Day for night

Baseball has seen its share of revolutions over the generations. From professional teams and fielders’ gloves in the 19th century to retractable domes and (occasional) video review in our century, the conditions under which baseball is played have been in considerable flux for a game that cherishes its traditions.

One of the greatest changes of all has been the introduction of night baseball. Conceived to give working folks greater opportunity to attend games and thus raise team revenue, the innovation succeeded wildly—and it also wrought changes to the game on the field. This article will examine some aspects of how the lights have changed game play, while taking some glimpses at the scenery on the road night ball has traveled.

And there was light

Many baseball fans, and likely most of our readers, know that the first major-league night game was played at Crosley Field in Cincinnati. What they may not know is that the lights could have come on earlier than May 24, 1935, against the Phillies. Much earlier. Over a quarter-century earlier.

Crosley Field’s predecessor in Cincinnati, occupying the same site during the first decade of the 20th century, was perhaps the most wonderfully named ballpark ever: The Palace of the Fans. (“Fan” was still a new and slangy term in those days, so this name was quite unconventional—but it’s already aged better than, say, O.co Coliseum or Enron Field. More stolid observers did insist on calling it League Park, though.) The park boasted high-end touches like Greco-Roman columns, along with what can fairly be called baseball’s first luxury boxes. They were broad ground-level booths, served by waiters, and with ample room beneath the grandstand to park your carriage (or your newfangled automobile, if you were really loaded).

Late in August of 1908, Reds owner and president Garry Herrmann announced his plans to play Reds games at night that season. He had already helped to incorporate the Night Baseball Development Company, and the plan was to erect five towers around the Palace of the Fans, mounted with searchlights. His plans were a little over-optimistic, as only three of the towers were in place by the end of September, with completion then expected in two weeks.

The towers did finally go up, and night baseball was played at the Palace—not by National League teams, but by a pair of Elks lodges. The game on June 18, 1909 drew between 3,000 and 5,000 spectators, including several Reds lingering from their afternoon game. This was not the first night baseball game by a long shot, a few amateur games having been played under lights even in the 19th century. Still, this successful experiment seemed to presage imminent acceptance at the highest level.

It didn’t happen. Maybe the quality of lighting was good enough for Elks, but not for Cubs or Cardinals. Perhaps the rapid deterioration of the Palace of the Fans deterred Herrmann, and he decided to concentrate his capital on building its successor, Redland Field (which would become Crosley Field in 1934). For whatever reasons, baseball’s leap into the future stalled for a generation in the major leagues.

But the lights finally came. President Franklin Roosevelt threw a switch in the Oval Office, and hundreds of bulbs began pouring nearly a megawatt of light upon Crosley Field. The Reds reaped early dividends from the great experiment, winning a 2-1 decision from the Phillies that evening.

Night baseball would gain an early reputation for suppressing offense. The numbers from the first three years of night games at Crosley Field support this, eventually.

I compared the 210 day games played there from 1935-1937 to the 21 night games they hosted. (Games at other fields provided different run environments, along with home teams with different offensive abilities than the Reds, so I exclude them.) Runs per game were almost equal for the first two seasons: 8.57 apiece in 1935, then 9.29 at night versus 9.32 in daylight during 1936. The following year produced some separation, day play scoring 7.75 runs a game while night contests scored an even seven runs a game. The underlying batting numbers follow a similar pattern.

Year    Crosley day games    Crosley night games
1935    .267/.319/.341       .259/.313/.350
1936    .277/.331/.349       .285/.327/.355
1937    .259/.318/.321       .223/.298/.265

More important in entrenching the reputation were the rawer numbers. Crosley Field played as a pitcher’s park, and the Reds were a subpar offensive team, for those three years. Scoring in Cincinnati was thus already depressed, and it did not take much added run suppression to create the appearance of an offensive drought in night games.

This chart of OPS in day versus night games (for all the major leagues) shows how dramatic that early spike was. Do note that the early numbers come from a very small proportion of games being played at night until the mid-1940s.

I will draw no broad conclusions from this chart yet, except to note how early impressions can linger (and sometimes even be accurate as well). We will return to it, though.

Entering the mainstream

Cincinnati’s monopoly on major-league night play ended on June 15, 1938. The Brooklyn Dodgers threw the switch that night at Ebbets Field—and they could not have suspected how dramatic the debut would be. Their opponents, appropriately, were the Cincinnati Reds, whose starting pitcher had thrown a no-hitter his last time out.

But surely Johnny Vander Meer wasn’t going to do that again.

When he did actually do that again, he did more than earn himself an unbeatable spot in the record books. He did even more than add to the anecdotal record of night games being low-scoring affairs. He made night baseball, at least for one evening in Flatbush, the most exciting kind of baseball there was. Any chance that the momentum of night games could be checked succumbed as easily as the Brooklyn batters.

