE29 - OtherE3E3 - Prices, Business Fluctuations, and CyclesE30 - GeneralE31 - Price Level; Inflation; DeflationE32 - Business Fluctuations; CyclesE37 - Forecasting and SimulationE39 - OtherE4E4 - Money and Interest RatesE40 - GeneralE41 - Demand for MoneyE42 - Monetary Systems; Standards; Regimes; Government and the Monetary SystemE43 - Determination of Interest Rates; Term Structure of Interest RatesE44 - Financial Markets and the MacroeconomyE47 - Forecasting and SimulationE49 - OtherE5E5 - Monetary Policy, Central Banking, and the Supply of Money and CreditE50 - GeneralE51 - Money Supply; Credit; Money MultipliersE52 - Monetary Policy (Targets, Instruments, and Effects)E58 - Central Banks and Their PoliciesE59 - OtherE6E6 -Macroeconomic Policy Formation, Macroeconomic Aspects of Public Finance, Macroeconomic Policy, and General Outlook E60 - GeneralE61 - Policy Objectives; Policy Designs and Consistency; Policy CoordinationE62 - Fiscal Policy; Public Expenditures, Investment, and Finance; TaxationE63 - Comparative or Joint Analysis of Fiscal and Monetary Policy; StabilizationE64 - Incomes Policy; Price PolicyE65 - Studies of Particular Policy EpisodesE66 - General Outlook and ConditionsE69 - OtherF- 国际经济学(F-International Economics)F0F0 - General F00 - GeneralF01 - Global OutlookF02 - International Economic Order; Noneconomic International Organizations ·Economic Integration and Globalization: GeneralF1F1 - TradeF10 - GeneralF11 - Neoclassical Models of TradeF12 - Models of Trade with Imperfect Competition and Scale EconomiesF13 - Commercial Policy; Protection; Promotion; Trade Negotiations; International OrganizationsF14 - Country and Industry Studies of TradeF15 - Economic IntegrationF16 - Trade and Labor Market InteractionsF17 - Trade Forecasting and SimulationF18 - Trade and EnvironmentF19 - OtherF2F2 - International Factor Movements and International Business F20 - GeneralF21 - International Investment; Long-T erm Capital MovementsF22 - International MigrationF23 - Multinational Firms; International BusinessF29 - OtherF3F3 - International FinanceF30 - GeneralF31 - Foreign ExchangeF32 - Current Account Adjustment; Short-Term Capital Movements F33 - International Monetary Arrangements and InstitutionsF34 - International Lending and Debt ProblemsF35 - Foreign AidF36 - Financial Aspects of Economic IntegrationF37 - International Finance Forecasting and SimulationF4F4 - Macroeconomic Aspects of International Trade and Finance F40 - GeneralF41 - Open Economy MacroeconomicsF42 - International Policy Coordination and TransmissionF43 - Economic Growth of Open EconomiesF47 - Forecasting and SimulationF49 - OtherG-金融经济学(G-Financial Economics)G0G0 - GeneralG00 - GeneralG1G1 - General Financial MarketsG10 - GeneralG11 - Portfolio Choice; Investment DecisionsG12 - Asset PricingG13 - Contingent Pricing; Futures PricingG14 - Information and Market Efficiency; Event StudiesG15 - International Financial MarketsG18 - Government Policy and RegulationG19 - OtherG2G2 - Financial Institutions and ServicesG20 - GeneralG21 - Banks; Other Depository Institutions; MortgagesG22 - Insurance; Insurance CompaniesG23 - Pension Funds; Other Private Financial InstitutionsG24 - Investment Banking; Venture Capital; BrokerageG28 - Government Policy and RegulationG29 - OtherG3G3 - Corporate Finance and GovernanceG31 - Capital Budgeting; Investment PolicyG32 - Financing Policy; Capital and Ownership StructureG33 - Bankruptcy; LiquidationG34 - Mergers; Acquisitions; Restructuring; Corporate Governance G35 - Payout PolicyG38 - Government Policy and RegulationG39 - OtherH-公共经济学(H-Public Economics)H0H0 - GeneralH00 - GeneralH1H1 - Structure and Scope of GovernmentH10 - GeneralH11 - Structure, Scope, and Performance of GovernmentH19 - OtherH2H2 - Taxation, Subsidies, and RevenueH20 - GeneralH21 - Efficiency; Optimal TaxationH22 - IncidenceH23 - Externalities; Redistributive Effects; Environmental Taxes and SubsidiesH24 - Personal Income and Other Nonbusiness Taxes and Subsidies H25 - Business Taxes and SubsidiesH26 - Tax EvasionH27 - Other Sources of RevenueH29 - OtherH3H3 - Fiscal Policies and Behavior of Economic AgentsH30 - GeneralH31 - HouseholdH32 - FirmH39 - OtherH4H4 - Publicly Provided GoodsH40 - GeneralH41 - Public GoodsH42 - Publicly Provided Private GoodsH43 - Project Evaluation; Social Discount RateH49 - OtherH5H5 - National Government Expenditures and Related Policies H50 - GeneralH51 - Government Expenditures and HealthH52 - Government Expenditures and EducationH53 - Government Expenditures and Welfare ProgramsH54 - Infrastructures; Other Public Investment and CapitalStock H55 - Social Security and Public PensionsH56 - National Security and WarH57 - ProcurementH59 - OtherH6H6 - National Budget, Deficit, and DebtH60 - GeneralH61 - Budget; Budget SystemsH62 - Deficit; SurplusH63 - Debt; Debt ManagementH69 - OtherH7H7 - State and Local Government; Intergovernmental RelationsH70 - GeneralH71 - State and Local Taxation, Subsidies, and RevenueH72 - State and Local Budget and ExpendituresH73 - Interjurisdictional Differentials and Their EffectsH74 - State and Local BorrowingH77 - Intergovernmental Relations; FederalismH79 - OtherH8H8 - Miscellaneous