MLB Iterative Power Rankings

The Iterative Power Rankings are a way of measuring a baseball team's strength, factoring in their basic offensive and pitching statistics as well as the strength of their opponents.

The IPR method is built on David Smyth's Base Runs framework. Even more than runs scored and allowed, Base Runs for and against give a true picture of a team's quality of play because the Base Runs formula accounts for each individual offensive action by a team and its opponent, as well as balancing against the luck involved in "bunching" or "scattering" hits in any given inning. Using BsR for and against and the Pythagenpat formula, we can come up with a "Base Winning %" (RawBsW%) for each team that is a truer reflection of their play than their real-life or even Pythagorean winning %.

However, you also must adjust for each team's opponents. To do that, we postulate that the log5 method holds true in predicting what a team's Win% vs. a .500 opponent would be. In the IPR system, Power = (OppPower*RawBsW%)/((2*RawBsW%*OppPower)-OppPower-RawBsW%+1). As you can see, this creates an iterative loop, and the computer cycles through the calculations hundreds of times until each team's rating stabilizes. The result is Base Winning %, but also adjusted for the Base Winning % of each team's opponents, and their opponents, and so on and so forth.

Check back in every week to see the latest version of the IPR. As the season goes on and each team plays more unique opponents, the sample size will increase and, consequently, so will the predictive accuracy of the ratings.