History of the doubleheader (Part 1)

We are living in the twilight of the doubleheader. It used to be a common part of baseball, with teams hosting numerous ones per year, but nowadays they are rarely ever scheduled and exist primarily as a way to make up for rainouts.

The rise, fall, and overall history of the doubleheader has been on my mind more than it ought to because for my SABR presentation next month (Searching for the Elusive Sunday Pitcher, to be given Thursday, Aug. 5, at 3pm) I did a considerable amount of research on doubleheaders.

The doubleheader has been around for so long, it’s easy to assume it always existed, and that they always occurred with the same regularity until recent decades made them an endangered species. That isn’t true. Like anything that lasts for well over a century, doubleheaders have had their ebbs and flows over time.

What I’d like to do here over the next two or three weeks at THT is trace the rise, crest, and fall of the doubleheader in MLB history. For starters, let’s look at its origins.

Getting a toehold

According to Peter Morris’ magnificent Game of Inches, doubleheaders predated professional baseball but died out when money came into the game, in part because teams didn’t want to give twice as much away for the price of one game.

When the National Association began in 1871, there were no doubleheaders. Nor were there any the next year. Professional baseball had its first one in 1873, and it would prove to be the only one in the five-year history of the NA. It took place on the Fourth of July, which was fitting because this would quickly become one of the great days for doubleheaders in baseball.

The National League initially continued this reluctance toward playing two games in one day. In its first two years, there was only one doubleheader. Though I can’t say for sure, it appeared that twinbill wasn’t even planned in advance, but caused by weather conditions. If that was the case, that made it the first in a long series of unintended doubleheaders.

It came on Sept. 9, 1876, when the Hartford squad hosted Cincinnati during the latter’s last eastward trek of the season. On Sept. 6, Cincy wrapped up a two game series against Boston. Neither team played on the seventh or eighth.

Doubleheaders didn’t gain a toehold in baseball until the 1880s, when teams started scheduling the double dips of baseball on Memorial Day and Independence Day. This doesn’t mean all teams played on these days, but for a while in the early 1880s, they were the only days anyone played doubleheaders.

Becoming established

Something finally happened in the 1880s to burst the doubleheader out of its holiday ghetto: competition. The American Association started up, challenging the NL’s monopoly on major league entertainment.

At first, the league competition helped solidify the holiday ghetto. Heck, it arguably created it, as there were only two doubleheaders in 1881 (the last NL-only year); both on Independence Day. By 1883, there were 16 doubleheaders: eight on Memorial Day, seven on Independence Day and one in September. In fact, May 30, 1883 has the distinction of being the first time every MLB team played in a doubleheader on the same day.

However, the leagues soon upped their doubleheader totals. If I had to guess, I’d assume the AA led the way in doubleheaders. In my own hazy knowledge of that league, they were the hungrier group, scrambling to establish themselves as equals to the NL. For example, the NL never played on Sunday, as that was the Lord’s Day. The AA played games on Sunday, though. They also had lower ticket prices and had the first clubs to sell beer at the ballpark.

Anyhow, I’d assume the AA would lead the way with doubleheaders, because that fits their hungry, fan-friendly reputation. That isn’t quite how it worked out though. The chart below shows what percent of games in each league were doubleheader games while both existed. (I’ll throw in the Union Association and Players League, to see how they fit in):

Year	NL	AA	PL	UA
1882	3.55%	0.85%		
1883	4.05%	4.10%		
1884	3.50%	3.95%		2.80%
1885	4.93%	3.15%		
1886	6.46%	5.75%		
1887	7.89%	5.82%		
1888	10.68%	8.03%		
1889	16.20%	11.81%		
1890	17.30%	14.07%	12.92%	
1891	11.59%	9.70%	

There’s a key trend of increased doubleheaders. It was like some sort of arms race: let’s see who can make the fans happiest with double ordering of games, without doing too much of it (for fear of giving fans such a good deal that it hurts the bottom line).

That said, the AA was not the driver of the doubleheader explosion. They had the most only twice, and once by only the slenderest of margins. Instead, it looks like doubleheaders were primarily used by the NL to make the fans happy. As the twinbill became more important, the NL established its leading role in organized baseball.

For the first half the 1880s, the games were mostly a holiday thing. From 1886 onward, doubleheaders became more common in September. I really have no idea if these games were primarily about making up rainouts from earlier in the year or attempts to coax out the faithful in the colder weather.

