• Yoenis Cespedes pinch-hit a game-tying, first-pitch three-run home run (on one leg),
• Neil Walker tie the Mets record for most home runs in April,
• Bartolo Colon set the record for oldest (and greatest) player to hit his first career home run,
• Cespedes cap off a 12-run inning with a grand slam,
• Noah Syndergaard hit two home runs against Kenta Maeda,
• and, of course, I’ve already documented the team’s early season back-to-back binge.

In fact the Mets currently lead the league in both most home runs hit and fewest home runs allowed (HRA) which made me curious to see how many teams have ever done that for a full season and what kind of chance the Mets have of doing it this year.

If home runs and home runs allowed were completely independent, you would expect a team to lead the league in both approximately once every 30 (the number of teams in the league) years. There are many reasons why they are not independent including the biggest factor, the designated hitter. The DH makes it much easier for an American League team to lead the league in home runs and for an National League team to lead the league in fewest home runs allowed. In fact, since the introduction of the DH in 1973, 37 AL teams have led (or tied) the league in home runs while only seven NL teams have done so. Conversely, 39 NL teams have led (or tied) the league in home runs allowed compared to only four AL teams. Right away the chances of a team leading the majors in both categories drops dramatically from that theoretical one team every 30 seasons.

Of course, we don’t need to rely on theory since it’s easy enough to figure out which teams have led the MLB in both home runs and home runs allowed. The most recent team to accomplish it was the 1906 Philadelphia Athletics who hit 32 home runs and allowed only nine. Seeing that no team has led in both statistics for 110 years makes me a bit less optimistic about the Mets chances.

Teams who led MLB in HR and HRA

If no team has led in both measure since 1906, that made me wonder which teams have come the closest since then. I first measured this in an absolute sense by combining the number of home runs behind the league leader and the number of home runs allowed above the league leader for each season. The list below is mostly composed of teams from a vastly different era than today. (Of the 11 teams who trailed by single digits, nine of them were in 1918 or earlier.) However, two teams stand out: the powerhouse 1927 Yankees and our very own 1988 Mets. Each of these teams missed leading the league by six total home runs with the Yankees leading in home runs but allowing six more than the league leader and the Mets leading in fewest home runs allowed but trailing the league leader by six home runs.

Teams Closest to Leading MLB in HR and HRA (Raw Totals)

The above chart is clearly biased towards older teams who played in an era with significantly fewer home runs. The 2016 Mets, a mere month into their season, have already hit more home runs than eight of the teams on the list. With such low home run totals, it isn’t surprising that older teams dominate this measure. To correct for this I also looked at teams by their percentage back. I calculated this by taking the combined number back and dividing it by the league leading number of home runs hit plus the fewest home runs allowed that year. This denominator would allow a team who led in both statistics would finish 0% back, which makes sense. The 1988 Mets jump to the top of the list when sorting teams this way.

Teams Closest to Leading MLB in HR and HRA (%)

I’ve established that the designated hitter is a huge impediment to a team leading MLB in both home runs and home runs allowed. One way to even the playing field is to look at the leaders broken down by their respective leagues. With consistent rules and fewer teams in each of these leagues we’d still expect a 1/n (where n is the number of teams in the league) probability of a team leading in each of those measures in a year, assuming independence.

There are, however, other factors beyond the designated hitter which make it less likely for a team to lead in both statistics. The main one which I can think of is the stadium which the team plays in. A team which plays in a hitter friendly park is going to have a higher chance of leading in home runs and a lower chance of leading in home runs allowed. Since 1960, only one team has led the American League in both measures: the 1994 Cleveland Indians. In the same time span five teams have led the National League in both, most recently the 2001 San Francisco Giants. I can’t think of an obvious reason why an NL team would be more likely to lead both so I’m going to chalk this up to chance for now (if you can think of any good reasons feel free to share them in the comments). The chart below contains all team 19 teams from the Lahman database who led their league in both home runs and home runs allowed.

