Throwing by the Numbers: Other determinants of high 4-seam fastball success

By Nic Osanic

In the first two articles, we talked about how pitch selection, location, and velocity can all impact the swing and miss ability of different pitches. We saw how velocity clearly makes a difference when it comes to the success of 4-Seam fastballs as long as the pitch is thrown high over the plate. However, guys like Aroldis Chapman, Noah Syndergaard, and Ken “100 Miles” Giles all throw triple-digit high heat but don’t appear on any of the high fastball success leaderboards in the previous articles. Why is that? Well, this likely means that there is more to fastball success than just location and velocity. So what other significant factors could we be missing? This is the question we will dive into in this article.

1. Spin rate

\"Keuchel.jpg\"

The first additional influencer of fastball success is spin rate, which has been somewhat of a buzzword in baseball for the last couple years. The general understanding is that the higher the spin rate, the more movement each pitch will have. Nevertheless, there is more to it than that.

When talking about spin rate, it is important to know that there are two types of spin. First, there is Active Spin (sometimes referred to as “Transverse Spin”) which is the amount of spin that results in the baseball’s directional movement. The other type is called Inactive Spin (sometimes referred to as “Gyrospin”) which does not impact the trajectory of the ball and works like the spiral of a football or bullet instead. This means that a pitcher with a high spin rate will not generate extra movement unless the majority of their spin is Active Spin.

Here are the 2019 leaders in fastball Active Spin %:

\"Active

Everyone on this list has an above-average fastball movement profile. This makes sense since their high Active Spin rate means that the majority of their fastball spin contributes to movement. The results of this chart do raise a question though, as both Colin Poche and Marco Estrada have elite vertical movement but below-average horizontal movement on their fastballs, whereas Caleb Smith is the opposite. This made me wonder whether there is an ideal movement profile that affects high fastball results.

In order to find this out, I used Baseball Savant’s 4-Seam Fastball Movement Above Average leaderboard and grouped every pitcher’s 2019 vertical and horizontal movement by intervals. For the sake of sample size, I only included pitchers with a minimum of 50 high 4-Seam fastballs to ensure that their swinging strike percentage was not inflated by factors such as facing a batting pitcher. Using movement above-average allows us to see who adds the most movement to their pitches as it compares each pitcher’s movement to that of pitchers with similar velocity and release extension – otherwise, pitchers who throw soft would head the leaderboard, as slower pitches tend to drop more (i.e. fastballs vs changeups). It is also important to note that less vertical drop is considered good for fastballs and is therefore classified as positive vertical movement.

This is what it looks like:

\"Fastball

Unfortunately, Pitch Movement Above Average is not available on individual pitches and can only be used as a season average. To see the potential problem with this, let’s use an example of a pitcher with above-average movement over the course of the season. 

If a pitcher generates a swinging strike on a singular straight fastball, then this swinging strike will still be counted as coming from a pitcher with above-average movement. Knowing that these results aren’t perfect, we can still see that having above 2-inches of vertical movement compared to the average on a high 4-Seam fastball generates nearly 5% more swinging strikes than a fastball thrown in the same location with 2-inches of vertical movement below average. This means that increasing the vertical movement on a pitcher’s fastball can significantly improve their swinging strike rate. On the other hand, horizontal movement has a much weaker correlation with swinging strike percentage but with an R2 of 0.0388, a pitcher in need of every possible advantage could likely increase their horizontal movement to boost their swinging strike percentage by a fraction of a percent.

After looking at the impact of movement on swinging strikes, it is only fair to do the same for quality of contact in 2019:

\"Vertical

Interestingly, there is pretty much no correlation between quality of contact and different amounts of both vertical, and horizontal movement. The R2 value for vertical movement is less than 0.01 and the value for horizontal movement is slightly negative (meaning that pitches with positive horizontal movement above average are slightly more likely to result in runs when hit into play). This finding is surprising given the drastic difference in swinging strike percentage with vertical movement.

This could mean that xwOBA on Contact (xwOBACON) is a volatile stat from year-to-year, as Justin Verlander and Josh Hader both had stellar seasons in 2019 but combined for an xwOBACON of .422, despite their combined xwOBACON of .337 from 2015-19. If this is the case, then it could partially explain why these numbers appeared as universal across each interval, as Hader and Verlander both had top-tier vertical fastball movement and potentially inflated the 2.0+ vertical movement interval by getting unlucky with a high xwOBACON. However, each movement interval accounted for at least 1800 balls in play so the data should be normalized to its true values. Either way, it is apparent that movement had no impact on high 4-Seam fastball quality of contact in 2019.

