It appears the batting strategy of lowering the O-Swing rate was effective. The benefits were limited.
It appears that the Diamondbacks’ batting strategy is to not swing at pitches outside the strike zone, and by waiting for pitches in the strike zone (presumably more hittable) raise the pitch counts and achieve above average hitting. The caveat is I don’t have inside information and my statement could be off-target.
A key measure will be the O-swing rate from FanGraphs. Let’s talk about what it is. It is a measure of how often a batter swings at pitches outside the strike zone. It helps tell the story of each batter’s plate discipline and approach. A high O-Swing rate could indicate either poor plate discipline or an aggressive batter who prefers a hit instead of a walk. It is calculated by dividing batter swings at pitches outside the zone by total pitches outside the zone. FanGraphs provides two versions of the O-Swing calculation, each with a different data source. One data source is the raw data from PITCH f/x. The other data source is Baseball Information System (BIS), which has data which was modified by “human coders.” My view is that the BIS data better reflects the ground truth, so that is the version that I referenced.
One key premise is that a lower O-Swing rate is better for the batter. Although it’s logical that less swings at pitches outside the strike zone helps the batter, confidence in the premise would be higher if the best 8 batters had a lower O-swing rate than the other batters for the Diamondbacks. For the 2021 Diamondbacks, batters with the best OPS+ had significantly lower O-swing rates. The PA weighted average O-Swing rate was 27.8% vs 32.8%. As a point of reference league average is about 30%.
An alternative premise is that there is an optimal range for O-Swing rate. That range could differ for different types of batters. That premise would seem to be supported by the following chart which shows the O-Swing rates and OPS+ for the Diamondback batters. The optimal range of O-Swing rate for the top batters could be 26.2% to 30.6%. The optimal range for the next tier of batters could be 28.4% to 35.8%.
Another premise is as O-Swing rate is lowered, the walks per PA are higher. For the Diamondbacks batters, the following graph seems to confirm that premise. I added a line to illustrate a linear relationship that seems to exist for O-Swing rates between 25% and 35%.
An interesting question is what are the characteristics of batters who had the most plate appearances (PAs)? Two insights follow:
- The top 8 batters (OPS+) accounted for only 52.1% of the Diamondbacks PAs through 14 June. Was it remarkably low? Let’ s look at the percentages for two other teams: Rockies: 75.8%, Giants: 53.2%. So it was low, but not unusually low.
- 57.6% of the Diamondbacks’ PAs were by batters who improved their O-swing rates in 2021. It appears that four of the bottom 13 batters, who improved their O-swing rates, were given significant PAs.
The Top 8 Batters.
The top batters had the highest OPS+. Let’s look at O-swing rate of the top 8 batters plus Eduardo Escobar (#9). The following table shows O-swing rates compared to the league average of about 30%, and which batters increased or decreased their O-swing rate compare to 2020.
I added the #9 batter (Escobar) to the chart because it suggests that if he further improved his O-swing rate to league average, he might crack the top 8 batters. Two other batters with potential, but high O-Swing rates were Christian Walker and Tim Locastro.
The chart show me that Rojas, Kelly, and Cabrera are best at avoiding swinging at pitches outside the strike zone and they improved in 2021.
For the top 8 batters, did opponents pitchers react by changing the percentage of their pitches in the zone (thereby raising their pitch counts)? The answer was split; 4 batters had a higher percentage and 4 batters had a lower percentage. The more interesting observation was that in 2021, seven of the top 8 batters faced a narrow range of pitches in the zone (39.9% to 45.0%). Andrew Young saw an unusually low percentage of pitches in the zone (35.2%) and he did well to keep his O-swing rate low at 28%.
Let’s compare those 8 batters to the other 13 batters.
As we covered earlier, the top batters had a significantly lower average O-Swing rate. It appears that the lower O-Swing rate was associated with more RBIs and homers.
- Did the top 8 batters hit significantly more RBIs per PA than the other 13 batters? Yes (.117 vs .095). The batters with the best OPS+ hit RBIs at a rate about 20% higher.
- Did the top 8 batters hit significantly more HR per PA than the other 13 batters? Yes (.032 vs .021). The batters with the best OPS+ hit homers at a rate about 50% higher.
Homers matter. In June the Diamondbacks had a 5 game streak and a 6 game streak without hitting a homer. During those streaks, the Diamondbacks averaged 2.8 runs per game, which compared unfavorably to the 4.2 runs per game for the season through 14 June.
Did these 8 batters receive a different percent of pitches in the zone than the other 13 batters? There was a small difference but it may not be significant (43.2% vs 42.9%). Despite an unusually low percentage of pitches in the strike zone, Andrew Young did well to keep his O-swing rate at 28%.
The top 8 batters in OPS+ had more RBIs and homers per PA, and had a lower average O-Swing rate compare to the bottom 13 batters. Most of them improved their O-swing rate compared to 2020. Half of them had a lower O-Swing rate than the league average.
Summary.
It appears the batting strategy of lowering the O-Swing rate was effective. The benefits were limited because only about half the PAs were by batters with an O-Swing rate below the league average. The upside is further gains could be achieved. Let’s end with some baseball humor about a pitch that was far outside the zone:
“Juuuusssst a bit outside.” — Bob Uecker’s version of Harry Doyle in Major League.