APPENDIX 3Why Dynamic?
I presented at a gaming summit with David Koch, who was then data scientist at a casino in Canada. Our session was about evolving analytical capabilities in casinos. I set up the discussion, much like I have done in this book, by describing the characteristics of an analytically driven organization and providing a roadmap to achieve that vision. David followed, with findings from a database for slot performance management he developed.
Much can be learned by examining a game’s history. The database exhausted all years of available production data but pared it to existing assets (machines) only. Asset attributes captured in production were cleaned and new ones introduced for better product differentiation. Capable of regeneration from production in moments, a rich, up-to-date historical database was always available. Not only the basis for forecasting, these game histories would prove insightful when viewed graphically against time.
Under the theme of the importance of asking how and why, following are findings David presented at the summit. For the purposes of this book, he has re-created here relevant portions of the presentation, with further discussion.
Context comes from a long-term perspective. Play is not constant. There are cycles.
For this first game (Figure A.1), season of the year was a determining factor in its betting activity. Note the recurring January (and sometimes July) lows.
At the same time, broader changes may ...