Controls to Make Data Analysis Blind

We want to measure how likely this is to be a random fluctuation. To do that, we need to make sure not to twiddle the knobs. We add a big random number to every collision so that we don't actually know what they mean. So it's almost like you corrupt the data on purpose,. But then at the end, you take out that plant that see the corruption. Yes, like, we're working with incryptit data, and then we type in the password, and it all becomes clear at the end.