A Prototype Prediction Market Game for Enhancing Knowledge Sharing among Salespeople
1 like804 views
This document proposes using a prediction market game with a comment system to enhance knowledge sharing among salespeople. It describes preliminary experiments comparing prediction market sessions with and without the comment system. The sessions with comments had significantly more transactions, suggesting comments activated more frequent information and knowledge sharing. Further experiments are needed to better evaluate the game and identify improvements, such as capturing comment relationships. The goal is to develop a serious gaming approach that motivates truthful information sharing to foster collaboration.
A Prototype Prediction Market Game for Enhancing Knowledge Sharing among Salespeople
- 1. Hajime Mizuyama A Prototype Prediction Market Game for Enhancing Knowledge Sharing among Salespeople Hajime Mizuyama & Kazuto Yamamiya Aoyama Gakuin University mizuyama@ise.aoyama.ac.jp ISAGA 2014 7/July/2014
- 2. Hajime Mizuyama Competitive atmosphere • In a company running a BtoB business, such as IT system integration, orders are usually obtained through negotiations with customer companies. • The mission of its sales division is – to keep winning an appropriate number of orders through negotiations – to accurately forecast the total number of orders to be achieved Research background Information and knowledge sharing in the division Information and knowledge sharing in the division
- 3. Hajime Mizuyama • In a company running a BtoB business, such as IT system integration, orders are usually obtained through negotiations with customer companies. • The mission of its sales division is – to keep winning an appropriate number of orders through negotiations – to accurately forecast the total number of orders to be achieved – to foster collaborative atmosphere in the division Research background Information and knowledge sharing in the division Collaborative atmosphere
- 4. Hajime Mizuyama The objective of this research is to develop a serious gaming approach using prediction markets for motivating truthful information and knowledge sharing and thereby fostering collaborative environments in a sales division. • Reward the members of the division for providing valuable information or knowledge truthfully. • Properly forecast the probability of winning a sales negotiation based on the information provided. • Improve/control the probability of winning a sales negotiation based on the knowledge provided. Research objective
- 5. Hajime Mizuyama • Research background and objective • Problem setting • What are prediction markets? • Proposed game design • Preliminary experiments and their results • Conclusions Agenda
- 6. Hajime Mizuyama Problem setting Sales division Sales negotiation 1 Sales negotiation 3 Sales negotiation 2 Sales negotiation 4
- 7. Hajime Mizuyama Forecast accurately the probability of winning the negotiation. Improve/control the probability of winning the negotiation. Collaborative information/knowledge acquisition and sharing Problem setting Sales negotiation 1 Owner Other participants
- 8. Hajime Mizuyama • Research background and objective • Problem setting • What are prediction markets? • Proposed game design • Preliminary experiments and their results • Conclusions Agenda
- 9. Hajime Mizuyama An ordinal prediction market setting Central Market Maker (LMSR) Market prices = Probabilities of being elected Candidate A Candidate B Candidate C WTA securities A fixed posterior payoff only for the elected candidate Bid and ask orders
- 10. Hajime Mizuyama Automated market maker LMSR by R. Hanson (2003) N i i bQbC 1 /explog)(Q )()(Cost QqQ CC N j j i i i bQ bQ dQ dC p 1 /exp /exp)(Q Numberofsec.：Q Cost function How much to charge for buying Δq Unitary price of ith security Whole numbers sold so far Unitaryprice：p Candidates Candidates
- 11. Hajime Mizuyama • Research background and objective • Problem setting • What are prediction markets? • Proposed game design • Preliminary experiments and their results • Conclusions Agenda
- 12. Hajime Mizuyama Outline of proposed system Central Market Maker (LMSR) Bid/ask orders Market prices = Probabilities of whether it will be won or lost Win Lose Win/Lose securities A fixed posterior payoff is given only for that corresponding to the actual outcome. Owner Other participants price
- 13. Hajime Mizuyama Outline of proposed system Central Market Maker (LMSR) Bid/ask orders & Comments Market prices = Probabilities of whether it will be won or lost Comments = Grounds for the prediction Win Lose Win/Lose securities A fixed posterior payoff is given only for that corresponding to the actual outcome. Highly evaluated comments are also rewarded. Owner Other participants comment comment comment comment comment comment comment comment Provide basic information. Set initial prices for the securities. Answer questions (Q&A function to be introduced). price
- 14. Hajime Mizuyama Collection and evaluation of comments Win Lose It will be effective to change the sales strategy such that… It will be effective to modify the proposal such that … The proposal does not seem competitive because … The competitor A has a stronger connection with the customer. Buy shares Buy shares
- 15. Hajime Mizuyama Collection and evaluation of comments Win Lose It will be effective to change the sales strategy such that… It will be effective to modify the proposal such that … The customer will accept our offer because … The proposal does not seem competitive because … The competitor A has a stronger connection with the customer. Buy 3 shares Buy shares
- 16. Hajime Mizuyama Collection and evaluation of comments Win Lose It will be effective to change the sales strategy such that… It will be effective to modify the proposal such that … The customer will accept our offer because … The proposal does not seem competitive because … The competitor A has a stronger connection with the customer. Buy shares Buy shares ✔ 2
- 17. Hajime Mizuyama Collection and evaluation of comments Win Lose It will be effective to modify the proposal such that … It will be effective to change the sales strategy such that… The customer will accept our offer because … The proposal does not seem competitive because … The competitor A has a stronger connection with the customer. Buy shares Buy shares
- 18. Hajime Mizuyama Collection and evaluation of comments Win Lose UncontrollableControllable A competitor has decided to withdraw their proposal. If the owner does this, the probability will increase. A new company had decided to enter the competition. If the owner does not do this, the probability will remain low.