A Hardball Times Update
Goodbye for now.

My hunch is that it’s some of both, but mostly the latter. If it was just early season rainouts, the shape would be different. For example, in 1887 19 out of 36 baseball doubleheaders came in September. But there were only two in August. If it’s just make up games, you would expect a few more doubleheaders in August.

The big upsurge in NL doubleheaders in late 1880s came because the league finally started scheduling doubleheaders throughout the summer. September remained the big month for twinbills, but they occurred throughout the year. Generally speaking, teams would rarely play doubleheaders before Memorial Day but a steady trickle of them went on until the September deluge.

By the end of the 1890s, doubleheaders had become an established part of the game. People expected them to be offered and baseball didn’t have enough of a guaranteed fanbase to pull back on them. A doubleheader downturn in 1891 proved to be temporary. In 1892, even though the NL had the field to itself with the fall of the AA, over one-fifth of all games occurred in doubleheaders. By the end of the 1890s, the NL put a quarter of its games in twinbills, as the chart below shows:

Year	NL
1892	20.85%
1893	14.52%
1894	16.27%
1895	16.02%
1896	16.54%
1897	20.47%
1898	24.54%
1899	25.14%

A key theme emerged in the 1890s: The worst teams played the most frequently in doubleheaders. This was an especially important development, because it remained true for decades.

That makes sense if you think about it. Poor teams need an added inducement to convince the fans to come out and see them. Perhaps more importantly, when they traveled on the road their opponents needed an extra bit of persuasion to convince rooters to see what promised to be some lackluster on-field performances.

For example, in 1890, the Cleveland Spiders set a record by appearing in 17 doubleheaders while slogging through a miserable 44-88 season (good for seventh place in the eight-team NL). That season’s last place Pirates also topped the existing record by appearing in 15 doubleheaders and a rare tripleheader.

In the 1890s, long-standing sad sacks Louisville and Washington both twice led the league in doubleheaders played, and were routinely near the league lead in other years. Rather fittingly, the decade ended with the worst team of all-time, the 1899 Cleveland Spiders (20-134), setting a new record for doubleheaders, with 60 of their 154 games came in twinbills. From 1892 to 1899, when the NL had 12 teams, the Cardinals, Senators, Colonels, and Spiders played in the most doubleheaders. One of these teams came in last seven out of eight times in that period.

While the worst teams played in the most doubleheaders, the converse was not necessarily true. The best teams varied in how many doubleheaders they played. In fact, in 1898, the powerhouse Baltimore Orioles set a record by appearing in 26 different doubleheaders. Then again, the year before the Orioles played in the fewest doubleheaders in the league.

The Cubs were the most doubleheader-averse team in the league, something that would continue to be true of them all the way until World War II. They appeared in 78 doubleheaders from 1892-99, easily the fewest in baseball. Philadelphia appeared in the second fewest, with 97.

Why the Cubs? Though they were a middle-of-the-pack team in the 1890s, they had the advantage of being one of the most established squads. They were one of two teams to have survived since 1876, and already possessed a half-dozen pennants. They had the game’s biggest name in Cap Anson. Simply put, they were a draw. They finished in the top four in attendance every year except one in this period, despite never factoring in the pennant races. They played 38 home doubleheaders and 40 on the road, indicating that they could draw well around the league. Chicago’s railway hub status may have helped (it was easier to get teams in and out), but all MLB cities had good rail lines, and Chicago was the western edge of it all.

Simply put, by the end of the 19th century, doubleheaders had established themselves as a regular feature in baseball. They were held across the year (at least from May-onward) with a special emphasis in the later months, and the worst teams were the ones most likely to play in them.

By and large, these trends would remain in place for the next several decades while doubleheaders maintained their central status in major league history. They would continue to develop and evolve while retaining their important place in the game, but that’s a story for next week.

References & Resources
This column was achieved by relentlessly raiding Retrosheet.


11 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
gdc
13 years ago

Possibly an unintended benefit of scheduling a DH vs a bottom dweller if you didn’t have a big fanbase is that if said bottom dweller happened to have one good ironman pitcher that you didn’t have to face him twice, and your probable advantage in depth would come into play (e.g. if you normally had 3 starters and had to use #4, their dropoff might be worse than yours).

Ace K.
13 years ago

Chris,

I have not read this one yet, but I can confidently say that when I check out a HardballTimes article it is one by you a lot of the time.  And always a very enjoyable and interesting read.  I like your style, research, and topics.

