by Matthew de Marte – August 8th, 2018
Last month, SOTG introduced its very own power rankings. The power rankings are derived from an algorithm that you can read about HERE – The goal of the rankings were simple: Create a ranking system that does a better job of predicting future success than current methods available to the public. While creating the rankings there were a few metrics I had to establish: weighted Run Differential (wRD) and wRD Win %. In our first edition, the top five teams in baseball were the Astros, Red Sox, Yankees, Cubs, and Braves, respectively. If you need to catch up on how the rankings were calculated, I suggest reading the first power rankings. Let’s take a look at the SOTG MLB Power Rankings entering play on August 8th. Team rankings in the previous edition of the ranking system are in parenthesis below:
|Rank||Team||wRD W-L||wRD W%||Home wRD||Away wRD||W-L||Win %||Luck|
Let’s break down this month’s rankings! I was surprised to see the Astros still clinging to the top spot over the Red Sox. Although there is not much separating them at the top, I expected the Red Sox to take over the top spot, especially sweeping the Yankees this past week. The biggest movers in this edition of the wRD Win % rankings were both from the AL West. The red hot Oakland A’s, who have gone 21-8 since the last rankings, climbed seven spots to enter the top 10, while the Mariners whose wRD Win %has never been very high in the first place fell seven spots to #20. The top 5 remained pretty much the same. The Braves fell out after their recent struggles, while Manny Machado and the Dodgers rose to the number three ranking, surpassing the Cubs as the top team in the National League.
Now we will take a look at how the predictive power of wRD Win % fared since the last rankings.To analyze this, we will look at Win %, Pythagorean Win %, and wRD Win % for the July 2nd power rankings and compare them to teams Win % since then. This will give us a good idea of which measure was the best at predicting team performance over the last month.
The following plot compares teams July 2nd Win % to their Win % from July 2nd through games played on August 7th:
Correlation = .6072219
The following plot compares wRD Win % on July 2nd and teams Win % from July 2nd through games played August 7th:
Correlation = 0.6771952
Unfortunately, this time I was unable to use Pythagorean Win % in this comparison, but I will be able to use it next time. wRD Win % had a correlation of 67.7% to future Win %, while Win % itself had just 60.7% correlation. Meaning wRD Win % was a better indicator of future performance than a teams Win %! I hope you all enjoyed this edition of the SOTG power rankings. As always, reach out to firstname.lastname@example.org if you have any questions!