2. Arm angles at release point

\"Pitch

Arm Angle is responsible for the difference between a sweeping curveball and a 12-6 curveball. Sidearm pitchers create more sidespin (sweeping motion) which forces the ball sideways and overhand pitchers create more topspin (downward motion), which forces the ball down. The same idea works with 4-Seam fastballs. A pitcher who releases a fastball with an overhand throwing motion is able to create more backspin on their pitches.

Fastballs with lots of backspin can help to offset the effects of gravity on the baseball by reducing the amount of drop on the pitch due to the Magnus Effect. This can give batters the illusion of the vaunted “rising” fastball as the pitch drops less than normal (studies have shown that no pitcher has ever thrown hard enough to make a fastball physically rise). Sidearm pitchers usually get sidespin on their fastballs which leads to more horizontal movement and limited vertical movement. Pitchers with a three-quarter delivery will tend to get a mixture of both vertical, and horizontal movement on their fastballs. A deeper understanding of pitch movement is available here.

Here is a visual representation of the connection between vertical release point and vertical movement:

\"Movement

Although there is a correlation, I think it is likely stronger than this shows. Due to limitations within the available data around vertical release points as they simply measure the distance between the ball at the point of release and the ground. This means that it does not account for variations in pitcher height or the angle of their arm when the pitch is made. To better explain this, here is an example of 6’8” Tyler Glasnow:

\"Glasnow.jpg\"

Glasnow’s release point is measured at 6’1”. Considering that he is 6’8”, raw release point data would suggest that Glasnow has a three-quarter/sidearm delivery since his release point is much lower than his height. However, the picture above suggests that Glasnow is an overhand pitcher who just doesn’t stand upright during his delivery. Unfortunately, I could not find any type of pitcher arm angle leaderboard so I cannot prove that the correlation between vertical release point and vertical movement should be higher. 

If we assume this anyways, then – in theory – pitchers can alter their release point to generate different types of spin. This appears to be the case with Lucas Giolito’s sudden breakout in 2019:

\"Giolito

In 2018, Giolito’s release point was a mess. He did not throw his pitches out of a consistent arm slot throughout the year and his peripherals failed to indicate any sort of impending breakout. As a result, he struggled mightily. Then things changed. In the 2018-19 offseason, Giolito played around with a Rapsodo camera to refine his throwing mechanics and pitch movement profiles. The results were incredible. Giolito condensed his release point to create more of a “tunneling effect” and altered it both vertically and horizontally. Here is a physical view of his changes:

\"Giolito

Now let’s look at the impact these changes made on his fastball:

\"Giolito

As expected, Giolito’s increased arm angle led to more vertical movement and less horizontal movement. This increase in vertical movement was partly responsible for his jump in swinging strike rate with the other part being his 2 MPH increase in fastball velocity. This increase can be attributed to Giolito’s offseason training which involved throwing weighted baseballs and his mechanical work with the Rapsodo camera, which caused him to shorten his arm stride (pictured below). Both of these methods are proven to increase velocity while simultaneously reducing the arm stress created from throwing thousands of pitches each year.

\"Giolito

Another example of the impact of arm angle at release point can be demonstrated through a comparison of Chris Sale and Josh Hader.

\"Sale

Both pitchers had one of the lowest vertical release points in the league last year and were nearly identical in terms of horizontal release point. Still, they somehow experienced different types of movement. By freezing video of these two pitchers at their moment of release, we may be able to come up with an explanation:

\"Sale

Despite Josh Hader having an unusually low release point, this image shows that his arm angle is more vertical than Chris Sale’s at release point. By doing this, Hader should be able to release the ball with his fingers somewhat on top of the baseball (creating backspin and vertical movement) whereas Sale’s fingers release from the side of the baseball (creating sidespin and horizontal movement).

