Top 5 nerdiest concepts in baseball statistics:

5) Replacement level theory.
The concept that baseball talent has a normal distribution-- in the entire population of the world. The major leagues represents the very tail end of that distribution.

4) Marginal cost/win.
A team's payroll efficiency determined by the formula
(Payroll - 25*Major League Minimum)/(Wins-54)
25*Major League Minimum is the least that a team could conceivably spend on payroll, and 54 is the number of wins for a team of replacement players.

3) Defense Independent Pitching Statistics (DIPS).
Has to do with the theory that a pitcher can only control outcomes that are independent of defense-- homeruns, walks, and strikeouts.

2) Pitcher Empirical Comparison and Optimization Test Algorithm (PECOTA).
Projection system based on comparisons with similar players in baseball history. Based on past production, usage, and phenotypic attributes.

1) The Pythagorean Thereom of baseball.
A team's expected record can be calculated by the formula
(Runs scored)^2/[(Runs scored)^2+(Runs allowed)^2]