IssuesH80 - GeneralH81 - Governmental Loans, Loan Guarantees, and CreditsH82 - Governmental PropertyH83 - Public AdministrationH87 - International Fiscal Issues; International Public Goods H89 - OtherI-卫生经济学教育经济学和福利经济学(I-Health, Education andWelfare)I0I0 - GeneralI00 - GeneralI1I1 - HealthI10 - GeneralI11 - Analysis of Health Care MarketsI12 - Health Production: Nutrition, Mortality, Morbidity, Substance Abuse and Addiction, Disability, and Economic Behavior I18 - Government Policy; Regulation; Public Health I19 - OtherI2I2 - EducationI20 - GeneralI21 - Analysis of EducationI22 - Educational FinanceI23 - Higher Education Research InstitutionsI28 - Government PolicyI29 - OtherI3I3 - Welfare and PovertyI30 - GeneralI31 - General Welfare; Basic Needs; Living Standards; Quality of Life; HappinessI32 - Measurement and Analysis of PovertyI38 - Government Policy; Provision and Effects of Welfare Programs I39 - OtherJ-劳动经济学和人口经济学(J-Labour and Demographic Economics)J00 - GeneralJ1J1 - Demographic EconomicsJ10 - GeneralJ11 - Demographic Trends and ForecastsJ12 - Marriage; Marital Dissolution; Family StructureJ13 - Fertility; Family Planning; Child Care; Children; Youth J14 - Economics of the Elderly; Economics of the Handicapped J15 - Economics of Minorities and Races; Non-labor Discrimination J16 - Economics of Gender; Non-labor DiscriminationJ17 - Value of Life; Foregone IncomeJ18 - Public PolicyJ19 - OtherJ2J2 - Time Allocation, Work Behavior, and Employment Determination and Creation; Human CapitalJ20 - GeneralJ21 - Labor Force and Employment, Size, and StructureJ22 - Time Allocation and Labor SupplyJ23 - Employment Determination; Job Creation; Demand for Labor; Self-EmploymentJ24 - Human Capital; Skills; Occupational Choice; Labor ProductivityJ26 - Retirement; Retirement PoliciesJ28 - Safety; Accidents; Industrial Health; Job Satisfaction; Related Public PolicyJ29 - OtherJ3J3 - Wages, Compensation, and Labor CostsJ31 - Wage Level and Structure; Wage Differentials by Skill, Training, Occupation, etc.J32 - Nonwage Labor Costs and Benefits; Private PensionsJ33 - Compensation Packages; Payment MethodsJ38 - Public PolicyJ39 - OtherJ4 - Particular Labor MarketsJ40 - GeneralJ41 - Contracts: Specific Human Capital, Matching Models, Efficiency Wage Models, and Internal Labor MarketsJ42 - Monopsony; Segmented Labor MarketsJ43 - Agricultural Labor MarketsJ44 - Professional Labor Markets and OccupationsJ45 - Public Sector Labor MarketsJ48 - Public PolicyJ49 - OtherJ5J5 - Labor–Management Relations, Trade Unions, and Collective BargainingJ50 - GeneralJ51 - Trade Unions: Objectives, Structure, and EffectsJ52 - Dispute Resolution: Strikes, Arbitration, and Mediation; Collective BargainingJ53 - Labor–Management Relations; Industrial Jurisprudence J54 - Producer Cooperatives; Labor Managed FirmsJ58 - Public PolicyJ59 - OtherJ6J6 - Mobility, Unemployment, and VacanciesJ61 - Geographic Labor Mobility; Immigrant WorkersJ62 - Job, Occupational, and Intergenerational MobilityJ63 - Turnover; Vacancies; LayoffsJ64 - Unemployment: Models, Duration, Incidence, and Job Search J65 - Unemployment Insurance; Severance Pay; Plant ClosingsJ68 - Public PolicyJ69 - OtherJ7J7 - Labor DiscriminationJ70 - GeneralJ71 - DiscriminationJ78 - Public PolicyJ79 - OtherJ8 - Labor Standards: National and International J80 - GeneralJ81 - Working ConditionsJ82 - Labor Force CompositionJ83 - Workers' RightsJ88 - Public PolicyJ89 - OtherK-法律和经济学(K-Law and Economics)K0K0 - GeneralK00 - GeneralK1K1 - Basic Areas of LawK10 - GeneralK11 - Property LawK12 - Contract LawK13 - Tort Law and Product LiabilityK14 - Criminal LawK19 - OtherK2K2 - Regulation and Business LawK20 - GeneralK21 - Antitrust LawK22 - Corporation and Securities LawK23 - Regulated Industries and Administrative Law K29 - OtherK3K3 - Other Substantive Areas of LawK30 - GeneralK31 - Labor LawK32 - Environmental, Health, and Safety LawK33 - International LawK34 - Tax LawK35 - Personal Bankruptcy LawK39 - OtherK4K4 - Legal Procedure, the Legal System, and Illegal Behavior K40 - GeneralK41 - Litigation ProcessK42 - Illegal Behavior and the Enforcement of LawK49 - OtherL-产业组织(L-Industrial Organization)L0L0 - GeneralL00 - GeneralL1L1 - Market Structure, Firm Strategy, and Market Performance L10 - GeneralL11 - Production, Pricing, and Market Structure; Size Distribution of FirmsL12 - Monopoly; Monopolization StrategiesL13 - Oligopoly and Other Imperfect MarketsL14 - Transactional Relationships; Contracts and Reputation; NetworksL15 - Information and Product Quality; Standardization and CompatibilityL16 - Industrial Organization and Macroeconomics; Macroeconomic Industrial Structure; Industrial Price Indices L19 - OtherL2L2 - Firm Objectives, Organization, and BehaviorL20 - GeneralL21 - Business Objectives of the FirmL22 - Firm Organization and Market Structure: Markets vs. Hierarchies; Vertical Integration; ConglomeratesL23 - Organization of ProductionL24 - Contracting Out; Joint VenturesL25 - Firm Performance: Size, Age, Profit, and SalesL29 - OtherL3 - Nonprofit Organizations and Public EnterpriseL30 - GeneralL31 - Nonprofit Institutions; NGOsL32 - Public EnterprisesL33 - Comparison of Public and Private