![]() ![]() ![]() Let’s look at a chart from the perspective of the three metrics listed above. Maximizing swings and misses (we’ll measure this with Swing & Miss %).Maximizing called strikes (we’ll measure this with Called Strike % on pitches taken).Limiting quality of contact (we’ll measure this with wOBAContact).Pitchers, when they choose the location of the pitch are optimizing for three outcomes: This author is a big fan of simplicity, so we’ll be taking all the pitch locations and summarizing them to the above metrics (distance from the center, zone yes/no), before we propose a slightly more nuanced version. We could also incorporate directionality into our model (inside vs. Now, we could make that a little more complex and separate the vertical and horizontal components of distance. Was the pitch in the probabilistic strike zone?.How far away from the center-center of the strike zone was the pitch?.Which arbitrary rectangle of the strike zone a pitch was thrown to isn’t what’s important. Further, as we discussed in part one, the corners are actually bad spots to throw to if you’re looking for a called strike, so they operate differently than the horizontal or vertical edge would, at the middle of the zone. Those pitches are just as difficult to hit as the ones that are on the square edge. If we look at the edge, we’re leaving out a large amount of surface area that is outside of the larger circle, but within the dotted square. Why are we using a pixelated methodology, when we can draw circles? The point on the far left of the inner circle, at the center of the vertical axis, is just as close to the center-center of the plate as any of the four corners in the square heart. If we constrain ourselves to using the square heart, we’re leaving out a lot of surface area that is equally as far away from the center of the zone. Let’s talk about the heart of the plate first. The inner circle surrounds the central square, which we’ll nominally refer to as the “heart” of the zone. The area between the outer square and the inner blue dotted line we’ll call the “edge” of the zone. We see a basic three-by-three grid, with a couple of circles. For visual clarity, we’ve drawn this as a square for the initial discussion. In part one of this series, “ It’s not a Square,” we redefined the strike zone as a superellipse, rather than a rectangle. It’s time to start thinking in terms of concentric circles. It’s time to stop thinking in terms of grids of squares and rectangles. Squares and rectangles have been silently misleading you for years. The main purpose of drawing a grid is to give you a sense of how far away the pitch is from the middle of the zone. However, this is not practical with the strike zone. We can of course draw a lot of boxes and increase our resolution, as we do with modern television. The four corners of a square are much farther away from the the center of the square than the midpoints of the vertices. The core flaw in the current models is that squares are an inaccurate shape to classify how far away a pitch is from the center of the strike zone. Today, we’re addressing a different concept, specifically our tendency to divide the zone into boxes. Other important strike zone research includes (but is certainly not limited to) Rob Arthur’s initial assessment of StatCast’s accuracy, as well as the Baseball Prospectus team’s detailed proposal for a universal strike zone, and its elucidation of the measurement challenges attendant with and the implications of a robotic strike zone. The very talented Bill Petti employs this standard approach on his excellent Edge % web app today’s article is in many ways an evolution of Bill’s Edge % work. This author wrote an entire article, wherein the basis of the research was where a pitch was located in an arbitrary square or rectangular zone. You might find some with slightly more complex shapes, with “L” shaped corners. Some have larger “hearts” where the inner square is larger others are symmetrical grids. If you browse any modern baseball website, such as FanGraphs, Baseball Prospectus, MLB.com, and ESPN, you’ll find that they all depict the strike zone as subdivided into some form of a grid. ![]() Let’s reimagine the strike zone, but not as a grid.įor many years, you’ve been led to believe that the strike zone should be divided into boxes, little boxes made of ticky-tacky arbitrary zones, little boxes in the strike zone, little boxes all the same. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |