After shuffling inputs and outputs, he found that 42% of transactions were bip69 when randomized. He believes that although bip69 is used, it is not as popular as randomization. Ryan also mentioned that the false positive rate for bip69 usage can be calculated using math, which he attempted to do with his limited math skills. On the other hand, Gregory Maxwell commented that bip69 follows around 60% of transactions, but according to Ryan's findings, bip69 is only followed by a minority of transactions. A two input randomly ordered transaction has a 50% chance of following bip-69.