Enterprises; Privatization; Contracting OutL39 - OtherL4L4 - Antitrust Issues and PoliciesL40 - GeneralL41 - Monopolization; Horizontal Anticompetitive Practices L42 - Vertical Restraints; Resale Price Maintenance; Quantity DiscountsL43 - Legal Monopolies and Regulation or DeregulationL44 - Antitrust Policy and Public Enterprise, Nonprofit Institutions, and Professional OrganizationsL49 - OtherL5L5 - Regulation and Industrial PolicyL50 - GeneralL51 - Economics of RegulationL52 - Industrial Policy; Sectoral Planning MethodsL53 - Government Promotion of FirmsL59 - OtherL6L6 - Industry Studies: ManufacturingL60 - GeneralL61 - Metals and Metal Products; Cement; Glass; CeramicsL62 - Automobiles; Other Transportation EquipmentL63 - Microelectronics; Computers; Communications Equipment L64 - Other Machinery; Business Equipment; ArmamentsL65 - Chemicals; Rubber; Drugs; BiotechnologyL66 - Food; Beverages; Cosmetics; TobaccoL67 - Other Consumer Nondurables: Clothing, Textiles, Shoes, and LeatherL68 - Appliances; Other Consumer Durables L69 - Other。
(转载)英语介绍⽃地主及其规则该篇⽂章转载⾃:IntroductionFight the Landlord (Dou Di Zhu) is a climbing game primarily for three players, but also playable by four. In each hand one player, the "landlord", plays alone and the others form a team. The landlord's aim is to be the first to play out all his cards in valid combinations, and the team wins if any one of them manages to play all their cards before the landlord. The game is said to have originated in Hubei province but is now popular all over China, and is also extensively played on line.Players, Cards and DealThe three-player game will be described first. The differences in the four-player game are explained near the end of the page.This game uses a 54-card pack including two jokers, red and black. The cards rank from high to low:red joker, black joker, 2, A, K, Q, J, 10, 9, 8, 7, 6, 5, 4, 3.Suits are irrelevant.As in most Chinese games the cards are not really dealt, but taken from the deck by the players.One of the players shuffles the cards, gives them to the player to his left to cut, and stacks them face down in the middle of the playing surface. One card is turned face up and inserted somewhere near the middle of the stack - this will determine who starts the auction. The dealer then draws the top card from the deck, looking at it but not showing it to the other players. The player to his right does the same, then the third player, then the dealer and so on counter-clockwise around the table until each player has 17 cards. The last three cards are left face down on the table until after the auction. It saves time if you look at your cards and sort them as you pick them up.AuctionThere is an auction to determine which player will be the landlord, and play alone against the other two. The possible bids are 1, 2 and 3. The player who drew the face up card in the "deal" is the first to bid. Each player in turn may either pass or bid higher than the highest bid so far. If everyone passes the hand is thrown in and there is a new deal. If there is a bid, the bidding continues counter-clockwise, each player passing or bidding higher than the previous bidder, until there are two consecutive players pass or someone bids 3, which ends the auction since it is the highest possible bid. The final and highest bidder is the landlord. This player now picks up the three face-down cards from the middle, for a total of 20 cards.PlayThe landlord plays first, and may play a single card or any legal combination. Each subsequent player in anticlockwise order must either pass (play no card) or beat the previous play by playing a higher combination of the same number of cards and same type. There are just two exceptions to this: a rocket can beat any combination, and a bomb can beat any combination except a higher bomb or rocket - see definitions below. The play continues around the table for as many circuits as necessary until two consecutive players pass. The played cards are then turned face down and put aside, and the person who played the last card(s) begins again, leading any card or legal combination.In this game, there are thirteen types of combination that can be played:Single card - ranking from three (low) up to red joker (high) as explained abovePair - two cards of the same rank, from three (low) up to two (high)Triplet - three cards of the same rankTriplet with an attached card - a triplet with any single card added, for example 6-6-6-8. These rank according to the rank of the triplet -so for example 9-9-9-3 beats 8-8-8-A.Triplet with an attached pair - a triplet with a pair added, like a full house in poker, the ranking being determined by the rank of the triplet - for example Q-Q-Q-6-6 beats 10-10-10-K-K.Sequence - at least five cards of consecutive rank, from 3 up to ace - for example 8-9-10-J-Q. Twos and jokers cannot be used.Sequence of pairs - at least three pairs of consecutive ranks, from 3 up to ace. Twos and jokers cannot be used. For example 10-10-J-J-Q-Q-K-K.Sequence of triplets - at least two triplets of consecutive ranks from three up to ace. For example 4-4-4-5-5-5.Sequence of triplets with attached cards - an extra card is added to each triplet. For