Throwing by the Numbers: An analytical take on the art of pitching

By Nic Osanic

Pitching philosophy in baseball has long been a game of conventional wisdom. Low fastballs and sinkers were thrown to induce weak contact and the job of a breaking ball was only to complete strikeouts. This philosophy always seemed to make sense as the lower the pitch, the easier it is to swing over the top and hit it on the ground. As more data has become available in recent years, batters have started to adjust to these low pitches by swinging on an upward plane (a.k.a. increasing their launch angle). These swing adjustments have transformed traditional groundball/contact hitters into power threats and has seemingly coincided with the league-wide spike in home runs.

Another development is the recent dominance of power fastball pitchers who live in the upper part of the strike zone, such as Gerrit Cole and Josh Hader. As the MLB’s data-driven pitcher/batter pendulum constantly swings in response to new adjustments, one must wonder if – in today’s game – traditional attitudes towards fastball locations are making good pitchers great, or great pitchers good.

Breakdown of pitch locations for the purpose of this article:

\"Screen

Do high 4-Seam fastballs lead to general pitching success?

To see if there is any connection between high 4-Seamers and pitching success, I found each team’s total number of 4-Seam fastballs in the high portion of the transitional Gameday strike zone during the last 5 seasons and divided it by each team’s total pitches thrown over that span. Then I simply compared it to their team SIERA (similar to ERA but removes defense from the equation so it more accurately accounts for pitching ability) over the last 5 years.

\"Screen

Since a low SIERA is better, this graph shows that there is a connection between high 4-Seam fastball usage and pitching success. Most of the league’s top pitching teams over the last 5 years lead the way in high 4-Seam fastball usage. As Toronto Blue Jays fan, one team that I have been forced to keep an eye on are the Tampa Bay Rays. Despite lacking the resources to consistently roster established star-level talent, the Rays’ forward-thinking mindset has allowed them to forgo conventional fastball wisdom and explore the benefits of high 4-Seam fastballs (which a few players and Jim Hickey – former Rays pitching coach – explain here).

It is also interesting to think that this correlation could be even higher. The three teams in the top right of the graphic above – Miami, Detroit, and Baltimore – have all committed themselves to rebuilding recently and have shipped away most of their quality pitching talent, which explains why they have had such little success. Once they complete their rebuilds, their valuing of high 4-Seam fastballs should allow them to achieve pitching success once again. Just for fun, if we were to remove Miami, Detroit, and Baltimore from the graph, our R-squared triples to 0.224.

What portion of the plate is best for 4-Seam fastballs?

My next step in this study is to compare the results generated between high 4-Seam fastballs and low 4-Seam fastballs over the last 3 years. I based this comparison off the last three years since the average velocity in the MLB has jumped from 92.4 mph in 2010 to 93.6 mph in 2019. Velocity has a huge impact on fastball success which is something that we will examine in the next article. Because of this, the numbers from earlier in the decade might not represent the success level of a fastball in today’s game.

Another reason for the three-year span is that the main metric I used – xwOBA – incorporates measurements from the MLB’s relatively new Statcast database. For anyone who is new to advanced statistics, xwOBA is an attempt to crunch offensive production into one number. xwOBA stands for expected, weighted, on-base average and looks the same as batting average in appearance. However, it also accounts for factors such as ballparks (Coors Field gives up a lot more home runs than Oracle Park due to the thin air of its high elevation), hit classification inequalities (home runs are worth more than singles but traditional stats such as batting average treat them the same way), and quality of contact (a broken-bat pop-up that falls for a hit should not be worth more than a 100 mph line-drive caught by a diving third baseman).

Generally, a pitcher with an xwOBA against between .310 and .330 is average, below .300 is great, and above .350 is bad. However, numbers in this article will be inflated as I am only including balls in play in the xwOBA calculation. The reason for this is that walks and strikeouts are dependent on the previous pitches in the at-bat so missing your spot with a 3-ball count results in a walk (an increase of .690 to xwOBA) while missing your spot with a 2-ball count does not make a statistical impact. Since xwOBA was created along with Statcast in 2015, the slightly less accurate wOBA (same as xwOBA but does not use Statcast data to calculate quality of contact) must be used for any performances before 2015.

\"Screen

After accounting for every high and low 4-Seam fastball thrown over the last three seasons, here are the results (we will factor in pitching locations later on).

\"Screen

As you can see, Location B is clearly the better spot for 4-Seam fastballs. It generates more than double the swings-and-misses produced by Location A, which also means that Location B will – on average – allow fewer balls in play per swing by the batter. Additionally, the balls in play allowed by Location B result in fewer hits and – when weighted equally – induces weaker contact. However, ISO – which stands for isolated power and is calculated by Slugging Percentage minus On-Base Percentage – tells us that Location B does result in slightly more extra-base hits. Although 4-Seam fastballs are thrown in Location B at a higher percentage, the lack of success with location A makes me wonder why it still accounts for almost a third of all 4-Seamers.

