We ran our program a million times on each contest in the 2012 NSW local government elections. The results are shown here in four tables:

• All contests
• Only those contests that had different results (who or order) in at least one run
• Only those contests that had different results (who was elected) in at least one run
• Only those contests where the most likely outcome differed from the official results (who or order)

Each contest is shown with a table of the elected people, for each showing the proportion of the time they are elected, the mean position that they were elected in amongst the times they were elected, and the official position elected.

This is meant to give an idea of the level of randomness that the election counting algorithm produces. We are not claiming that we understand the legislation properly (it is somewhat ambiguous), nor the NSWEC specification (it is somewhat ambiguous and, we think, self-contradictory - we try to follow the text, not the pseudocode), nor that our program is without errors. So the results may not be correct. We do not warranty them in any way. You should avoid trusting programs when you can, including ours. But on the bright side the results are somewhat neatly presented. Our program is available for you to look at, you can run it yourself, and, best of all, you can write your own program and check it against ours.