Restoration Shaman Raiding Talents

• LB_CI: Lower bound of a 95% confidence interval, calculated on the full dataset. This metric is chosen because it provides a conservative estimate of player performance, accounting for variability in the data.
• MEAN: Calculated by summing all values and dividing by N, representing the average performance metric across all players.
• MAX: Maximum value of the metric, indicating the highest performance achieved in the dataset.
• N: Number of players in the dataset, providing the sample size for statistical calculations.

Tiers are formed by distributing specs using the Ckmeans algorithm, which is chosen for its optimal one-dimensional clustering capabilities. Specs are binned in clusters (tiers) based on the LB_CI value.

This algorithm is particularly effective for datasets where the goal is to create distinct, non-overlapping groups based on a single variable, in this case, the LB_CI value.

The algorithm ensures optimal and deterministic clustering results, meaning tier boundaries remain consistent when run multiple times with the same data. The number of clusters is 5: S, A, B, C, D, with each tier representing a different level of performance.