The Cardinals have been on a hot streak since the beginning of July, winning 12 out of their last 13 games. The question is: does this mean they are playing better? Not necessarily. Over the course of any season, teams will often experience streaks where they play well for a period but regress back to their true talent level shortly thereafter.
This article discusses how managers use special teams in baseball to mitigate the effects of regression and maintain an advantage over opponents who do not employ them successfully.
Special teams are an important part of baseball offense. Teams have the 9th hitter in the lineup come up with runners on base more often than any other position to take advantage of his high OBP even though it results in worse lineup balance. A team’s first basemen usually has a below average glove, but managers put them there because they hit well. First base is often vacated by the designated hitter in American League leagues where teams give up an excellent hitter for one who can play defense when their lineups are filled with right handed, low power hitters.
Despite the importance of special teams in offense, managers do not treat them equally when setting lineups or determining defensive responsibilities. This is because they are subject to regression, which is the tendency of statistics to move towards the average of all players in a sample.
This effect is easy to see among special teams. A hitter’s batting average on balls in play (BABIP) tends towards .300 while his strikeout rate tends toward 20 percent of his at bats. The BABIP of players who attempt to steal (rushing attempts plus stolen bases) is 70 percent higher than league average. Teams’ BABIPs are three times higher with men on base and two times higher with the first batter of an inning at the plate rather than the ninth.
Regression can be used to understand how team performance changes over periods of time. It shows that teams have been better than their true talent throughout the season. In other words, since the beginning of July, the Cardinals’ offense has been performing at a level higher than they were capable of before then. Before this surge, they had been performing even worse than their low standard from earlier in the year because their production fell towards their true talent level.
The Cardinals’ case is an extreme example of two phenomena that affect all managers: improved performance by teams over a period and the disappearance of value from players who were acquired for a reason, but have since stopped producing. The lineups of most teams move from being balanced with respect to expected OBP to being unbalanced in favor of power hitters who will not be affected by regression because they likely see fewer chances to contribute to their teams. It is unbalanced because the improved performance of the team over a period increases the expected benefit from high OBP players and thus reduces that from low OBP ones.
The lineup changes explained above are only one example of how managers deal with the problems of regression. They also alter defensive alignments, pinch hit for players who regress more than expected, and make trades to replace under-performers on their rosters. Each manager must determine how he will address these issues without exposing his team to the risk of not meeting his own performance benchmarks or appearing weak to agents, teammates, or fans.
Regression can be used to explain how teams change performance over periods of time. They improve (and their players’ performances regress) towards their true talent level, which is often lower than what it appears to be when looking at a period that is still relatively short. Teams’ success and under-performance during these periods may cause managers to change their lineups by being less balanced with respect to OBP and more reliant on power hitters who will not regress as much.
The Cardinals’ offense has been playing above their expected level since the beginning of July. Before then, they were performing under expectations because players’ production declines towards their true talent level. The Cardinals have been balancing their lineups less with respect to OBP and more with power hitters who will not regress as much. This is because the better performance of the team over a period increases the expected benefit from high-OBP players, so there is more pressure on low-OBP ones.