First off a little definition. The complete game is where the starting pitcher pitches the entire game and faces every batter without help from a relief pitcher. A pitcher can pitch the whole game and still lose.

I have started to research some parts of baseball history and found a few interesting things concerning complete games in baseball and how they have steadily decreased over the years. What got me interested initially to think about this subject is all the great pitching performances we have had in the playoffs leading up to this year’s world series.

This post is more about looking at the numbers and how they have changed over time. Likely in a future post, I will delve into why there has been a decline in complete games.

As I started to look for the historical data on complete games for each season I noticed that for the most part, complete games were going down over time. In general, it was a fairly gradual decline to the number of complete games in a season with a few areas that are a more dramatic decline, and some areas where they briefly spiked back up. To dig into this more, first here are two charts to help show what I am talking about.

The charts are interactive if you click on them and they will open up in a new window.

What you can see from both graphs but more easily from Graph two is the steady decline in the number of complete games. From 1904 to 1913 there is a decline of 33%, the most dramatic period of decline in the entire graph.  After that dramatic drop from those nine years, things start to level off and even come back a little bit for a few years. Starting in 1921 we reach a period where for the next 25 years to 1946 where things move at a much more gradual downward slope losing 10%, going from 52% in 1921 to 42% in 1946.

Another chart behind the link.

I would breakdown the rest of the years into four rough brackets.  We have two eras where the declines are not steep but more gradual, 1947-1967 going from 39% to 24% a 15% drop. The grouping of 1968-1979 is the outlier of these four groups, in that the decline levels off and fluctuates up and down from 28% to 22% and has several years stretches where it just stays at 27-28%.

That pattern breaks in our next group with our next drop from 1980 to 1992, going from 20% to 10%. From 1993 to 2013 the present day we are always under 10% complete games a year and have reached roughly the bottoming out of the decline once we reach 2001. From 2001 to 2013 we reach 4% and fluctuate from 2-4% over the next decade-plus 2001-2013.

Moving on to Chart one, most of the major spikes in the graph can be explained. Starting with the two early in the graph in 1884 and 1890, 1884 is a smaller spike but that year the number of teams in the National League and the American Association increased to 21 teams from 16 teams, and then decreased back to 16 teams the following year of 1885. For 1890, again there was an increase in teams from 16 teams the year before to 25 teams and then decreased back to 17 teams for the year 1891. The reason for the increase in teams in 1890 is that a third league came into existence for one year called the Players League that existed alongside the National League and the American Association. The large dip we see in 1900 is because the National League lost 4 teams before the season, with the league dropping down to 8 teams for the 1900 season compared to 12 for the 1899 season and then doubling to 16 teams for the 1901 season.

Our next major spike comes in the two years of 1914 and 1915, with the creation of another try at a third major baseball league in the Federal League that tried to challenge the National League and the new American League. The Federal League was the last baseball league to try to attain major league status that took the field. There was an attempt at creating a third league in 1959-1960 but it folded before games were ever played for it. For the years 1914 and 1915, the number of teams that existed went to 24 teams from the previous 16 that existed before and after the Federal League existed for two years. The American League and National League did not expand in the number of teams until 1961 when they added 2 teams, partially in response to the attempt to create a third league as mentioned, which was called the Continental League during its brief existence.

The last chart below is a zoomed part of Chart one from 1979 to 2013.

What I want to highlight in this last chart is the difference the two baseball labor strikes had on the number of complete games in 1981 and 1994.  The major difference between the 1981 strike and 1994 strike is that the number of complete games rebound from the 1981 strike, and do not dip below the 1981 level of 510 complete games until the 1989 season when there were only 483 complete games. This is compared to the 1994 strike and the rebound after that season. As you can see the rebound after the 1994 season is much more subdued. After the 1981 season, there was a 31% rebound in the 1982 season, and for the 1995 season, there was only  7% rebound after the 1994 strike. The first time baseball falls below the shortened 1994 complete game number is in 1999 where there are 236 complete games for the year compared to the 255 for the 1994 season.

Another way to put it into context is that for every season beginning in 1999 and onward there were fewer complete games than the 1994 strike year even though there was a total of 813 lost regular-season games in the 1994 season. Those 813 games amounted to a loss of 34% of a normal regular season of baseball.

One final point to help illustrate how much baseball has changed is the fact that this baseball record will very likely never be broken unless the game drastically changes. The career complete game record of pitcher Cy Young’s 749 complete games over 22 seasons. For perspective, the current leader for active players is Roy Halladay at 67 complete games throughout 16 seasons so far.

In closing, hopefully, this piece has given you a little more insight into baseball and how the baseball complete game has become a lost art.

The underlying data for this post came mainly from Lahman’s Baseball Database.  Addition data from fangraphs.com and baseball-reference.com