Here is what their fingers should look close to if we could zoom-in:

\"Sale

Obviously this analysis isn’t perfect. It is subject to a guess of their finger positioning, and that those freeze-frames are an accurate representation of Hader and Sale’s entire season’s worth of pitches. Nevertheless, it does make sense intuitively that Hader’s 3-quarter arm angle and finger positioning would give him more vertical movement and less horizontal movement than Sale. It should also be noted that Hader’s active spin percentage is 96.7% whereas Sale’s is 87.9%. This means that Hader has nearly maximized the movement potential of his pitches and that Sale could generate greater amounts of both vertical and horizontal movement by slightly tweaking his throwing motion. 

3. Deception

\"Trevor

This is probably the hardest fastball success factor to quantify since it all revolves around the perception of the batter. The longer a pitcher can “hide” the ball from the batter, the less time the batter will have to see and react to the pitch. This can be especially useful for pitchers who have a slight variance in release point as batters will not know where the pitch is coming from until it has been thrown.

Deception is likely the reason why pitchers with below-average vertical movement and velocity – such as the Milwaukee Brewers’ Brent Suter – can still manage to generate strong whiff rates on high fastballs. Furthermore, a combination of deception, above-average vertical movement, and average velocity is becoming the recipe to surviving in the major leagues for fastball-exclusive pitchers such as Sean Doolittle, and Colin Poche.

This is probably the hardest fastball success factor to quantify since it all revolves around the perception of the batter. The longer a pitcher can “hide” the ball from the batter, the less time the batter will have to see and react to the pitch. This can be especially useful for pitchers who have a slight variance in release point as batters will not know where the pitch is coming from until it has been thrown.

Deception is likely the reason why pitchers with below-average vertical movement and velocity – such as the Milwaukee Brewers’ Brent Suter – can still manage to generate strong whiff rates on high fastballs. Furthermore, a combination of deception, above-average vertical movement, and average velocity is becoming the recipe to surviving in the major leagues for fastball-exclusive pitchers such as Sean Doolittle, and Colin Poche.

For those who have not heard of Colin Poche, he is a 26-year-old relief pitcher for the Tampa Bay Rays who was the ‘player to be named later’ in the trade that sent Steven Souza Jr. to the Arizona Diamondbacks. Poche put up a pedestrian 4.70 ERA in 51.2 innings pitched during his rookie year in 2019, and his best pitch – a fastball which averaged only 92.9 MPH – was thrown a whopping 88.5% of the time last season (highest fastball usage rate in the majors). However, Poche sports a career 2.22 ERA in the minor leagues which includes 275 strikeouts in 174.2 innings pitched (40.4 K%) and his Statcast’s expected ERA in the majors last season was 2.86 which is impressive for a reliever. 

So, how did Poche produce these eye-popping numbers despite overly relying on a seemingly below-average fastball? The answer is a combination of his deceiving delivery and his ability to create vertical movement. Both of these factors allowed Poche to fool batters despite them essentially knowing that a fastball is coming each time. Poche could likely survive with only one of these fastball aids, but having both of them should make him a special reliever over the next couple of years (unfortunately this year will not be one of them as he tore his UCL and just underwent Tommy John Surgery in July 2020).

Now that we have seen how deception helps fastballs with average and below-average velocity, here is an example of a pitcher with lots of deception and above-average velocity – Josh Hader:

\"Hader.jpg\"

One of the key contributors to Josh Hader’s dominance is his deception. As seen in the picture above, Hader does a great job hiding the ball from the batter behind his body before releasing each pitch. This makes it difficult for batters to pick up the baseball until it is too late. With Hader also generating vertical movement from his low arm slot and averaging 95.5 MPH on the radar gun, he has been able to become arguably the best reliever in all of baseball.

Something else that I find interesting is the high fastball success of sidearm and submarine style pitchers. As a Blue Jays fan, this became apparent whenever I would watch the Orioles’ Darren O’Day pitch to anyone not named Jose Bautista. O’Day typically sits in the mid-to-high 80’s with his 4-Seam fastball but has averaged over a strikeout per inning throughout his career. The reason for his success is the high location of his fastballs which complement his elite slider. His high fastballs seem to be even more effective from his low, submarine-style arm slot.

