There are two types of labels for starting pitchers: team-dependent and team-independent. Team-dependent labels, as the name suggests, are contingent on variables specific to that team, most notably the skill of the four other members of the rotation. Contrarily, team-independent labels look past roles on individual teams to assess the entire population of major league starting pitchers.
Consider the 2015 Dodgers’ rotation. From a team-dependent perspective, Clayton Kershaw was the number one starter (and perhaps consequently the ace) while Zack Greinke was the number two starter. From a team-independent perspective, both Kershaw and Greinke, as two of the top three pitchers in the National League, were good enough to claim a number one starter label (and possibly the title of an ace). A conflict exists between Greinke’s team-dependent label and team-independent label that is far from unique.
These conflicts represent a broad issue that perhaps we can address from the top down. In what follows, multiple theories about the definition of an ace, which may or may not be the same as a number one starter, will be analyzed from both team-dependent and team-independent perspectives. The goal is to find a suitable theory for what constitutes an ace pitcher.
MORE: Every team's fan base, ranked from most to least miserable
Theory one: The best starting pitcher on each team
The word ace literally means one and has traditionally been used in a team-dependent context to refer to the number one starter on each team. By this theory, there are always exactly 30 ace pitchers who are always defined in a team-dependent context and most easily discovered by checking the probable pitchers each Opening Day.
By this measure, Zack Greinke was not an ace in 2015, but thanks to his move to the Diamondbacks, stands to gain that title in 2016. From a team-independent perspective, from a team-dependent perspective this is simple, easy, and apart from differences in picking the best pitcher on each team, can be objective.
Negatively, and among the reasons why the traditional ace definition is no longer an exclusive definition, are the contradictions arising when applying the team-dependent role to a team-independent context. Continuing with the Greinke example, by this definition the former Dodger was not an ace in 2015 but Jorge de la Rosa, whose ERA was roughly 2.5 times higher than Greinke’s, his held the title.
This theory has some initial merit, but the inability to reasonably compare players in the same role on different teams – not to mention free agents – gives this option a major blow. When analyzing aces, Greinke, as the league’s ERA leader, should be included, while de la Rosa and has no place.
Theory two: The top 30 starters in the league
Consider yourself in an alternate universe where baseball fans and analysts agree on an objective measure of rating and ranking pitchers. By this measure one could objectively line up all 150 current starting pitchers from best to worst, call the top 30 starters aces, the next 30 number two starters, the next 30 number three starters, and so forth.
This method creates a fundamental shift in team-dependent versus team-independent role allocation. Instead of exclusively using the team-dependent label, theory two allows that label to stay, albeit with much less significance, and places the more important distinction in an objective league context.
For example, de la Rosa remains the Rockies’ best starter but is not a top-30 starter in the league and is therefore, more importantly, not an ace. Greinke, meanwhile, may have been the Dodgers’ second-best starter, but as one of the top 30 starters in the league consequently receives the ace label.
Practically speaking, this framework is an improvement but could still faces challenges. Most significantly, the assumption that the top 30 starters are all aces is not necessarily a fair one, as it leads to questionable consequences such as Marco Estrada, Kyle Hendricks, and J.A. Happ possibly receiving the title of an ace (depending on the methodology of the rankings).
Theory three: Better than a number one starter
Perhaps one agrees with the theory two’s principle that aces should be anointed in a team-independent context but disagrees that we should simply pick the best 30. The crux of this argument is regarding a potential difference between the label of a number one starter and an ace. If all number one starters are aces then theory three has no merit, but if an ace is something more special than the 30th best starter in the league, perhaps we must raise the stakes.
There are few quality options for raising the stakes. We should certainly agree that the best pitcher in baseball is an ace, as is the second, the third, and the fourth, but what about the 10th, 15th or 20th best pitcher? Where is the line? This version of the paradox of the heap is unsolvable without arbitrary distinctions, and arbitrary distinctions should not be the guide for answering this question.
Theory four: A statistical threshold
Perhaps we can put rankings to the side and instead construct a statistical threshold to separate aces from non-aces. There are no fixed number of aces in this model, and everyone achieving this level of statistical dominance would earn the title of an ace.
How can we construct an objective statistical threshold? Of the many tell-all single number stats, BP’s DRA-based WARP is my choice as the best of the bunch, but even with this measure, how would we draw the line? Is there a way to establish a line without arbitrarily asserting that an ace is anyone with at least a 4.00 DRA-based WARP?
Another possibility is to propose that aces have certain positive traits that can be measured against league averages. Perhaps all aces throw hard, have good control, or feature three plus pitches. Finding traits that are better than league average gives us a non-arbitrary measuring stick, but this method tends to only reward certain types of pitchers and reward the means rather than the end of being an ace. For example, R.A. Dickey was certainly an ace during his Cy Young 2012 season but only featured one plus pitch, Ubaldo Jimenez has gone on dominant runs in spite of his below average control, and Greg Maddux remained dominant in spite of below-average fastball velocity. While featuring plus velocity, control, and multiple above-average pitches can be very helpful to reach ace status, it is essential to recognize these attributes appropriately as means and not as ends in themselves. Statistical models should not penalize unusually effective pitchers but rather reward overall effectiveness.
Wrap
There are no comprehensive, objective answers to the question of what qualifies as an ace pitcher. Each of the four theories listed above has merit, but each also contains fatal flaws preventing its adoption as widespread definition.
Until a theory gains widespread acceptance or a better theory comes along, the term will continue to be used in a vague manner. There is no objective definition, and unless one of the examined theories is drastically improved or a new theory emerges, it may be a very long time until we truly know how to identify an ace.
Dan Weigel is a contributor at Sporting News focusing on pitching. Follow him on twitter at @DanWeigel38.