View Single Post
  #16 (permalink)  
Old 16th May 2008, 05:40 PM
Jufo Jufo is offline
Senior Member
 
Join Date: Sep 2007
Location: Finland
Posts: 159
WTGs: 0
WTGd at 1 Time in 1 Post
Thanks: 13
Thanked 66 Times in 40 Posts
Reputation Points: 340
Rep Power: 5
Jufo is a jewel in the roughJufo is a jewel in the roughJufo is a jewel in the roughJufo is a jewel in the rough
Player community auditing

Quote:
Originally Posted by vinylweatherman View Post
Rusty,
Rather than requesting a BS "independent audit" which will show you got as close as expected to 95% given the number of spins you played, request your playlogs in machinable format, such as Excel, so that you can do your OWN analyses on whatever statistic takes your fancy. If MG have nothing to hide, they would have no problem with this.
I would take this one step further and let the player community do the audit themselves. Why rely on some independent audit when you can collectively gather data, combine it into one huge play session and then analyse that? So it would be an Open Source audit conducted by the players themselves. For such a huge sample any statistical bias in the probabilities or streaks would be revealed immediately. Just recall the English Harbour case and the big role of player community in solving that.

I have a done a model template to test game fairness and natural streakiness in Blackjack but the same model can be used in any other game. The model calculates actual number of streaks, compares it to expected values and determines whether the observations match expectations closely enough. You can try out the model at

http://koti.mbnet.fi/sunu/streak_simulation.zip

Right now the model just calculates frequencies from random data. What is needed is to replace this random data with real playing data.
Reply With Quote
The Following User Says Thank You to Jufo For This Useful Post:
Rusty (16th May 2008)