\"Screen

The best spot to locate 4-Seam fastballs is high in the strike zone or above it. To further back up this point, here is a chart of all the swings & misses in the MLB last year:

\"Screen

This graphic does not show us the total quantity of swinging strikes in each section, but it does show us the percentages. The dark blue (lower whiff rates) becomes lighter as you go higher in the strike zone (neon green box) and the majority of red and white (higher whiff rates) is above the strike zone. It is worth noting that small sections will generally mean a smaller sample size which can influence percentages. However, there is undeniably a trend here which supports the previous findings in the high 4-Seamer vs low 4-Seamer comparison.

Although some pitchers may see the increase in power (ISO) and become afraid to pitch higher in the zone, they must remember that conventional pitching philosophy is also no secret to batters. As a result, batters have increased their average launch angle to allow them to take uppercut swings and lift the low fastballs into the air more frequently. This has left hitters susceptible to the high fastball as they often swing well underneath the ball and hit weak pop-ups or miss the pitch altogether. You may give up a few more extra-base hits if the batter barrels the baseball, but the resulting increase in swings & misses plus the overall weaker quality of contact will have an even greater impact. Altering pitching philosophy is what data-savvy teams such as the Houston Astros have done recently to make their pitchers so successful, and there may not be a greater example of this than with Gerrit Cole.

Gerrit Cole’s transition from good to great

\"Screen

Gerrit Cole has always been an amazingly talented pitcher. He was drafted twice in the first round of the MLB’s First-Year Player Draft – once in 2008 out of high school by the New York Yankees (but did not sign) and once as the 1st overall pick by the Pittsburgh Pirates out of college in 2011. Historically, his greatest weapon is an electric 4-Seam fastball that has averaged above 96-MPH in each of the seven seasons he has played. Unfortunately for Cole, the Pirates did not value the high 4-Seamer and insisted that their players “pitch to contact” instead. This meant throwing low fastballs and sinkers to induce more ground balls, instead of unleashing his best pitch. Cole’s arm talent gave him the ability to suppress runs until 2017 when the new approach of batters finally caught up to him and he posted a mediocre 4.26 ERA. At this point, the Pirates decided to sell while Cole still had value, and traded him to Houston.

\"Screen

Under the expert film analysis of the Houston Trashtros, Cole ditched his sinker almost entirely – throwing it only 4.2% of the time. He has also nearly doubled his usage of the high 4-Seam fastball. Despite allowing the same quality of contact on his high 4-Seamers as in Pittsburgh, his swinging strike rate nearly doubled (possibly due to an increase in fastball spin rate which has been – controversially – the Astros’ recent specialty). Cole also slightly increased the usage of his wipeout slider (from 18.1% in Pittsburgh to 21.6% in Houston) and nearly doubled his curveball total (10.4% to 17.3%). This led to Cole’s surge in strikeouts from an average of 8.44 K/9 (strikeouts per 9 innings) in Pittsburgh to 13.13 in Houston.

By reducing the amount of sinkers he threw, it appears that Cole also made the pitch better as batters were less likely to expect it. Amazingly, Cole was worth the same number of Wins Above Replacement (WAR) – according to Fangraphs – in 2 years with Houston as he was in his 5 years in Pittsburgh. Just like we saw with the high 4-Seamer vs. low 4-Seamer comparison, Cole also did experience an increase in home runs allowed. His HR/9 saw a significant increase from 0.77 in Pittsburgh to 1.05 in Houston. However, the gains he made by switching his pitches and locations clearly outweighed that increase in hitting power, as the Yankees just made him the MLB’s highest-paid pitcher by signing him for $324 million. If the Pirates researched pitch data instead of following their unproven philosophy, they could have received more than just Joe Musgrove, Colin Moran, Michael Feliz, and Jason Martin in return for Cole.

To see just how dominant Cole’s fastball has been during his tenure in Houston, here is a list of the MLB’s top starting pitchers in terms of high 4-Seam fastball effectiveness from 2018-19.

\"Screen

This list shows that a major contributor to Cole’s breakthrough has been his increase in high fastball usage as he possesses the best high 4-Seam fastball among starting pitchers in today’s game. Although it may not have been as dominant in Pittsburgh, Cole’s high 4-Seam fastball was not utilized to its fullest extent and, as a result, Cole’s numbers suffered significantly as his low fastballs and sinkers got crushed. Although Cole is just one example, the direct fastball location comparison clearly shows that traditional pitching philosophies are not as effective in the “launch angle era” and are holding back pitchers who are slow to adapt.

Despite the above conclusion, this article has not demonstrated what makes Cole’s high fastball the best. We know that something caused his high fastball swinging-strike rate to dramatically increase after he was traded to Houston, but that cause is not definitively known. To find these answers, we must dig deeper. Over the next couple of articles, we will uncover the factors that make high fastballs effective, how pitchers can better leverage these factors, more examples of past data-driven pitcher breakouts, and pitchers who could potentially breakout if/when they make these adjustments.

Statistics retrieved from Baseball Reference, Baseball Savant (Statcast), Fangraphs, and MLB.com

Leave a Comment

Your email address will not be published. Required fields are marked *