Examining The Time Through the Order Penalty (Part 1)

This is part 1 in a series where I will attempt to better understand what causes the Times Through the Order Penalty. Part 2 of this series will take focus on individual pitchers in an attempt to better understand the causes of the TTOP.

If you’ve spent time watching baseball you are probably at least somewhat aware of the idea that starting pitchers generally lose effectiveness as they face the opposing lineup multiple times. This is commonly referred to as the Time Through the Order Penalty (TTOP) and while others have attempted to try to understand what causes this, there are some factors that may have not been previously considered. While it is fair to consider pitch quality, there are other factors involved, such as pitcher command and hitters pitch recognition/approach at the plate that may better help explain what causes the TTOP.

Fangraphs splits leaderboards allows us to view statistics for all Major League starting pitchers sorted by each time through the order. For the purposes of this exercise, only data from 2019 starting pitchers will be considered. By examining the data there are certainly some noticeable trends that emerge.

Time Through Order	Total Batters Faced	wOBA	FIP	xFIP	K%	BB%	HR/9	BABIP	Hard%
1	43027	.310	4.19	4.24	24.3%	7.9%	1.3	.293	37.3%
2	40067	.326	4.63	4.58	21.5%	7.6%	1.5	.299	38.6%
3	23543	.336	4.92	4.79	20.1%	7.5%	1.6	.301	40.1%
4	858	.299	3.73	4.18	21.8%	5.5%	1.0	.296	33.0%

While it should come as no surprise that starting pitchers tend to get worse the 2nd and 3rd time they face an opposing lineup, there is certainly some noticeable information uncovered. The first statistic that stands out is walk rate. As pitchers go through the order multiple times they actually start walking batters less frequently. Given how much worse pitchers get overall each of the times through the order, this is certainly surprising.

One of the other trends noticeable at first glance is seen in the fourth time a pitcher faces an opposing order. While it might seem off at first that starting pitchers actually did much better the fourth time facing an opposing lineup in 2019, there are certainly some potential explanations for this. Since starters very rarely face an opposing lineup a fourth time through nowadays, usually only starters pitching at their very best will get the opportunity to do this. Since the fourth time through the order is such an outlier and a much smaller sample of pitches than the others, it will be ignored for the rest of this exercise.

To begin to understand the effects of the time through the order penalty, one can begin by querying a Statcast pitch-by-pitch database for all 2019 starting pitchers in an attempt to determine if some aspects of pitch quality is decreasing as a starter faces a lineup multiple times through. One of the first things people oftentimes consider when evaluating pitches, especially fastballs, is velocity.

TTO	Fastball Velo (MPH	Two-Seam Velo (MPH)	Sinker Velo (MPH)	Cutter Velo (MPH)	Change-Up Velo (MPH)	Curveball Velo (MPH)	Slider Velo (MPH)
1	93.08	92.24	91.91	87.75	84.37	78.44	84.78
2	92.84	92.16	91.4	87.6	84.28	78.08	84.56
3	92.90	91.25	91.46	87.44	84.27	78	84.62

While it definitely appears that average pitch velocity decreases on average the second time through the order, the average pitch velocity actually increases for a some of the pitches the third time against an opposing order. While it would certainly make sense that velocity would decrease over the course of a start as the pitcher tires, the results shown above are again likely due to something else happening with the data. Better pitchers are going to face an opposing lineup the third time more frequently and, as a result of this, taking a macro level look at this problem may in fact be a problem in and of itself. There are certainly other trends around the league that also must be considered, like use of the opener strategy, when attempting to study the TTOP.

There are many other factors that can help begin to understand the quality of a pitch. Since the implementation of the Statcast tracking system prior to the 2015 season, many analysts have gone to spin rate in an attempt to easily quantify pitch quality. While all pitch types will be considered for the purposes of this exercise, please note that raw spin rate is usually not something to look at in isolation, but usually works a little better when evaluating a pitcher’s individual pitches. Given this fact, there are certainly many other factors that must be considered when attempting to understand pitch aerodynamics.

TTO	Four-Seam Spin Rate (RPM)	Two-Seam Spin Rate (RPM)	Sinker Spin Rate (RPM)	Cutters Spin Rate (RPM)	Change-Up Spin Rate (RPM)	Curveball Spin Rate (RPM)	Slider Spin Rate (RPM)
1	2273	2163	2138	2333	1818	2531	2420
2	2272	2170	2142	2334	1802	2522	2418
3	2276	2172	2134	2336	1793	2530	2426

There certainly isn’t much of an effect noticeable here and most of the variations seen between the separate times through the order are likely a result of the previously mentioned problems with taking a look at this issue from this level. There are certainly other metrics that can be considered in attempt to better understand the quality of a starters pitches.

