Looking Beyond the Defensive Spectrum

When I think of the defensive spectrum in baseball, sometimes I picture a baseball diamond-shaped table disrupted by constant tremors. In the center of that table are many orb-like objects (marbles, beans, very small rocks, whatever) which represent the players, like tokens on a board game. While all those objects start toward the center of this peculiar table in my mind’s eye, those relentless tremors slowly but persistently disperse the table’s contents. Some tokens tend to settle at the center of the table and some toward the margins of the table. Others fall off it entirely.

This admittedly strange conception of the defensive spectrum appeals to me because it highlights the non-linear nature of that spectrum. As players ascend the professional ladder, players increasingly are relegated, not necessarily down a linear pipeline (although the defensive value at all positions can, of course, be order-ranked), but to the margins: shortstops are edged over and become third basemen, centerfielders shift over to left or right, those who started at the edges might not earn the jump to the next level of professional baseball at all.

What isn’t discussed so often is what I will refer to as the pitching spectrum. Position players are far from alone in taking less demanding roles as part of their rise through the professional ranks – or in an effort to elongate their careers. However for pitchers, I would argue that things do tend to be a bit more linear. Most often, that spectrum is made up of the varying amounts that one gets to pitch.

While the most skilled position players probably played as shortstops or catchers or centerfielders with amateur teams, only to be pushed aside in the face of new and heightened competition upon reaching pro ball, the vast majority of pitchers were likely one thing as amateurs: starters.

Not everyone can be a starter, though. Erratic control might mean sliding down the spectrum to the bullpen, and so too could injuries or limited pitch mixes or fastballs that tend to play up in shorter bursts. As pitchers work their way up through the minor leagues, those pressures appear at every turn. In scouting reports, starting pitchers are dinged specifically for profiles that feature “bullpen risk.”

That position isn’t really unfair either; those pitchers who can manage to continue starting games longer into their professional career tend to have higher ceilings in the majors. The scatterplot below illustrates this point.

In order to quantify and visualize the pitching staff spectrum, data were pulled from FanGraphs and thereafter joined in order to juxtapose the roles of pitchers over the course of their minor and major league careers. On the minor league side, data at FanGraphs goes back to 2006, so data featured comes from the 2006-2022 seasons.

Getting at the heart of the pitching staff spectrum, the scatterplot below compares the percentage of games pitchers start in MiLB versus MLB.

Generally speaking, those who start a greater fraction of their games in the minor leagues go on to start a greater fraction of their games in the major leagues. However, the concentration of data in the lower right of the above scatterplot relative to the upper left indicates that pitchers tend to start a smaller percentage of games in the major leagues than they did in the minor leagues. That is the pitching spectrum in effect: pitchers talented enough to start most games in, say, AA might not be so talented as to start so often at the game’s highest level – they slide down the pitching spectrum to the bullpen or, worse yet, fall from the roster entirely.

Beyond visualizing, it would be helpful to quantify this phenomenon. Below is a table that aims to do just that; it provides the median percentage of games started as each minor league level (starting at full season single-A) among those pitchers who eventually pitched at least one game in the major leagues. Worth noting is that this data features survival bias: only pitchers who made it to the major leagues at one point are included.

This table is quite telling. Just over half of all pitchers who appeared in MLB from 2006-2022 started a single game, as evidenced by the median percentage of games started being <1%. However, rewind to single A, and the median pitcher (who went on to pitch at least once in MLB, remember) was starting in more than 4 out of 5 of the games they appeared in. This median figure drops precipitously with the leaps to AA and then to AAA: clearly, lots of pitchers are descending the pitching staff spectrum even as they climb the minor league ranks to the major leagues.

Medians can be misleading though, and distributions can be a bit more telling. As evidence of this, below is a histogram of the distribution of % Games Started (for eventual major leaguers) at AA.

This distribution is extremely bimodal: it has two peaks at either end of the spectrum. Put another way, plenty of pitchers start exactly none of their games or all of their games at AA. Still, the median pitcher, who reached the major leagues by some point in 2022, was starting 3 out of 5 games at this level. Distributions for the other minor league levels look similar to this one, although the percentage of players who threw 100% of their games as a starter is higher the lower the minor league level, generally.

While the pitching staff spectrum is prevalent among pitchers who ascend to MLB, that does not mean there aren’t exceptions to the rule. As a final view, those players who most dramatically bucked the trend are ranked.

Luis Castillo, who has started every single game of his major league career to this point, only started about a third of the time throughout the minor leagues. Nate Robertson and Brandon Beachy are the only other pitchers included in this dataset that had a difference of >50% between their % games started in MLB versus MiLB. Still, there is something hopeful about those players who come to pitch the most only upon reaching the major leagues.

Ascending to the major leagues is exceptionally difficult, it should go without saying. In order to get there, players contend with ever-intensifying competition among their peers. As the supply of roster spots shrinks, demand for those best suited to offer the most innings spikes. Some players are fortunate and talented enough to position themselves at the top of the ladder (or center of the table, if you will) until the very end, but for others it is beyond impressive to have found a spot on the ladder at all.