On the 2017 MLB Records: Home Runs, Strikeouts, Complete Games, Relievers, and Time

The 2017 Major League Baseball season saw several records set — not season or career records for individual players, but rather records for MLB as such. The one that has understandably gotten the most attention is regarding Home Runs. Overall, more HRs were hit this year than in any season in MLB history — yes, even more than during the height of the PED-era. The grand total was 6,105 which is significantly more than the 5,610 hit last year, and truly an explosion compared to 4,909 in 2015 or the relative recent low of 4,186 in 2014. The previous record was 5,693 in 2000.

That said, because there are so many more teams now than there were in decades past, to better compare HR trends over time it is helpful to consider the number of Home Runs hit per game played by each team (HR/G). Here also a record was set this year as there were 1.26 HR/G, compared to 1.16 HR/G last year — which itself nearly matched the record of 1.17 HR/G in 2000. To put these recent numbers in perspective, the range from 1950-1989 was a low of 0.61 HR/G in the pitching strong year of 1968 to a high of 1.06 in  hitting-heavy year of 1987. Before that, from 1921 to 1949 the range was from a low of 0.35 to a high of 0.69 HR/G. And before that, during the Dead-Ball era, the range was 0.10 to 0.26 HR/G.

The burst of HR in recent years, most notably the past two seasons, has not been accompanied by a similar record setting pace for other traditional offensive metrics, such as Batting Average or even Runs scored per game. In both 2017 and 2016 the overall Batting Average in MLB was .255, which is about the average for the past eight years. This was noticeably lower than the range from 1993 to 2009, which was .261 to .271, not to mention the high seasonal averages from the 1930s, which ranged from .275 to .296.

For Runs per Game, the 2017 season saw an average of 4.65 Runs per Game (for each team). That is higher than any season since 2008, but it is far from the modern (since 1901) record of 5.55 set in 1930, or the more recent high mark of 5.14 in the 2000 season.  Other offensive stats didn’t set such overall MLB records this year either — not Doubles, Walks, OBP, or any other standard statistic.  Except one – Strikeouts. With so many hitters swinging for the fences, there was also a record set for the most strikeouts: 40,104 which was more than the 38,982 strikeouts in 2016. That works out to 8.25 strikeouts per team per game in 2017, which is more than the 8.03 in 2016. It seems feast or famine is the name of the game these days.

But as this post’s title notes, this season did set several other interesting all-time MLB records. Obviously, if hitters struck out more than ever before, then pitchers get credit for striking out more hitters too. The 2017 season also saw an all-time low in Complete Games pitched (CG) — yes, even including the strike-shortened 1981, 2004, and 2005 seasons. This continues a recent trend as 2015 set an all-time low for CG with 104, then 2016 broke that with 83, and 2017 shattered that mark with a grand total of only 59. As recent as 2003 there was 209 CG, in 1992 there was 419 CG, and in 1978 there were 1,034. Again, to better compare we need to look at CG per G, which in 2017 was a paltry 0.012. In 2003 this ratio was 0.043, in 1992 it was 0.099, and in 1978 it was 0.245.

All baseball fans know that relief pitchers are used more now — and are more specialized — than in decades past. But while the rate of complete games did gradually decline from the 1920s to the 1970s, it was a slow decline. What we have seen in recent years is a sharp drop off — the leaders this year were Corey Kluber and Ervin Santana who had 5 complete games each. No one else had more than 2, and only six pitchers even had 2 (that is 8 with 2+ CG in 2017, compared to 16 with 2+ CG in 2016).

So it seems that the “Complete Game” for a pitcher is in rapid decline. But a couple of related records were also set in 2017. More pitchers were used this year than ever before, to the tune of 755 different pitchers, compared with 742 last year, 735 in 2015, and a range of 606 and 692 from 2002 to 2014. On a per team, per game basis the recording setting number in 2017 was 4.22, compared to 4.15 in 2016, 4.11 in 2015, and a range of 3.63 and 3.99 from 2002 to 2014.

Having more pitchers could mean the very best pitchers are used less often, and this could be one factor in the surge in HR allowed. But it doesn’t explain why there are more strikeouts than ever before, nor why batting averages and other offensive statistics look fairly normal (and in some cases rather low) compared to the past few decades. So a lot more analysis would be needed to connect the HR surge on the use of more relief pitchers, especially since those are fresh arms and are often focused on just getting specific hitters out or at most going one or two innings only.

