Does Pitch Tunneling Actually Work?

Pitch tunneling is among the many sizzling phrases thrown across the baseball area. Analysts, gamers and coaches, commentators, and followers alike all appear to have totally different opinions about its effectiveness in addition to its legitimacy as a complete, however what’s pitch tunneling anyway? Though generally made to look like a darkish artwork or a sophisticated science idea, pitch tunneling at its core isn’t all that complicated, and is just one other approach for pitcher’s to deceive hitters. On this article, I’ll be analyzing whether or not or not pitch tunneling truly works, and if that’s the case, how a lot.

Pitch tunneling is the idea of matching pitched balls’ flight paths intently sufficient to make them seem like the identical pitch lengthy sufficient to idiot the batter earlier than they transfer in several instructions. This can be complicated to visualise by yourself, so right here’s a video of Shane Bieber performing some tunneling, courtesy of Pitching Ninja. There are a number of components that go into tunneling 2 pitches collectively; a very powerful being launch level, pitch trajectory, and plate location or the place the pitch finally ends up. As tunneling depends on 2 pitches touring the identical path after which breaking in several instructions, it will be unattainable to tunnel 2 pitches launched from 2 totally different arm slots. Nevertheless, even when 2 pitches are launched from the very same level, it does no good if one is a spiked curveball and the opposite is a fastball that sails over the catcher’s head, so the two pitches should additionally journey comparatively related paths for so long as potential to deceive the hitter. Moreover, if the two pitches don’t find yourself doing something otherwise motion smart, then that additionally defeats the aim of the pitches being tunneled, because the ball leads to the identical place regardless of which pitch is thrown.

This all appears fairly superior at first look. Throw 2 pitches down the identical path and the hitter is totally fooled proper? Properly possibly if all pitchers have been robots and have been able to doing that everytime then sure, pitch tunneling can be fairly foolproof. Nevertheless even within the large leagues, no pitcher hits his spot wherever close to everytime, and even when they did, a number of instances hitters merely sit on sure pitches based mostly on that pitcher’s tendencies and can take something that isn’t what they’re on the lookout for. As well as, 2 pitches may very well be completely tunneled however be balls, which once more defeats the aim of the observe. So does pitch tunneling truly work outdoors of idea? Let’s reply that query with slightly little bit of statistical evaluation.

First, we want a metric to measure how effectively 2 given pitches are tunneled. As I stated earlier, pitch tunneling primarily boils down to three components: launch level, flight path, and plate location. An incredible instance of a primary metric combining these 3 components is the Tunnel Rating which is printed on this nice article. Tunnel Rating takes into consideration the distinction in launch level between 2 pitches by calculating the euclidean distance between the two factors of launch. A better distance will lead to a decrease rating and vice versa. Tunnel Rating then makes use of the ratio between the distinction in location of the two pitches with motion and the distinction in location of the two pitches with out motion to see how related they have been early of their flight and the way totally different they have been on the finish of their trajectories. Which means that a decrease distinction location with out motion and the next distinction with motion will lead to the next Tunnel Rating and vice versa. With every issue accounted for, the entire formulation for the Tunnel Rating metric is: (precise distance tunnel distance) – launch distance with precise distance being the distinction within the pitch areas with motion and tunnel distance being the distinction between the areas with out motion. 

There are 2 adjustments I’ve comprised of the unique formulation. The primary being that I’ve chosen to measure precise distance and tunnel distance in inches as an alternative of ft. It’s because I imagine {that a} smaller change in motion must be magnified within the tunnel rating calculation, whereas launch distance ought to nonetheless be measured in ft as a result of a small change can have minimal results on the tunneling impact of the pitch. The second change I’m going to make is I’m going to scale Tunnel Rating to 100 to make it into Tunnel Rating+, that means that the typical rating will grow to be 100, a rating of 115 can be 15% above common, and a rating of 85% can be 15% beneath common. This won’t solely make it simpler to grasp what is an efficient or dangerous rating, however it would additionally make it simpler to make additional calculations later within the evaluation.

