KimSS,
Alternatively, let me describe the test you could run on your playcheck data to check whether or not an algorithm like the one you describe has been used.
You suggest that a casino would influence when wins are awarded. That implies that the spacing of the wins after the 'tampering' is no longer uniform. The best way to determine if that is the case is by measuring the distribution of the spacing of the wins. One way to do this that results in a graph that you can 'visually' check is the following :
1. express all wins in multiples of the betsize.
2. create categories for each ocuring winsize (2xbet, 3xbet etc)
3. for each category calculate the average nr of spins between two wins in that category
4. for each category construct a distribution graph, with #occurences on the y-axis and on the x-axis the nr of spins.
All the graphs constructed in step 4 should be bell curves around the average from step 3, when provided with sufficient data. (you can make less detailed categories by grouping if you don't have enough data). If your theory is correct you will find win categories (the higher ones) that would show anomalies there.
Cheers,
Enzo
Alternatively, let me describe the test you could run on your playcheck data to check whether or not an algorithm like the one you describe has been used.
You suggest that a casino would influence when wins are awarded. That implies that the spacing of the wins after the 'tampering' is no longer uniform. The best way to determine if that is the case is by measuring the distribution of the spacing of the wins. One way to do this that results in a graph that you can 'visually' check is the following :
1. express all wins in multiples of the betsize.
2. create categories for each ocuring winsize (2xbet, 3xbet etc)
3. for each category calculate the average nr of spins between two wins in that category
4. for each category construct a distribution graph, with #occurences on the y-axis and on the x-axis the nr of spins.
All the graphs constructed in step 4 should be bell curves around the average from step 3, when provided with sufficient data. (you can make less detailed categories by grouping if you don't have enough data). If your theory is correct you will find win categories (the higher ones) that would show anomalies there.
Cheers,
Enzo