The analytics team of the Pittsburgh Pirates led by Dan Fox leaves no stone unturned in searching for ways to maximize the performance of Pirate players. The most recent stone they turned over belongs to the Golden State Warriors of the NBA. As reported by Jayson Stark, the Pirates have been looking closely at the success of the Warriors. The Warriors have the best record in the NBA this season despite decreasing the minutes played by their core players. The Warriors have seemingly found a sweet spot where the number of minutes of rest in a game for their core players optimizes the collective performance of the team. Interesting stuff for sure, but how can minutes played per an NBA game translate to baseball? There is no player reentry in an MLB game. A player can not sub out for a few innings and then reenter to hit in the 9th inning. Rest for baseball players has always been thought to have value as a cumulative effect. Taking a day or two off in April keeps a player fresher for the September stretch run, or at least that is the theory. It is hard to measure if that is truly the case. But perhaps resting a player has a more immediate effect in baseball just like it does in the NBA.
If there is a more immediate benefit to resting a player in baseball it seems like the most logical place to look for it would be in games that followed an off day. So that is what I did. I looked at the the game logs over the last two seasons for the Pirates’ position players that started more than half of the games in both the 2013 and 2014 seasons. There were five players that fit the criteria: Andrew McCutchen, Neil Walker, Starling Marte, Pedro Alvarez, and Jordy Mercer. From the game logs of these players I extracted the games that followed a rest day and calculated their OPS for those games. There were two types of rest days I considered:
- Scheduled team off days and rain outs
- Games in which the player did not start and played 4 innings or less
Here are the results spanning the 2013 and 2014 seasons:
|Games Following Rest (GFR)||OPS in GFR|
So now the question is how do these numbers differ from when the players were playing with no rest the day prior – or in other words games on consecutive days in the starting lineup? I gleaned those numbers as well and then compared the two.
|Games Not Rested (GNR)||OPS in GNR||OPS in GFR|
As the data shows 4 of 5 core everyday Pirates position players had a significant bump in OPS when playing on a day after rest. Pedro Alvarez, Starling Marte, and Jordy Mercer each posted an OPS of more than 100 points higher on such days. Only Neil Walker performed worse, albeit only slightly. Andrew McCutchen was an interesting case. In 2013 Cutch mashed on days following rest (1.193 OPS), but in 2014 he posted just a .763 OPS on such days, which was well below he usual performance. As with all splits these stats have sample size limitations and would be expected to vary from year to year, perhaps significantly. This is by no means a complete study that proves anything conclusively. But it does show that perhaps the Pirates are on to something by looking at how the Golden State Warriors rest their core players to optimize performance. There is potentially a competitive advantage that can be gained if the Pirates can find the sweet spot in the correlation between rest and maximum performance on the diamond.