In both 1939 and 1940, four teams at three venues began playing under the lights. (Shibe Park [1939] housed both the A’s and Phillies; Sportsman’s Park [1940] hosted the Browns and Cardinals.) By the 1941 season, only five teams didn’t have night ball, and it was due to become at most four in 1942, as Wrigley Field had its lights ready for installation. A slight difficulty at Pearl Harbor intervened, and owner Philip Wrigley famously donated those lights to the war effort the day America declared war on Japan.

The march toward illumination resumed after the war, and it is interesting to note the last four ballparks to make do without lights. In order of eventual adoption, they are Yankee Stadium, Fenway Park, Tiger (then Briggs) Stadium, and finally Wrigley Field. These four venues would, in the closing decades of the 20th century, be hailed as the last great holdovers from a golden age of ballpark design, a rebuke to multi-use concrete ashtrays, sterile domes, and artificial turf. It is fitting, perhaps, that these bastions of old-style baseball had been the last to surrender the tradition of always playing in daytime.

The New Normal

Night baseball was becoming familiar, even common. But when did it become normal? Dave Studeman gave one answer at this site years back, when he identified 1961 as the first year that night games outnumbered day games. The chart below shows that progression, along with how (as Dave also noted) we’ve been at a plateau the last 30 years.

But I had another definition in mind. When did players become as comfortable playing under lights as in the sunshine?

The answer does not depend on an exact equality of statistics in day and night play, because the game is played under different circumstances, making for different performance environments. The night environment itself was not always stable: lighting equipment surely improved over the decades since the towers first went up at Crosley Field. We may not expect an exact answer, but we can still make a few good passes at approximation.

The most telling statistic measuring quality of play would seem to be errors. That first night game in 1935 was played error-free, but one can argue that in dimmer light, you’ll see the ball less well, judge its trajectory more poorly, and misplay it more often. Error splits for day and night games are not available until 1948, but those that are available reveal a surprising pattern.

(Note: the following charts show the day/night ratio. A number over 1 means the result is happening more frequently in day games; under 1 means it is more common in night games.)

The chart shows the ratio of day error rates to night error rates (said rates measured as errors per ball in play). Going year by year, or especially with moving averages, we see a generally higher error rate for day games.

This is not the common-sense result we’d expect. Perhaps there could be a subjective factor: scorekeepers might be less forgiving of misplays made in the light of day, especially in the early seasons of night play. Maybe the contrast of a lighted white ball on a darker background makes it easier to judge after all (though this would seem to be contradicted by the walk numbers, below). In any case, there is no evidence here that ballplayers had any difficulties adjusting to night play as early on as 1950. If anything, it is counter-intuitively the opposite.

Let’s try again, using hit-by-pitches this time. Batters could have a harder time at night picking up the ball and thus getting out of its way. (The effect of night play on pitchers should be far less. Moving objects are one thing, but home plate is not going anywhere.) The moving average plot is especially useful here in the early years, when small sample size produced high volatility in the ratios. (Rates for HBP, and for the stats in following charts, are per plate appearance.)

And our expectations are confounded again. In the developing years of night ball, players got plunked much more often in daytime. The ratio would eventually reverse during the 1960s, but flip back afterwards, so that for the last three decades, batters have been mildly safer at night.

There is no subjective judgment influencing these numbers, unless you count the umpire’s ability to tell when a batter has been nicked. Perhaps that explains the difference; perhaps pitchers are a bit more careful at night, or batters a little quicker to dodge an inside pitch. However we slice it, there is no evidence of a long-term fall-off of quality of play at night in this category.

Let’s stay with the pitcher-batter relationship, and examine walks and strikeouts. As observed before, the pitcher’s skill at hitting the stationary target of the strike zone should scarcely be affected. The batter’s ability to judge the ball in flight, and either make contact with it or let it pass, seems more vulnerable. If this is the case, we should expect fewer walks and more strikeouts at night.

Strikeout data dates only from 1938.

Well, that’s a little more like it. We have fairly conclusive results this time, though one of them ends up a shocker again. With infrequent exceptions, batters have been able to take walks better by day, and after an early rate drop, the trend line has been wandering around the same general level since 1950. The numbers for strikeouts start off much the same, showing the batters struggling at night. However, the trend line did not flatten out (if it ever has flattened) until batters had improved their night performance so much that they have consistently struck out less under the lights for over 20 years.

We now look at home runs, and here we have more than the difference in lighting to consider. Night games are played in a cooler environment than day games, and this materially affects the behavior of batted balls. The ball travels farther in hot weather, as well as when the ball itself is warmer. Thus, day games should feature more home runs, assuming the performance of the players is equal.

