User blog comment:Felza/40% Troll vs 50% Troll/@comment-5266525-20131029153836

You ultimately end up failing to fulfill the goal you presumably set out to achieve with the blog.

The 44%/56% claim is completely true and mathematically provable. Sample size is totally irrelevant - it simply states that in any given game the enemy team will always have a higher chance of having a troll. Throwing a list counterexamples at it in an attempt to disprove it is unsound logic unless that list is exhaustive.

You've, somehow, completely misunderstood it to mean that for any given game the enemy team WILL always have more trolls.

It's like hearing that the chance of flipping a coin and getting heads is 50% and then flipping two heads in a row and immediately calling the theory bullshit. That's basically what you are doing.