example 7-7-7-8-8-8-3-6. The attached cards must be different from all the triplets and from each other. Although triplets of twos cannot be included, a two or a joker or one of each can be attached, but not both jokers.Sequence of triplets with attached pairs - an extra pair is attached to each triplet. Only the triplets have to be in sequence - forexample 8-8-8-9-9-9-4-4-J-J. The pairs must be different in rank from each other and from all the triplets. Although triplets of twos cannot be included, twos can be attached. Note that attached single cards and attached pairs cannot be mixed - for example 3-3-3-4-4-4-6-7-7 is not valid.Bomb - four cards of the same rank. A bomb can beat everything except a rocket, and a higher ranked bomb can beat a lower ranked one.Rocket - a pair of jokers. It is the highest combination and beats everything else, including bombs.Quadplex set - there are two types: a quad with two single cards of different ranks attached, such as 6-6-6-6-8-9, or a quad with two pairs of different ranks attached, such as J-J-J-J-9-9-Q-Q. Twos and jokers can be attached, but you cannot use both jokers in one quadplex set. Quadplex sets are ranked according to the rank of the quad. Note that a quadplex set can only beat a lower quadplex set of the same type, and cannot beat any other type of combination. Also a quadplex set can be beaten by a bomb made of lower ranked cards.Note that passing does not prevent you from playing on a future turn.Example Player A (the landlord) leads 3-3-3-9 to get rid of some low cards, player B passes, player C plays 5-5-5-7, player A plays K-K-K-J and player B plays A-A-A-3. C and A pass, so B can start again with anything. He leads a single 4.Note B could have played his aces on his the first turn, but preferred to pass to give his partner a chance to get rid of some cards. C will now play if possible, so as not to give the landlord (A) a free chance to lead again. Having beaten A's second play, B leads a low card to give C the choice of playing another unwanted card or putting the landlord under pressure by playing a high card.ScoringIf the landlord runs out of cards first he has won, and each opponent pays him the amount of the bid - 1, 2 or 3 units - provided that no bomb or rocket was played. If one of the other two players runs out before the landlord, the landlord loses and must pay the amount of the bid to each opponent. For each occasion when any player played a bomb or rocket, the payment for the hand is doubled. So for example in a hand in which two bombs and a rocket were played, a player who bid 3 will win 24 points from each opponent for going out first, or pay 24 to each opponent if another player goes out first.Note that since the opponents of the landlord stand to win or lose equally, they form a temporary partnership. When playing against the landlord it is just as profitable to help your partner to run out of cards first as to win yourself. Because of this the partners will usually not beat each other’s cards, and the weaker partner will play to help the stronger partner.Four-Player GameThe four-player form of Fight the Landlord is played mainly in Zhejiang and Jiangsu provinces, including Shanghai. It uses a double deck, including two red and two black jokers - 108 cards altogether. Each player takes 25 cards and 8 cards are left over for the landlord, who plays alone from a hand of 33 cards against the other three players in partnership.The combinations that can be played differ from those in the three-player game, listed above, as follows:Single card attachments are not permitted - i.e. combination types 4 and 9 are exluded.There are no quadplex sets - combination type 13 is excluded.A bomb (type 11) can consist of four or more cards of equal rank, and a bomb with more cards beats a bomb with fewer cardsirrespective of the ranks of the cards.For a rocket (type 12) you need all four jokers.You cannot play a red and a black joker together as a pair, but you can use a pair of red jokers or a pair of black jokers either as a pair by itself or to attach to a triplet.The payment for a hand is doubled for each bomb of 6 or more cards and for each rocket, but bombs of 4 or 5 cards do not affect the payment.。
网络拍卖英文流程Ah, the digital bazaar of dreams and nightmares! Welcome to the world of online auctions, where you can bid on anything from a vintage typewriter to a haunted house (just kidding, we hope). Here’s a step-by-step guide to the wild, wacky, and sometimes worrisome world of online auctions, all in the good ol' American English.Step 1: The Lurking PhaseFirst things first, you gotta do some recon. Browse the listings like a detective, looking for that one-of-a-kind item that’ll make your heart go "ka-ching!" Just remember, what you're looking at is a digital mirage until you see it in person (or at least in a high-res photo).Step 2: The Registration ShuffleBefore you can start bidding, you gotta sign up for the auction site. It's like joining a club, but instead of a secret handshake, you need a credit card and a good credit score. Fill in the blanks, hit submit, and voilà! You're ready to rumble.Step 3: The Bidding BattleNow for the main event! Place