90% sucess rate and still missing

[Dakre]

OP is just frustrated, so I can understand it because I've felt the same. But if you think it isn't working correctly and they think it's fine, how do you prove a random chance is not so random? The other question is, if it happens that often, how many times an hour/day does it happen?

The reason behind those questions is to clarify the difficulty of fixing it, because if you take a random number generator and get similar results, you proved yourself wrong. If you get different results with the same number of tries, then you still have to prove their system isn't a 90% hit rate. Otherwise, people don't care as much without proof, cause ideas are a dime a dozen.

[Conradus]

OP is just frustrated, so I can understand it because I've felt the same. But if you think it isn't working correctly and they think it's fine, how do you prove a random chance is not so random?

By accurately recording all results as they happen, getting a statistically significant sample (you'll want thousands of results, at a minimum), and subjecting your data to a rigorous statistical analysis.