Subsequent, we want a metric to grade the standard of every pitch itself from a non-tunneling standpoint. For this we’ll merely use the run values of every pitch as supplied by Baseball Savant. In doing this, we get a measure of how good every one among a pitcher’s pitches are by itself, not sequenced with one other pitch. I’m additionally going to scale the run values to 100 for every pitch sort with a view to put them on the identical scale as Tunnel Rating+.

To make use of these 2 metrics to grade the effectiveness of tunneling in including worth to the pitches which can be sequenced collectively, I’ll first randomly choose 100 pitchers with a minimal of three totally different pitch varieties thrown, and 200 pitches thrown this season. This may give me sufficient totally different mixtures of pitch sequences in addition to a bunch of pitchers chosen with out bias in order that I’ll theoretically have some pitchers who tunnel very effectively, and a few who don’t, in addition to some who’ve nice particular person pitches and a few who don’t. I’ll then calculate the Tunnel Rating for every 2 pitch sequence they threw all through all the season, and get the run worth of every pitch they threw this season. Then, to check the results of tunneling on the worth of the pitches, I’ll test the correlation between the units of values, on the lookout for a sample indicating a rise in run worth with a rise in Tunnel Rating, or the opposite approach round. Now it is time to check out my outcomes and see if tunneling had a constructive impact on the worth of the pitches. 

My random choice of pitchers labored as meant, as I had some names like Jack Leiter and Kenta Maeda who’ve abysmal run values for his or her Fastballs, in addition to some names like Zach Wheeler and Cade Smith, who had very excessive run values on theirs. The most typical sequences of pitches have been Fastball-Changeup, Sinker-Slider, Fastball-Curveball, and every of these pair’s reversed variations, with the best Tunnel Scores coming from Sinker-Slider and Fastball-Curveball mixtures. This is sensible, as every of these pitch sort pairs (apart from Fastball-Changeup) include Fastballs and Breaking Balls which transfer in several instructions completely. The shock, nevertheless, got here within the last outcomes of my evaluation; the check of the correlation between the Tunnel Scores and Run Values. Whereas I anticipated there to be a considerably robust correlation between the two, that means that I anticipated profitable tunneling to extend the values of the pitches tunneled, I discovered that there was truly virtually no relationship in any respect between the two with the correlation between them being simply 0.07. Which means that regardless of whether or not a sequence of pitches was tunneled very effectively or very poorly, the run values remained largely the identical. So why is that this?

There are a selection of the explanation why tunneling didn’t contribute a lot to adjustments in run worth. For example, it is vitally potential that the tunneling did work on some events however didn’t work on others, thus balancing out the connection. As tunneling requires such high-quality command of each motion and launch out of the hand, I discover it laborious to imagine that even the very best arms in MLB may do it constantly sufficient for it to make a lot of an influence. One other drawback with tunneling is that it usually leads to a number of pitches being thrown in the identical location repeatedly, which provides a bonus to the hitter, particularly if they can work out the sequence by which the pitcher is making an attempt to tunnel his pitches. Nevertheless, the largest cause for my part why there wasn’t a powerful correlation is that tunneling will finally by no means outweigh the standard of the pitch itself. Even when 2 pitches observe the identical trajectory after which break free from one another, completely tunneled, if the pitches are sluggish, poorly situated, or have poor particular person motion profiles, then the values of the pitches will nonetheless be low.

In conclusion, I imagine that whereas the idea of tunneling sounds nice, it merely isn’t a possible deception observe by way of constantly having the ability to carry out it at a excessive degree and, even when perfected, won’t ever be extra essential than the standard of particular person pitches themselves. Clearly, extra analysis can be completed on tunneling by followers {and professional} organizations alike, however as of now there isn’t a particular profit in making an attempt to carry out it, and when it’s completed efficiently, it’s extra probably a coincidental phenomenon than a repeatable sample.