your bid with the confidence of a poker player and the stealth of a ninja. Butwatch out, because the auction isn’t just you against the item—it’s you against a bunch of other bidders who are just as desperate to win.Step 4: The Nerve-Wracking WaitAfter you place your bid, it’s time for the waiting game. Keep an eye on the clock, and if you’re feeling brave, maybe even place a higher bid to scare off the competition. Just don’t blink, or you might miss the action!Step 5: The Winning WahooCongratulations, yo u’ve won the auction! Now it’s timeto pay up and wait for your prize to arrive. It’s like Christmas morning, but you already know what you’re getting.Step 6: The AftermathOnce your item arrives, inspect it like a pro. If it’snot what you bargained for, you might have to go through the hassle of returns or refunds. But if it’s everything you dreamed of and more, you can sit back and bask in the gloryof your successful online auction conquest.So there you have it, folks. The thrill of the chase, the agony of defeat, and the ecstasy of victory—all in thecomfort of your own home. Happy bidding, and may the odds be ever in your favor!。
Multi-player and Multi-round Auctions with Severely Bounded CommunicationLiad Blumrosen1,Noam Nisan1,and Ilya Segal21School of Engineering and Computer Science.The Hebrew University of Jerusalem,Jerusalem,Israel.liad,noam@cs.huji.ac.il2Department of Economics,Stanford University,Stanford,CA94305ilya.segal@Abstract.We study auctions in which bidders have severe constraints on the size of messages they are allowed to send to the auctioneer.In such auctions,each bidder has a set of k possible bids(i.e.he can send up to t=log(k)bits to the mechanism).This paper studies the loss of economic efficiency and revenue in such mechanisms,compared with the case of unconstrained communication.For any number of players, we present auctions that incur an efficiency loss and a revenue loss of O(1data(for example,how much they are willing to pay for a certain resource).The auction mechanism must somehow elicit this information from the selfish par-ticipants,in order to achieve global or“social”goals(e.g.maximize the seller’s revenue).Recent results show that auctions are hard to implement in practice. The reasons might be computational(see e.g.[12,8]),communication-related ([13]),uncertainty about timing or participants([4,7],)and many more.This, and the growing usage of auctions in e-commerce(e.g.[14,21,15])and in various computing systems(see e.g.[11,19,20])led researchers to take computational effects into consideration when designing auctions.Much interest was given in the economic literature to the design of optimal auctions and efficient auctions.Optimal auctions are auctions that maximize the seller’s revenue.Efficient auctions maximize the social welfare,i.e.they allocate the resources to the players that want them the most.A positive correlation usually exist between the two measures:a player is willing to pay more for an item that is worth a higher value to her.Nevertheless,efficient auctions are not necessarily optimal,and vice versa.In our model,each player has a private valuation for a single item(i.e.she knows how much she values the item,but this value is a private information for herself).The goal of the auction’s designer (in the Bayesian framework)is,given distributions on the players’valuations, tofind auctions that maximize the expected revenue or the expected welfare, when the players act selfishly.For the single item case,these problems are in fact solved:the Vickrey auction(or the2nd-price auction,see[17])is efficient; Myerson,in a classic paper([9]),fully characterize optimal auctions when the players’valuations are independently distributed.In the same paper,Myerson also shows that Vickrey’s auction(with some reservation price)is also optimal (i.e.revenue maximizing),when the distribution functions hold some regularity property.Optimal auctions and efficient auctions were studied lately also by computer scientists(e.g.[4,16]).Recently,Blumrosen and Nisan([1])initiated the study of auctions with severely bounded communication,i.e.settings where each player can send a message of up to t bits to the mechanism.In other words, each bidder can choose a bid out of a set of k=2t possible bids.The players’valuations,however,can be any real numbers in the range[0,1].Here,we generalize the main results from[1]for multi-player games.We also study the effect of relaxing some of the assumptions made in[1],namely the simultaneous bidding and the independence of the valuations.Severe constraints on the communication are expected in settings where we need to design quick,and cheap auctions that should be performed frequently. For example,if a route for a packet over the Internet is auctioned,we can ded-icate for this purpose only a small number of bits.Otherwise,the network will be congested very quickly.For example,we might want to use some unused bits in existing networking protocols(e.g.IP or TCP)to transfer the bidding infor-mation.This is opposed to the traditional economic approach that views the information sent by the players as real numbers(representing these can take infinite number