Complete Games in Decline, but Disproportionately So

It’s no surprise that the amount of innings that starting pitchers throw each year is on the decline. It has been covered already here, in fact. That prior post emphasized an innings threshold that fewer and fewer starters are surpassing on an annual basis. This one focuses in part on this site’s namesake: the complete game. 

The following data will primarily be presented visually and should offer a glance at the declining state of complete games in baseball today. This first figure ranks which teams, beginning in the 2000 season, have benefited most and least from the complete game:

There are a couple key takeaways. The first point that sticks out is the Padres. Once you’ve mentally stacked the Rays/Devil Rays and Nationals/Expos data points, it is clear that the Padres as a team have had far and away the fewest amount of complete games this century.  In fact the Rays, circa 2008, have more through this point in 2019. It’s at least fairly interesting to see from the Padres given that their ballpark is pitcher friendly and would supposedly be fairly conducive to more 9-inning outings. 

On the other end of the spectrum, the Blue Jays and Indians have received more complete game efforts from their pitchers by a fair margin. That said, this phenomenon is more likely due to a few superstars than an organizational philosophy. In the 21st century, it doesn’t take many to skew CG data. From 2000 on, Roy Halladay alone pitched 47 (!) complete games for the Jays. For his career Halladay had 67 complete games, more than the Rockies, Brewers, and Padres organizations have this century.

In Cleveland, two names come quickly to mind: Corey Kluber and Cliff Lee. Lee started and completed 10 games for the Indians; Kluber has 17 to his name. If you remove these three names, the Indians and Blue Jays recede comfortably into the middle of the pack. 

Next, team totals have been broken down by season:

Here, we can more broadly see the considerable decline in complete games dating back to the year 2000. In the early 2000s, years went by without a single team going a full season without a complete game. Now every year since 2015, at least one team hasn’t had a complete game, and more often many more than just one. 

The 2011 Phillies rotation threw the most complete games in a season with 18 cumulatively. That is more than 10% of the games in a season, a fairly wild total in today’s game. It is less surprising when one considers that both Roy Halladay and Cliff Lee were a part of that vaunted rotation. 

Fast forward, and 2018 was the first year this century in which no team’s starters together pitched more than 5 complete games. Is this due to the preponderance of games started by Openers in 2018? While it’s tough to make any causal claims, we can clearly state that Openers, by definition, drain the pool of potential games wherein a single pitcher can go the distance.

Next, team data has been rolled up to feature cumulative complete games by year. As you can see, the decline is sharp.

While there is no denying the wider trend, there are exceptions to the rule. From 2007-2011 complete games had a half-decade renaissance, growing each season. 2010, which some dubbed The Year of the Pitcher, feels long gone now though as complete games have continued their recession. While 2019 is on pace to feature more complete games than 2018, it won’t likely eclipse that total by much.

So far, only complete games have been considered. This next figure is near identical to Figure 3, but features only shutouts.

Alas, the trend for shutouts, as one might expect, roughly mirrors the trend for complete games. However, shutouts are tapering at a slower rate than complete games of all varieties. This last figure illustrates shutouts as a percentage of all complete games.

For the first time this century, shutouts are threatening to make up 60% of all complete games. For the third time in eight years, shutouts might be rarer than complete games wherein at least one run was allowed. 

What’s motivating this trend? That is tough to say, but here is some wildly ad hoc conjecture via anecdotal evidence: three times this year Justin Verlander has gone 8 innings while giving up a single run by way of the solo shot. I wonder if any of those games are outings he might have been allowed to finished if a home run hadn’t taken the shutout off the table. In my eyes, it is conceivable that shutouts still have enough allure that managers are willing to let their starters go the distance, whereas “typical” complete games do not quite justify running a starter out there for so long. 

Whatever the case, each statistic is on the decline, simply at differing rates. That said, there can be no shutouts without complete games, so this trend will be worth watching as either type of pitching performance continues to be squeezed.