Keep up the good work.

Ace K. (fake Internet name)

Paul
13 years ago

Baltimore may have played more doubleheaders because they were what we would describe today as a small-market team.  They finished first 1894-1896 then second the next two years, but despite that they ended up being bought out by Brooklyn as part of a syndicate scheme.  What would be interesting is a breakdown of the DHs by home and away – if most of their DHs are at home, this would be consistent with the theory.

The 1899 Cleveland Spiders started playing their home games on the road starting in July, which likely played a role in the DH count.  Louisville did the same “play at home on the road” in a prior year, but I don’t have details.

Also, there was a weird schedule in 1899.  Someone thought it would be a good idea to have a DH where the home team played team 1 for the first game and team 2 for the second game.  I don’t think they ever tried that again.

Chris J.
13 years ago

Why thank you, kind sir!

Chris J.
13 years ago

Paul,

Interesting theory about Baltimore – have to disagree, though.  Baltimore wasn’t large market, but it wasn’t small.  It was smaller than NYC, Chicago, Brooklyn, Philly, St. Loius & Boston – putting it 7th out of 12 in the 1890 census.  That’s middle of the pack.  What’s more, it was barely behind Boston and St. Louis (within 10K of Boston and 20K of St. Louis).  They were a bit further behind in 1900, but they weren’t small market. 

Baltimroe had over 100,000 more than Cleveland all decade, and an even bigger lead on Pit & Cincy.  Washington and Louisville were the real small ones.

Michael Caragliano
13 years ago

I’m sorry I won’t be at this year SABR convention. This sounds like one of the presentations I would’ve liked to see. Ted Lyons was the first name that came to mind when I saw your topic: IIRC, all of his starts in 1942 came on Sundays.

Paul
13 years ago

Census numbers are interesting, but they do not tell the whole story.  In 1898 Baltimore finished in second place, 6 games back of Boston.  Assuming my numbers from Total Baseball are accurate, in 1898 Baltimore drew 123,416 fans which was 9th in the league.  That was just barely ahead of Brooklyn, a team that went 54-91.  They had worse attendance than St. Louis, which finished dead last at 39-111.  They sold less tickets than Louisville, a really small market. 

When the team was started back up in the American League, they finished 6th in attendance, barely ahead of Milwaukee and Cleveland, who were the two worst teams in the league and both smaller cities.  The 1902 season is unique in its circumstances, though I’ll note that they finished last in the AL, both in wins and fans, though did do better than the NL Boston and Philly franchises.  (The NL squads probably lost ticket sales to their AL counterparts that were actually competitive.)

As John McGraw would later say, “Baltimore was never a town to support a loser.”  It may have not been a small market in population, but there is a significant difference between total population and people actually attending games.  Otherwise, we’d still have a team in Montreal.

Oh, and great article.  Good stuff.

Daniel Evensen
13 years ago

Very nicely written, Chris. I wonder what a bit more research could yield, especially in the archives of old newspapers. I haven’t looked much at the TSN scans of the 1890s, though I imagine they’re far from readable (considering what they look like in the 1900s). Perhaps something from the Chicago Tribune or Boston Globe, or (albeit less likely) New York Times from that era would shed a bit of light on the development of the doubleheader (yes, I do have access to ProQuest).

I’m looking forward to this series. I’ve wondered about the history of doubleheaders for quite some time now, and there honestly isn’t much literature on the subject. One thing I’ve always wondered is who in the world thought up that tripleheader in 1920.

Daniel Evensen
13 years ago

By the way, I won’t be at the SABR convention, either. I wish I were there. Perhaps someday SABR will be more than a hard-to-obtain Paper of Record subscription to me. wink

Chris J.
13 years ago

Paul,

Good point on attendance, but that’s less a matter of being small market than not being as much of a baseball town.

Dan,

There were some other triple-headers.  I know the Pirates played in one around 1890 or so. 

If you can’t be at SABR, they put the presentations online now.  (Not right away, but eventually they go up).  Here’s mine from last year:

http://convention.sabr.org/archive/sabr39/presentations/72-the-baseball-philosophy-of-charles-comiskey

Mike Bailey
13 years ago

I don’t think the numbers of doubleheaders can be accurately analyzed without taking into account the number of makeup games. To take one example,1894 (according to Retrosheet’s original schedule files), all teams scheduled doubleheaders only on Memorial Day, July 4th, and Labor Day. Thus all other doubleheaders were makeups.