What's particularly interesting about the data is that the Dot attempted to eliminate as many variables as possible from its numbers, and even found data that would logically suggest blue should be losing more games. For example, the red team killed more Dragons than the blue team, and often killed Dragons before the blue team. This should give the red team a boost to their gold, allowing them to buy items and become stronger faster than their opponents. Yet clearly this didn't happen, as blue team won more often against teams ranked lower than them and against teams ranked higher than them.
There's no solid answer as to why this is happening. And while we could spend all day looking at bar graphs, it's important to remember that correlation does not imply causation. Just because it looks like the rain in Howard County, Missouri causes people to get hit by a railway train doesn't mean that's what actually happens. Likewise, there's probably more to this data than a first glance would suggest.
Or, maybe blue is just a more powerful color, like how the Green Lantern used to be weak to the color yellow. Who knows!