- 19. Hajime Mizuyama Posterior wealth = Payoff from owned security − Cost of buying security + Revenue from selling security Comment contribution Game scores = evaluation counts of comment 𝑐𝑜𝑚𝑚𝑒𝑛𝑡𝑠 𝑒𝑛𝑡𝑒𝑟𝑑 𝑏𝑦 𝑡ℎ𝑒 𝑝𝑙𝑎𝑦𝑒𝑟
- 20. Hajime Mizuyama • Research background and objective • Problem setting • What are prediction markets? • Proposed game design • Preliminary experiments and their results • Conclusions Agenda
- 21. Hajime Mizuyama • Participants were 6 undergraduate students in Aoyama Gakuin University. • Only fixed participation fee, no additional monetary incentive depending on the game score, is provided. • Simple either-or problems: – Red-and-white-ball setting – Imaginary sales negotiation scenario • Before a game session started, the players were secretly provided different information. Then, they enjoyed the session for about 15 min. • Market game sessions with and without the comment system are compared. Experimental setting
- 22. Hajime Mizuyama Red-and-white-ball setting 60 70 80 90 100 110 120 130 140 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Number of transactions 赤証券 白証券 60 70 80 90 100 110 120 130 140 1 4 7 10 13 16 19 22 25 28 31 34 37 40 Number of transactions 赤証券 白証券 WithoutcommentsystemWithcommentsystem PricePrice Red White Red White 60 70 80 90 100 110 120 130 140 1 4 7 101316192225283134374043 Number of transactions 赤証券 白証券 60 70 80 90 100 110 120 130 140 1 2 3 4 5 6 7 8 9 10 11 12 13 Number of transactions 赤証券 白証券 PricePrice Red White Red White 60 70 80 90 100 110 120 130 140 1 4 7 10 13 16 19 22 25 28 31 34 Number of transactions 赤証券 白証券 60 70 80 90 100 110 120 130 140 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Number of transactions 赤証券 白証券 PricePrice Red White Red White
- 23. Hajime Mizuyama Imaginary sales negotiation scenario 60 70 80 90 100 110 120 130 140 1 3 5 7 9 111315171921232527293133353739414345 Number of transactions 60 70 80 90 100 110 120 130 140 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Number of transactions With comment systemWithout comment system Win Lose Win Lose Price Price
- 24. Hajime Mizuyama Without comment system With comment system Red-and-white-ball #1 14 42 Red-and-white-ball #2 13 43 Red-and-white-ball #3 14 35 Imaginary sales negotiation scenario 14 46 Average 13.75 41.50 Number of transactions in each game session
- 25. Hajime Mizuyama • Research background and objective • Problem setting • What are prediction markets? • Proposed game design • Preliminary experiments and their results • Conclusions Agenda
- 26. Hajime Mizuyama • A prediction market game with a comment system is proposed for enhancing information and knowledge sharing among the salespeople of a company. • A simple prototype of the game is developed and preliminary laboratory experiments using it found that the comment system activates transactions and hence the frequencies of sharing information and knowledge. • Further experiments are necessary for properly evaluating the game and identifying how to improve it. • It is also an important challenge how to capture the relationships among the comments, for example, by introducing gIBIS structure. Conclusions
- 27. Thank you for your kind attention! Questions and comments are welcome.