of bits!).Low communication also serves as a proxy for other desirable properties:with low communication the interface for the auction issimpler(the players have a small number of possible bids to choose from),the information revelation is smaller and only a small number of discrete prices is used.In addition,understanding the tradeoffs between communication and auc-tions’optimality(or efficiency)might help usfind feasible solutions for settings which are currently computationally impossible(combinatorial auctions’design is the most prominent example).Under severe communication restrictions,[1]characterizes optimal and ef-ficient auctions among two players.They prove that the welfare loss and the revenue loss in mechanisms with t-bits messages is mild:for example,with only one bit allowed for each player(i.e.t=1)we can have97percent of the ef-ficiency achieved by auctions that allow the players to send infinite number of bits(with uniform distributions)!Asymptotically,they show that the loss(forboth measures)diminishes exponentially in t(specifically O(1k2)wherek=2t).These upper bounds are tight:for particular distribution functions,the expected welfare loss and the expected revenue loss in any mechanism areΩ(1k2),and we show that for some distributionfunctions(e.g.the uniform distribution)this bound is tight.We also extend the framework to the following settings:–multi-round auctions:By allowing the bidders to send the bits of their messages one bit at a time,in alternating order,we can strictly increase the efficiency of auctions with bounded communication.In such auctions,each player knows what bits where sent by all players up to each stage.However, we show that the same extra gain can be achieved in simultaneous auctions that use less than double amount of communication.–Joint distributions:When the players’valuations are statistically depen-dent,we show that we cannot do better(asymptotically)than a trivial mech-anism that achieves an efficiency loss of O(1k ).–Bounded distribution functions:We know([1])that we cannot con-struct one mechanism that incurs a welfare loss of O(1B1AB wins and pays0133Fig.1.A matrix representation for a mechanism with two possible bids.E.g.,when Alice bids′′1′′and Bob bids′′0′′,Alice wins the item and pays1Definition3A strategy s i for player i in a game g∈G n,k describes how the player determines his bid according to his valuation,i.e.it is a functions i:[0,1]→{0,1,...,k−1}.Denoteϕk={s|s:[0,1]→{0,1,...,k−1}}(i.e.the set of all strategies for players with k possible bids).Definition4A real vector c=(c0,c1,...,c k)is a vector of threshold-values if c0≤c1≤...≤c k.Definition5A strategy s i∈ϕk is a threshold-strategy based on a vector of threshold-values c=(c0,c1,...,c k),if c0=0and c k=1and for every c i≤v i<c i+1we have s i(v i)=i.We say that s i is a threshold strategy, if there exists a vector c of threshold values such that s i is a threshold strategy based on c.We use the notations:s(v)=(s1(v1),...,s n(v n)),when s i is a strategy for bidder i and v=(v1,...,v n).Let s−i denote the strategies of the players except i,i.e. s−i=(s1,...,s i−1,s i+1,...,s n).We sometimes use the notation s=(s i,s−i). 2.1Optimality MeasuresThe players in our model choose strategies that maximize their utilities.We are interested in games with stable behaviour for all players,i.e.such that these strategies form an equilibrium.Definition6Let u i(g,s)be the expected utility of player i from game g when bidders use the strategies s,i.e.u i(g,s)=E v∈[0,1]n(a i(s(v))·(v i−p i(s(v))))Definition7The strategies s=(s1,...,s n)form a Bayesian-Nash equilibrium in a mechanism g∈G n,k,if for every player i,s i is the best response for the strategies s−i of the other players,i.e.∀i∀ s i∈ϕk u i(g,(s i,s−i))≥u i(g,( s i,s−i))Definition8A strategy s i for player i is dominant in mechanism g∈G n,k if regardless of the other players’strategies s−i,i cannot gain a higher utility by changing his strategy,i.e.∀ s i∈ϕk∀s−i u i(g,(s i,s−i))≥u i(g,( s i,s−i)) We say that a mechanism g has a dominant strategies equilibrium if for ev-ery player i there exists a strategy s i which is dominant.Clearly,a dominant strategies equilibrium is also a Bayesian-Nash equilibrium.Each bidder aims to maximize her expected utility.As mechanisms’designers, we aim to optimize“social”criteria such as welfare(efficiency)and revenue. The expected welfare from a mechanism g,when bidders use strategies s,is the expected valuation of the winning players(if any).Definition9Let w(g,s)denote the expected welfare in the n-player game g when bidders’strategies are s,i.e.w(g,s)=E v∈[0,1]n( n i=1a i(s(v))·v i)Definition 10Let r (g,s )denote the expected revenue in the n -player game g when bidders’strategies are s ,i.e.r (g,s )=E v ∈[0,1]n ( n i =1a i (s (v ))·p i (s (v )))Definition 11We say that a mechanism g ∈G n,k achieves an expected welfare (revenue)of αif g has a Bayesian-Nash equilibrium s for which the expected welfare (revenue)is α,i.e.w (g,s )=α(r (g ,s )=α).Definition 12We say that a mechanism g ∈G n,k incurs a welfare loss of c ,if there is a Bayesian-Nash equilibrium s in g such that the difference between w (g,s )and the maximal welfare with