Compared to a typical overhand pitcher, O’Day releases the baseball at waist-height and literally has to throw his fastball in an upward direction in order to make it reach the high portion of the strike zone. Gravity will still make the pitch drop somewhat downwards but the apparent upward direction of the baseball is enough to deceive batters and generate swinging strikes when thrown up in the zone. Darren O’Day is one example but not the only one. Below is a list of the high fastball results from notable low arm slot pitchers since the creation of Statcast:

\"Sidearm.jpg\"

All of these pitchers tend to have a release point near or below their shoulders. They also all have a high 4-Seam fastball swinging strike rate above the Major League average which is significant considering many of them don’t throw as hard with their low arm slots. In the end, most of them are established big leaguers and have pitched in the majors for a long time thanks to their high fastballs which play up due to their deceptive deliveries.

Putting it all together

\"Gerrit

If you were to sum up everything in this article that could increase a pitcher’s performance into one example, it would be through Gerrit Cole. Previously, we saw how his change in pitch usage from sinkers to high 4-Seam fastballs allowed him to get better results. Now let’s look at why that is.

\"Cole

In 2017, we can see that Cole’s fastball was hard – averaging nearly 96 MPH – but it was relatively flat and had a similar movement profile to that of Chris Sale (lots of horizontal movement but lacking vertical movement). His spin rate was nothing special and unfortunately, Statcast did not start tracking Active Spin Rate until 2018. All of this resulted in a – rather lacklustre – high 4-Seam fastball swinging strike rate of 13.5%.

In 2018, we saw major changes. Cole’s fastball velocity increased by 0.6 mph and his spin rate went up as well which tends to happen when pitchers add velocity. Suddenly, Cole had above-average vertical movement on his fastball for the first time in his career. As a result, his horizontal movement also dropped. As we’ve seen, this can sometimes be the result of mechanical changes in a pitcher’s delivery so let’s look for any changes in Cole’s delivery:

\"Cole

After joining the Astros in 2018, it appears that Cole made his release point more consistent (his individual fastballs come from a smaller condensed area now) and slightly raised his vertical release point. In an article I recently read (highly recommended read), it was also suggested that Cole appeared to be throwing from a slightly higher vertical release point when his fastball RPM was at its highest in Pittsburgh. Although there is definitely evidence of a change in vertical release point, it looks relatively minor so it might not be the direct cause of his massive jump in fastball effectiveness. 

Instead let’s look at Gerrit Cole’s 4-Seam Fastball grip:

\"Cole

At first, both pictures look very similar. Cole slightly increased his spin axis with the Astors by raising his arm angle vertically but once again, that does not seem to be a significant adjustment. However, there may be a difference in his grip on the laces if you look closely:

\"Cole

Although this is really hard to see, a different view of Cole on the mound can back up this claim:

\"Cole

In this shot, you should notice the direction of the baseball’s seams and the Major League Baseball printing. Every single picture I have seen with Cole throwing a 4-Seam fastball in Houston has the printing going horizontal from this point of view. In Pittsburgh, some of his 4-Seam fastball grips had the printing vertical, while others had it horizontal. This could be important because, during his time in Pittsburgh, Cole threw nearly 12.5% of his 4-Seam fastballs at 2300 RPM or higher. When he threw his 4-Seam fastball high with the same spin rate, he generated a 19.3% swinging strike rate – nearly identical to his 2018 mark in Houston of 19.4%. Maybe there is a connection between his 4-Seam fastball grip and his abnormally high spin rates.

It could be possible that Cole was unknowingly throwing his 4-Seam fastball with the Major League Baseball logo horizontal when he would hit 2300 rpm. There would be few differences with the grip of his fingers other than the “horseshoe” of the seams being on one side of hand vs the other. This could allow this minor detail to go unnoticed.

Here is how this change in grip could affect his fastball:

\"Cole

These images are not of Gerrit Cole’s actual grip but represent what it would look like from this angle based on the previous images of his grip. Notice how the stitches are on the left side of the index finger when he was with Pittsburgh but are on the right side of his middle finger in Houston. You might also notice that the second image is the 4-Seam fastball grip of Justin Verlander – Cole’s teammate – who helped him with his transformation to the Astros system.

Here is the difference a right-handed, 3-quarter, 4-Seam fastball grip can make at release point:

\"Fastball

Cole’s fastball grip in Pittsburgh is the left image whereas his grip in Houston is the right one. There is a noticeable difference in the direction of the stitches which could lead to slightly different movement patterns.

