Bias

A bias is a systematic deviation that pulls a result away from what chance would normally produce. It is not a one-off error but a recurring tendency: some results are favored and others disadvantaged, again and again, in the same direction. It is this regularity that defines bias and makes it dangerous for the fairness of a draw.

Concrete images are plentiful. A loaded die lands too often on the same face; an unbalanced wheel stops more readily on certain sectors; a poorly shuffled list leaves the first items at the top in a predictable way. In each case, a hidden cause quietly distorts the game without it being noticed at first glance.

The decisive distinction sets bias against simple fluctuation. Chance, by nature, almost never produces perfectly regular results: over a handful of flips of a fair coin, getting a clear majority of heads is nothing abnormal. This passing imbalance fades as trials accumulate. Bias, by contrast, persists: it does not disappear with the number of trials, but instead is confirmed by them.

This is precisely where the most common error of judgment lies. People cry foul at a surprising run of a few draws, when a short-lived coincidence proves nothing. Conversely, one should be concerned when a deviation holds up over a very large number of trials: a coin that lands on heads about seven times out of ten after thousands of flips reveals a real bias, not a whim of chance.

On the site, tracking down bias is a constant requirement. A poorly designed list shuffle, a generator that would favor certain values, or unbalanced team formation would betray their mission. Identifying and then eliminating these systematic deviations is what guarantees that a draw stays fair and that each item keeps its rightful chance.