Teams who Led the AL or NL in HR and HRA

I started this article wondering if the Mets had a chance to do something special this year. The numbers I’ve looked at so far have served two purposes–they have made it seem way more special for a team to lead the league in both home run measures and they have made me significantly more pessimistic that the 2016 Mets have a legitimate shot at doing so. What they haven’t done is give me an estimate of just how possible it is.

For that I turned to PECOTA, BP’s projection system. PECOTA does not do its own estimates of how likely a team is to lead the league in a particular stat but it does provide estimates (in this case from 5/8/16) for home runs, plate appearances, home runs allowed and total batters faced over the rest of the season. I decided to write a very simple simulator, one which takes the projected plate appearances for each team, gives each of those plate appearance that team’s projected chance of hitting (or allowing) a home run and adds each team’s year to date home runs to those totals. I simulated 100,000 seasons for each team (both batting and pitching), keeping track of how often they led the league in each home run measure. While we’ve established that these are not independent PECOTA takes into account things like the DH and park factors, so I believe that a team’s chance of leading the league in both measures is accurately stated by multiplying their probability of leading in each of the two measures.

Results of 100,000 simulations based on ROS PECOTA projections

The Mets, being a National League team with strong pitching, have a much better chance of leading the league in home runs allowed than they do in home runs hit. That being said, the lineup is filled with power hitters (although it took a hit losing Travis d’Arnaud to his shoulder injury) and stranger things have happened. In a world where Bartolo Colon can hit a homer at Petco Park I’m not going to totally write off the Mets’ 0.2 percent chance of doing something which hasn’t been done in over a century. If extreme long shots aren’t your thing you can take solace in knowing that the 2016 Mets have a much better chance (about 15 percent) of joining the 1988 squad in becoming the home run kings of the National League.

Thanks to Rob McQuown of Baseball Prospectus for research assistance and for providing PECOTA projections. Thanks to Coral Marshall (@Coral_Rae) and Jared Cross (@steamerpro) for invaluable input and discussions about the topic.

You can find the R code which I used to prepare the Lahman data for the above retrospective analysis here.

You can find the Python simulator which I used to generate the probabilistic predictions here.

Photo Credit: Anthony Gruppuso-USA TODAY Sports

My first instinct was to ground myself in some sense of how rare it is to hit back-to-back home runs–for the record, that’s just when two consecutive batters homer. This could be measured as simply as how many pairs a team usually hits in a season or how many games or plate appearances teams usually go between them. I figured that a good way to estimate this would be to take the probability of a home run being hit in a plate appearance and square it. This should represent the probability of home runs being hit on consecutive plate appearances (if home runs were independently and identically distributed throughout the season). I knew that actual back-to-back home runs would be more frequent than this (due to factors like lineup construction, pitcher skill-level differences, ballparks and weather) but it would give me a baseline to start from.

In 2015, the league hit 4,909 home runs over 183,705 plate appearances. That means that a random plate appearance from the 2015 season had a 2.7% chance of being a home run, which roughly equates to a .07% chance of consecutive plate appearances each yielding a home run. To put it another way that’s one pair of back-to-back home runs every 1,400 plate appearances. The Mets, on the other hand, had just hit four pairs of back-to-back home runs in approximately 200 plate appearances. This radical discrepancy made me even more interested in investigating how often teams actually hit back-to-back home runs and how many teams have had a streak like the one the Mets had.

I loaded up my Retrosheet database and searched the play-by-play data for every time batters hit back-to-back home runs. It turns out that in 2015, teams hit 136 pairs, one every 1351 PA. This means that they occurred about four percent more frequently than my rule of thumb calculation would have predicted. It turns out that 2015 was a bit of an outlier in this regard, as the average year between 1974 and 2015 has featured about 31 percent more back-to-back home runs than the rule of thumb would predict, as shown on the chart below.