Another possible difference between the two grips is the amount of finger contact on the laces of the baseball. If we look back on the image of Cole’s grips, his new 4-Seam fastball grip allows his ring finger to apply pressure against a portion of somewhat horizontal stitching. This extra grip could generate more speed during the ball’s transition from his grip to his fingertips in the milliseconds before each pitch, thus creating more spin and explaining the dramatic spin rate increase seen in the immediate aftermath of the Cole trade.

\"Fastball

Either way, there are not enough publically available resources to determine if this grip change could generate more spin or movement. Given what we already know about spin rates, however, different pitch grips have been proven to make these types of increases happen.

A familiar story

\"Lucas

Cincinnati Reds pitcher Lucas Sims also enjoyed spin rate fueled success last season. Although his 2019 ERA of 4.60 was slightly below league average, Sims should have been much better. Accounting for quality of contact, his expected ERA (xERA) was 3.15 and his 32.2% strikeout percentage was a career-high. Sims also ranked 8th in 4-Seam fastball spin rate after something resulted in a yearly average increase of 249 RPM between 2018 and 2019. Let’s have a look at what could have caused this starting with his release points:

\"Lucas

Like Cole, there is not much of a change in Sims’ horizontal release point on fastballs. His vertical release point appears to have dropped slightly but this likely had a minimal impact on his spin rate.

Interestingly enough, a look at his fastball grip also shows an obvious difference:

\"Lucas

This image clearly shows that Sims did change his 4-Seam fastball grip and – like Cole – experienced an immediate increase of 200 RPM to his fastball’s spin rate. Other fastball performance indicators before and after the grip change are also very similar between Cole and Sims:

\"Lucas

These results show that Cole and Sims both saw large increases in their Bauer Units (spin rate divided by velocity) – which tends to be incredibly hard to significantly change. This likely indicates that a major adjustment of some kind (legally or illegally) has taken place. 

Other notable pitchers who throw their 4-Seam fastballs with the same grip include: pitch design guru Trevor Bauer, Cole’s teammate Justin Verlander, spin artists Corbin Burnes and Dillon Maples, and Rich Hill – who appears to have started throwing his fastball with a different grip at some point last season. 

Although these grip changes may have ultimately not played a factor in their spin rate increases, there must be some reason why both pitchers would decide to change their 4-Seam fastball grips. It is also likely not a coincidence that Cole made his adjustment with the forward-thinking Astros and Sims made his with the Reds – who have gone all in on player development by hiring team personnel such as Kyle Boddy (founder of Driveline Baseball). 

Either way, there does not appear to be much research into the possibility of an optimal fastball grip other than a few Driveline studies such as the one included earlier in this article. This simple concept of different fastball grips could unlock extra spin potential for pitchers who are currently using a suboptimal grip. However, we will not know until more tests are done with conclusive evidence to back it up.

Conclusion

\"Bregman.jpg\"

In this article, we unlocked the hidden elements that create an elite fastball. It is not always enough to just feature top-notch velocity; it is also important to have a high spin rate. However, not all spin is equal as only Active Spin generates movement. A high Active Spin rate will make the ball move more in the direction of the baseball’s spin axis.

To generate swinging strikes from that movement, a pitcher needs to have the arm angle and spin axis to produce vertical movement. Even then, a pitcher isn’t guaranteed to be good. In order to be consistently successful with the fastball, a pitcher must have a combination of proper location, premium velocity, a high spin rate which generates vertical movement, and an element of deception during their delivery. 

Relief pitchers with most of these traits such as Sean Doolittle and Josh Hader have already proven that pitchers can survive throwing fastballs upwards of 80% of the time. This is encouraging as last season saw many starting pitchers throwing 4-Seam fastballs above 50% of the time and many of those starters had above average velocity and/or vertical movement. Unfortunately, many of them made the first mistake of focusing on the bottom of the plate and paid the price (Antonio Senzatela, Vince Velasquez, Glenn Sparkman, and Jeff Hoffman were among the starting pitcher leaders in fastball usage).

If this high fastball trend continues, we should hopefully see some of these fastball-first starting pitchers use the upper part of the zone more and reap the rewards. We already saw what happened when Gerrit Cole doubled his high fastball usage to 30%, imagine the strikeout potential that many current “contact pitchers” would have if the majority of their fastballs weren’t so easy to hit being down the middle or at the bottom of the zone. In the next article, we will look at more pitchers who have made these changes in recent years and those who still can before they vanish from the league.

 

 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 *