unbounded communication is c .We say that g incurs a revenue loss of c ,if there is an individually-rational Bayesian-Nash equilibrium s in g ,such that the difference between r (g,s )and the optimal revenue,achieved in an individually-rational mechanism with Bayesian-Nash equilibrium in the unbounded communication case,is c .Recall that an equilibrium is individually rational,if the expected utility of each player,given his own valuation,is non negative.The mechanism described in Fig.1has a dominant strategy equilibrium that achieves an expected welfare of 353,i.e.she bids “0”when her valuation is below 13is dominant for Bob.We know ([17])thatthe optimal welfare from a 2-player auction with unconstrained communication is 23−3554.3Multi-player MechanismsIn this section,we construct n -player mechanisms with bounded communication which are asymptotically optimal (or efficient).We prove that they incur losses of welfare and revenue of O (1Definition14A game is called a modified priority-game if it has an allo-cation as in priority-games,but no allocation is done when all players bid0. Definition15An n-player priority-game based on a profile of threshold values’vectors−→t=(t1,...,t n)∈×n i=1ℜk+1(where for every i,t i0≤t i1≤...≤t i k)is a mechanism that its allocation is as in a priority game and its payment scheme is as follows:when player j wins the item for the bids vector b she pays the smallest valuation she might have and still win the item,given that she uses the threshold strategy s j based on t j.I.e.p j(b)=min{v j|a j(s j(v j),b−j)=1}.We denote this mechanism as P G k(−→t).A modified priority game with a similar payment rule is called a modified priority-game based on a profile of threshold values’vectors, and is denoted by MP G k(−→t).For example,Fig.1describes a priority game based on the threshold values(0,13,1).When Bob bids0,the minimal valuation of Alice for whichshe still wins is1k2).In[1],a similarupper bound was given for the case of2-player mechanisms:Theorem1[1]For every set of distribution functions on the players’valua-tions,the2player mechanism P G k(x,y)incurs an expected welfare loss of O(1k2)in any mechanism.Here,we prove that n-player priority games are asymptotically efficient: Theorem2For any number of players n,and for any set of distribution func-tions of the players’valuations,the mechanism P G k(−→t)incurs a welfare loss of O(1k2:Theorem3When valuations are distributed uniformly,and for any(fixed) number of players n,any mechanism g∈G n,k incurs a welfare loss ofΩ(1Proof.Consider only the case where players1and2have valuations greaterthan12.This occurs withthe constant probability of1k2)will still be incurred(the fact that in theorem1the valuations’range is[0,1]and here it is[1k2).3.2Asymptotically Optimal MechanismsNow,we present mechanisms that achieve asymptotically optimal expected rev-enue.We show how to construct such mechanisms and give tight upper bounds for the revenue loss they incur.Most results in the economic literature on revenue-maximizing auctions,as-sume that the distribution functions of the players’valuations holds a regularity property(as defined by Myerson[9],see below).For example,only when the val-uations of all players are distributed with the same regular distribution-function, it is known that Vickrey’s2nd-price auction,with an appropriately chosen reser-vation price,is revenue-optimal([17,9,3]).Definition16([9])Let f be a density function,and let F be its cumulative function.We say that f is regular,if the function v(v)=v−1−F(v)every regular distribution,there is a mechanism that incurs a revenue loss of O (1k 2.Theorem 5Assume that all valuations are distributed with the same regular distribution function.Then,for any number of players n ,MP G k (−→t )incurs a revenue loss of O (1k 2).4Bounded Distributions and Joint DistributionsIn previous theorems,we showed how to construct mechanisms with asymptot-ically optimal welfare and revenue,given a set of distribution functions.Can we design a particular mechanism that achieve similar results for all distribu-tion functions?Due to [1],the answer in general is no.The simple mechanism P G k (x,x )where x =(0,1k ,...,k −1k )and no better upper bound can be achieved.Nevertheless,we show that if the distri-bution functions are bounded from above or from below,this trivial mechanism for two players achieves an expected welfare which is asymptotically optimal.Definition 17We say that a density function f is bounded from above (below)if for every x in its domain,f (x )≤c (f (x )≥c ),for some constant c .Proposition 3For every pair of distribution functions of the players’valu-ations which are bounded from above,the mechanism P G k (x,x ),where x =(0,1k ,...,k −1k 2).For every pair of distribution functions which are bounded from below,every mechanism incurs an expected welfare loss of Ω(1k )for the efficiency loss in 2-player games.Theorem 7The mechanism P G k (x,x )where x =(0,1k ,...,k −1k for any joint distribution φon the players’valuations.Moreover,for every k there is a joint distribution function φk such that every mechanism g ∈G 2,k incurs a welfare loss ≥c ·1B1AA,04134Fig.2.