The 136 back-to-back home runs from 2015 means that teams averaged just over 4.5 pairs of back-to-back home runs last year. The Mets have averaged 3.7 pairs of back-to-back home runs per season since 1974. Three times since 1974, the Mets didn’t hit a single pair of back-to-back home runs (1977, 1978, 1980), which means they once went four consecutive seasons (1977-1980) with just one pair of back-to-back dingers. No wonder this statistical quirk jumped out at me–the 2016 Mets hit just about a year’s worth of back-to-back home runs in only six days. The chart below shows the 12 seasons where the Mets have had the most back-to-back home runs. The 2016 squad seems very likely to end up somewhere on this list.

Next, I decided to look at the best overall six-day (or fewer) periods I could find. Unsurprisingly, only 18 teams have had four or more pairs of back-to-back home runs in such a short period of time since 1974 (including the 1995 Rockies, who did it twice). I also found five teams* who did it before 1974 … including one team, the 1964 Twins who also managed it twice. I’m not including overlapping windows; I’ve filtered the list to include the earliest period with the maximum number of back-to-back home runs for each team during each period of time. For chains longer than two, each pair of consecutive home runs is counted, so back-to-back-to-back (B3B in the chart below) counts as two pairs of back-to-back home runs.

* – incomplete play-by-play data means there could be others

Incredibly, two teams managed to hit five pairs of back-to-back home runs in only three days. The 2002 Cardinals and the 1996 Dodgers share this distinction. Each of those teams played the first two games of their record-setting spans in Coors Field, which definitely helped. The two teams hit a total of 27 home runs in the six games that make up their respective streaks.

Some interesting numbers can be gleaned from this chart. The average number of back-to-back home runs in a season by a team on this list is just over 12, well above the league average of four–even if you remove the ones included in the qualifying time spans. The 22 teams on the list averaged 195.2 home runs. A number of Mets fan favorites contributed to these streaks including: Juan Uribe, Moises Alou, Carlos Delgado, Jeff Kent, Fernando Tatis, and Mike Piazza. Piazza was the driving force in the 1996 Dodgers’ record-setting streak, hitting four of the nine total home runs involved. Neil Walker was part of the two most recent teams on this list, contributing two of the home runs when the 2014 Pirates hit four pairs of back-to-back home runs in six days. Walker, of course, also hit three of the home runs during the Mets’ streak this year.

The record for the most pairs of back-to-back home runs in a season is 19, set by the 1996 Mariners. The rest of the top 10 can be seen on the chart below.

I had a feeling that the Mets’ barrage of back-to-back home runs was an unlikely and overlooked story. These rare events always stand out to me even if they aren’t entirely unique or as celebrated as something like hitting for the cycle or throwing a no-hitter. Baseball provides us with a huge number of events each year which means that a lot of rare and interesting things happen every season. It’ll be fun to keep track of how many times the Mets go back-to-back this year and to see if they can surpass the franchise record of 13 set by the 2000 team. This lineup is deep and heavy on power, a good combination to keep piling up back-to-back home runs.

Author’s Notes

I thought people might enjoy seeing the sql commands which I used to fetch the data for this article. I’ve put them in a publically accessible git repository which you can find here: https://github.com/crglaser/B2B_article. I used PostgreSQL database which I set up using the data file provided on the py-retrosheet github which can be found here: https://github.com/wellsoliver/py-retrosheet.

Between the time this article was written and published the Mets went back-to-back an additional time. In the first inning of the 5/2 game against the Braves Lucas Duda followed up Yoenis Cespedes’s home run with one of his own. It was the Mets fifth back-to-back of the season, leaving then only one shy of joining the list of most back-to-back home runs in a season above.

Thanks to Rob McQuown of Baseball Prospectus for research assistance, and a big thank you to Andy Dolan (@adolan73 on Twitter) for his help in finalizing and verifying many of the numbers in this article!

Photo Credit: Anthony Gruppuso-USA TODAY Sports