The ability of a pitcher to get swings and misses consistently is one method that can be used in an attempt to easily quantify how good a pitchers stuff is and also generally serves as a good predictor of future pitching success. By using the Statcast 2019 pitch-by-pitch database, one can calculate the Whiff% (total pitches swung at and missed/total swings) for each of the pitches throughout the different times through the order.

TTO	Four-Seam Whiff%	Two-Seam Whiff%	Sinker Whiff%	Cutter Whiff%	Change-Up Whiff%	Curveball Whiff%	Slider Whiff%
1	21.37%	14.13%	16.07%	24.38%	32.14%	33%	36.37%
2	20.66%	12.87%	14.16%	22.21%	28.60%	29.56%	33.61%
3	19.47%	12.65%	12.5%	21.51%	29.02%	29.65%	32.19%

So we can see here that for most pitch types starters are getting fewer swings and misses the more times they face an opposing lineup. Pitchers missing fewer bats and pitching to more contact certainly might, at least in part, begin to explain why starters get so much worse the second and third time facing an opposing lineup. While Whiff% does appear to be consistently down across the board each of the times through the order, there are certainly still other factors that could be in play here.

The Statcast pitch-by-pitch database can also be utilized in an attempt to look at how command is effected as a start progresses. We can further divide our data into what Baseball Savant calls attack zones. These attack zones are divided into heart, shadow, chase and waste pitches (see below). Pitchers are most often going to try to work in what is referred to as the shadow zone. This makes a lot of sense and we can prove pitchers are mostly trying to locate pitches in the zone by looking at where catchers will most frequently set up pre-pitch. More than 42% of all Major League pitches were located in this shadow region in 2019 which is much more than the next region, the heart region, in which closer to 25% of all Major League pitches were thrown in 2019.

TTO	Heart%	Shadow%	Chase%	Waste%
1	25.7%	42.67%	22.63%	8.28%
2	24.99%	42.86%	22.97%	8.42%
3	24.18%	42.74%	23.46%	8.94%

While we have previously considered if some aspect of pitch quality can explain the time through the order penalty, the database can further be utilized to begin to understand if command is also effected. As a pitcher tires over the course of a start, one might think this would have an impact on his ability to consistently locate his pitches. While the rate of pitches located in the shadow zone increases the second time through the order, the rate actually decreased the third time through the order in 2019. One should also consider the pitches thrown in the other attack zones as well and there are certainly less pitches thrown in the heart of the plate and more in the chase and waste zone as a start progresses. It certainly looks like there could possibly be evidence to show decreased command might also explain some of the time through the order penalty. However, as previously mentioned, the third time through the order data is likely skewed due to the quality of starter that most frequently gets the opportunity to face an opposing order a third time through. Part 2 of this series will primarily focus on breaking down pitch quality and attack zone percentages by individual players in an attempt to better understand the causes of the TTOP.

We can begin to understand how batters change their approach at the plate each of the times through the order by looking into some commonly used plate discipline metrics. Again using Statcast pitch-pitch data for all 2019 starters, one can calculate some of the more commonly used plate discipline metrics for each TTO.

TTO	Swing%	O-Swing%	Z-Swing%
1	45.9%	27.9%	64.8%
2	47.5%	29.2%	67.6%
3	48.6%	30.2%	69.4%

Hitters overall are swinging more frequently as they face an opposing starter multiple times. This makes sense because as a hitter sees a pitcher a couple of times he likely has a better feel for the pitchers arsenal and the pitches he is seeking out. We have previously seen that strikeout rates, whiff rates and walk rates all decrease as a pitcher faces an order multiple times. While it certainly isn’t ideal that hitters begin chasing more pitches out of the zone, this fact, combined with the other evidence, provides some proof that hitters have a better idea of a pitchers arsenal and what certain pitches he is searching out at the plate the more times he sees an opposing pitcher.

It is almost undeniable at this point that pitcher effectiveness will generally decrease the second and third time he faces an opposing order. While taking a macro level look at the pitch-by-pitch data certainly didn’t uncover much in terms of overall pitch quality decreasing, this can perhaps be explained by overall starter quality and league wide trends having an influence on the data. While there is evidence presented here that certainly might begin to explain the TTOP, focusing on individual pitchers in part 2 of this series might uncover some more useful results.

All statistics and data courtesy of fangraphs.com and BaseballSavant.com