One thing that having so few complete games, and having so many more relief pitchers used, definitely does correlate with: a new record in 2017 for the average amount of time for each game. (This is because each time a relief pitcher comes in they have to run in from the bullpen, pitch a few final warm-up pitches, etc.) This past season’s average game clocked in at 3 hours, 8 minutes, which was barely longer than the previous high of 3 hours, 7 minutes in 2014. From 1987 to 2012 the average length of a game ranged between 2 hours, 49 minutes and 3 hours, 1 minute. Games from the mid-1950s through the 1970s averaged only about two and half hours, and prior to 1934 games averaged less than two hours each.

I’ll note that the modern issue with the length of games is not due to extra-inning contests. In fact, 2017 tied the mark (with the 2005 season) for the lowest percentage of extra innings games since 1948. So that is clearly not the culprit.

One factor in the increasing length of games is of course the introduction of instant replay in 2014. The average time of games went up by about 3 minutes per game that year. New commissioner Rob Manfred instituted some changes in 2015 — notably including a countdown clock in between innings and a rule that prevents players from stepping fully out of the batter’s box. Those changes helped, as the average game length dropped by seven minutes in 2015. But then it crept upwards in 2016 and again 2017, even with another change coming from the Commissioner’s office: allowing a manager to simply signal to an umpire when an intentional walk was desired, instead of going through the pointless and time-wasting exercise of having the pitcher throw four far-outside pitchers. The impact of this change was not huge though, especially as the use of intentional walks has been trending downwards in recent years, with the 2014-17 seasons having the lowest IBB rates since records started for this stat in 1955.

While instant replay clearly adds some time to baseball games, the surge in the use of relief pitchers — as evidenced by the rapid decline in complete games thrown, and the number of relief pitchers used per team per game — is likely the biggest factor in MLB games setting new highs in average game time length. For any given team in any give season there can be many reasons for the need to use more relievers: a weak starting rotation, a relatively strong set of relief specialists, and injuries or potential injury concerns (i.e., limiting pitch counts for starters and now even some relievers). But the ever increasing use of advanced analytics and more data than we’ve ever seen before is likely also leading to managers and pitching coaches to make the call to the pen more often, and thereby — unintentionally — adding to game time even while the Commissioner’s office is actively trying to decrease game time.

In the October 4 issue of the Wall Street Journal,  Brian Costa and Jared Diamond wrote an article titled “Baseball Learns Data’s Downside” ( available here , but requires subscription), where they make a many of the same points regarding this game length issue. One additional data point they noted is that relief pitchers take a bit longer than starters, on average, between pitches — about 1.5 seconds longer per pitch. Fans and experts have long debated the relative value of strikeouts compared to ground outs and fly outs, but one thing is known: strikeouts on average take longer to occur. And Costa and Diamond note that just as 2017 broke the record for the highest strikeout rate, so too it broke the record for the most pitches thrown per batter (3.9).

I don’t like seeing the average length of games getting longer and longer — I love baseball, but I don’t consider longer games due to more play stoppages to be more of a good thing. That said, Costa and Diamond report that earlier this year MLB appointed a 16-member committee comprising  owners, team presidents, general managers and field managers to suggest further changes to address the issue. They say that “According to one member, the committee has explored a range of possibilities that could diminish some of the impact analytics have had on the pace of play. Among the options discussed were a ban on defensive shifts, restrictions on pitching changes, and shrinking the strike zone.”

I definitely don’t like the idea of banning defensive shifts — that is a part of strategy that managers continue to experiment with, and I think it is fun to see it both succeed and fail. And far more people than only baseball purists would fight any change to the strike zone, as that would be a major break from tradition. Given the data and arguments above regarding the impact of increased use of relief pitchers on the length of games, I’d be open to rules that restrict the number of relief pitchers that can be used. I lean against it, but if instituted they would need to be very nuanced and have exceptions for situations like injuries and extra-inning games.

My preference would be for the kind of less dramatic changes that Commissioner Manfred is pushing for. As Costa and Diamond note: “In addition to the pitch clock, he said at a recent news conference, MLB is discussing with the players’ union a limit on visits to the pitcher’s mound and shortening breaks between innings.” Those are changes I could live with, as few would mind shorter inning breaks, a pitch clock seems like a fair change given the restriction on batters in the batter’s box, and I definitely dislike what seem like gratuitous mound more so than the increasing use of relief specialists. If those changes could keep 2018 and beyond from setting new records for the average time of games, then bring them on!