Reply #15: Here are my latest simulation results. [View All]

[eomer]

First simulation:

Simulation based on a range of turnout percentages from 33.62 to 74.63 where the turnout is assumed to follow a uniform distribution. Average number of adjacent matches = 8.9 out of 3800.

Probability of at least 14 matches if the rate is 8.9 out of 3800 = 1-BINOMDIST(13,3800,8.9/3800,TRUE) = 6.9%.

Second simulation:

Simulation based on turnout assumed to follow a normal distribution with mean of 54.12 and standard deviation of 10. Average number of adjacent matches = 10.94 out of 3800.

Probability of at least 14 matches if the rate is 10.94 out of 3800 = 1-BINOMDIST(13,3800,10.94/3800,TRUE) = 21.3%.

These results show that the question of whether the data follows a uniform distribution, a normal distribution or some other distribution has a significant effect on the answer.

My results seem to fall into the DMZ where they could probably be taken to support either side of the argument. I'll just let everyone draw their own conclusions.