A Game Theory Approach for Data Driven Business Decisions
summary
Builds a virtual market from customer histories plus declared preferences, estimates attribute-level utilities to translate choices into euros, and runs Monte Carlo game-theory simulations over three competitors and five tiers. Single-player tweaks converge quickly (~300–400 iterations) and show up to 25% profit uplift while rivals lose ~17%; allowing all players to react produces cyclic equilibria and a simulated 15–30% profit drop and suggests removing underperforming mid tiers and concentrating on tiers 4–5.
AI-written summary · about
chapters (17)
- 0:00Objective of this presentation
- 0:47Telefonica Key Figures
- 8:34What do customers want?
- 9:17What a Virtual Customer is?
- 11:13And attributes are divided into le
- 14:17The "Utility" is calculated
- 17:15What can we offer? Easy!
- 18:28What does optimization mean
- 19:03We introduce a Game Theory Vis
- 21:11Classical Prisoner Dilemma
- 22:06Game Theory: Nash Equilibrium
- 22:59Game Theory: Pareto Optimal P
- 24:21Rock Scissors Paper (RSP) Ga
- 27:08Optimization: Player A Profit incre
- 27:52Player A: Tiers' Market Share
- 30:47Several players: Price war 3000
- 33:38What we saw there
related talks
25mData-Driven Game-dev Business: Building a Profitable Portfolio
54m3 Ways To Understand Your Game's Audience
27mHow To Make A Successful Business Out Of Your Game
27mUnlocking Product Growth With Big Data, Data Science And Machine Learning
34mOptimizing E-commerce: Data Science for Better CX & Sales
28mThe Value of UX Applied to Company Data
42mAugmenting Human Decision Making with Data Science
58mPremium Steam Game Biz Dev Evaluation – A Proper Execution