(h1)This sequential game(when A bidsfirst,then B)achieves higher expected welfare than any simultaneous mechanism with the same communication complexity (2bits).The welfare is achieved with Bayesian-Nash equilibrium.5Multi-round AuctionsIn previous sections,we analyzed auctions with bounded communication in which players simultaneously send their bids to the mechanism.Can we get better results with multi-round(or sequential)mechanisms?I.e.mechanisms in which players send their bids one bit at a time,in alternating order.In this sec-tion,we show that sequential mechanisms can achieve better results.However, the additional gain(in the amount of communication)is up to a factor of2.5.1Sequential Mechanisms Can Do BetterThe definitions in this section are similar in spirit to the model described in section2.For simplicity,we present this model less formally.Definition18A sequential(or multi-round)mechanism is a mechanism in which players send their bids one bit at a time,in alternating order.In each stage,each player knows the bits the other players sent so far.Only after all the bits were transmitted,the mechanism determines the allocation and payments. Definition19The communication complexity of a mechanism is the total amount of bits which are sent by the players.Definition20A strategy for a player in a sequential mechanism is the way she determines the bits she transmits,at every stage,given her valuation and given the other players’bits up to this stage.A strategy for a player in a sequential mechanism is called a threshold strategy if in each stage i of the game,the player determines the bit she sends according to some threshold value x i;I.e.if her valuation is smaller than this threshold she bids0,or bids1otherwise.Denote the following sequential mechanism by h1(see Fig.2):Alice sends one bit to the mechanismfirst.Bob,knowing Alice’s bid,also sends one bit.When Alice bids0:Bob wins if he bids1and pays1;If he bids zero,Alice wins again,but now she pays14Proposition4When valuations are distributed uniformly,the mechanism h1 above has a Bayesian-Nash equilibrium and an expected welfare of0.653. Proof.Consider the following strategies:Alice uses a threshold strategy basedon the threshold value14when Alice bids“0”andthe threshold3s=(s n)that achieves at least the same welfare with h as s does.Theorem8Let h be a2-player sequential mechanism with communication com-plexity m.Then,there exists a simultaneous mechanism g that achieves at least the same expected welfare as h,with communication complexity of2m−1.Proof.Consider a2-player,sequential mechanism h with a Bayesian-Nash equi-librium,and with communication complexity m(we assume m is even,i.e.each player sends mlog(αA(m)+1)+log(αA(m)+1)<log(2m−1)+log(2m)=2m−1In the full paper([2])we show that the new strategies forms an equilibrium. Acknowledgments.The work of thefirst two authors was supported by a grant from the Israeli Academy of Sciences.The third author was supported by the National Science Foundation.References1.Liad Blumrosen and Noam Nisan.Auctions with severely bounded communica-tions.In43th FOCS,2002.2.Liad Blumrosen,Noam Nisan,and Ilya Segal.Multi-player and multi-round auc-tions with severely bounded communications,2003.Full version.available from http://www.cs.huji.ac.il/˜liad.3.Riley J.G.and Samuelson W.F.Optimal auctions.American Economic Review,pages381–392,1981.4.Andrew V.Goldberg,Jason D.Hartline,and Andrew petitive auc-tions and digital goods.In Symposium on Discrete Algorithms,pages735–744, 2001.5.Papadimitriou C.H.Algorithms,games,and the internet.ACM Symposium onTheory of Computing,2001.6.Paul Klemperer.The economic theory of auctions.Edward Elgar Publishing,2000.7.Ron Lavi and Noam petitive analysis of incentive compatible on-lineauctions.In ACM Conference on Electronic Commerce,pages233–241,2000.8.Daniel Lehmann,Liadan Ita O’Callaghan,and Yoav Shoham.Truth revelation inrapid,approximately efficient combinatorial auctions.In1st ACM conference on electronic commerce,1999.9.R.B.Myerson.Optimal auction design.Mathematical of operational research,pages58–73,1981.10.Noam Nisan.Algorithms for selfish agents.In STACS,1999.11.Noam Nisan,Shmulik London,Ori Regev,and Noam Camiel.Globally distributedcomputation over the internet–the popcorn project.In ICDCS,1998.12.Noam Nisan and Amir Ronen.Algorithmic mechanism design.In STOC,1999.13.Noam Nisan and Ilya munication requirements of efficiencyand supporting lindahl prices,2003.working paper available from http://www.cs.huji.ac.il/˜noam/mkts.html.14.Web page.ebay..15.Web page.vertical-net..16.Amir Ronen and Amin Saberi.Optimal auctions are hard.In43th FOCS,2002.17.W.Vickrey.Counterspeculation,auctions and competitive sealed tenders.Journalof Finance,pages8–37,1961.18.Rakesh Vohra and Sven de binatorial auctions:A survey,2000.Availail-abe from /faculty/vohra/htm/res.htm.19. C.A.Waldspurger,T.Hogg,B.A.Huberman,J.O.Kephart,and W.S.Stornetta.Spawn:A distributed computational economy.IEEE Transactions on Software Engineering,18(2),1992.20.W.E.Walsh,M.P.Wellman,P.R.Wurman,and J.K.MacKie-Mason.Auction pro-tocols for decentralized scheduling.In Proceedings of The Eighteenth International Conference on Distributed Computing Systems(ICDCS-98),1998.merce-one..。