How many more home runs? It depends on the temperature difference between day and night games, and this is tricky to estimate. The hottest part of the day is several hours after noon, later during summer months and earlier in winter. Day games in earlier times tended to begin a 3 p.m. local time, but today are likelier to start at 1 p.m., as opposed to a general 7 p.m. start for night games. There is also Daylight Saving Time, which became law in 1967 (as well as during the World War II years).

After wading through all these variables, and the actual meteorological records, I estimated that night games average five degrees cooler than day games. As for the balls in play, they would be in a climate-controlled environment before the game, but would have some time to approach ambient temperature before being pitched and batted. I estimated that baseballs in night games would be two and a half degrees cooler, half the margin for the overall temperature.

In The Physics of Baseball, Robert Adair estimated that every 10 degrees additional air temperature would add four feet to the flight of a standard 400-foot home run ball, and thus raise home run probability by seven percent. The same adjustment occurs for the warmth of the ball in play. My estimates thus total an added oomph of three feet to a standard home run (two from air temperature, one from ball temperature), and a higher home run rate that I round off to five percent.

This means we expect about five percent more homers to be swatted in day games, regardless of any day-night fluctuations in player skill. Therefore you should consider 1.05 the true baseline in the following chart:

This follows a pattern quite similar to the strikeout numbers. Batters struggle early on, make (not always steady) improvement, and roughly by 1980 begin doing better under lights than under the sun. Speaking of 1980, you’ll note that is the first year that batters homered more often in night games than day games. Scroll back to the OPS chart, and you will see that was also the first year that teams scored more in night games than day games. The two are naturally connected: homers are a great way to produce runs.

What the numbers mean

Adjustment to night baseball came in two distinct phases. The first, ending around 1950, saw a general acclimation to night play that favored the defense. Error and walk rates were level or better at night, and had ceased trending in any particular direction. The second phase saw the batter’s side of the game improve. He struck out less at night, homered more, and pushed his overall production even with day games, possibly better. Offense has now fully overcome the environmental disadvantage inherent to night baseball, and is still gaining, though at a reduced and likely flattening rate.

Swinging the bat, and to a degree judging the strike zone, were the elements of baseball slowest to adjust to the lights, which is perhaps not surprising. What is surprising is that night hitting has now not just matched, but exceeded the daytime baseline. Practice makes perfect, so maybe the greater number of games played at night lets batters feel more comfortable then.

Of course, they are also taking plenty of swings during batting practice which, coming hours before game time, is far more often conducted under daylight conditions. One might think that would balance the scales, but it doesn’t. One leans toward the conclusion that BP is not an adequate substitute for batting in a real game—and I suspect most players would agree with that idea.

The Old Guard, in retreat

Of what we can call the Original Sixteen teams, not one of them played their first night game on Opening Day. This makes sense. Night games were a strategy for boosting attendance, and Opening Day was already the most reliable day for high attendance. (The Fourth of July might be better, but half the time you’d be playing at someone else’s field on that day.) This old wisdom persisted during the first wave of expansion. The Angels, new Senators, Mets, and Colt .45s all waited past Opening Day to play their first night games.

The San Diego Padres broke that tradition. Their first game ever was a 2-1 victory over Houston on Tuesday night, April 8, 1969. (Totally unrelated bonus trivia: all four expansion teams that year, the Padres, Expos, Royals, and Seattle Pilots, won their first games.) But the Padres were probably wise to double up on the attractions of Opening Day. They drew just under half-capacity for their opener (23,370), and the following two games, both at night, drew fewer than 5,000 fans.

The old ways hung on for a while. The Toronto Blue Jays’ home premiere was a daylight affair, though this was probably to minimize the chances of arctic weather ruining the proceedings; they didn’t have a home night game scheduled until May. (It didn’t work: they played that day on a snow-covered field with wind-chills at 10 below zero.) The Seattle Mariners debuted with a night game, but under the Kingdome, what was the difference? The Marlins and Rockies enjoyed sun for their first home games in 1993, but five years later the D-backs and be-domed Devil Rays went the opposite way. Indeed, of the eleven games played that March 31 Opening Day, the only two night games involved the newest expansion clubs.

The first World Series night game was in 1971, Game Five between the Orioles and the Pirates. (Our Chris Jaffe gave a thorough rundown here.) So strong was the need to accommodate demand for evening World Series play, largely from the television networks airing them, that the last scheduled World Series day game came a mere 13 years later, as Sparky Anderson’s Tigers closed out the Padres in Game Five of the 1984 Series. It was a late afternoon start, however, and Kirk Gibson’s door-slamming homer off Goose Gossage came under the lights.

There was one later World Series day game, but with two major caveats. First, it was Game Six of the 1987 Twins-Cardinals Series, played in the Metrodome, not under sun and sky. Second, it was rescheduled for the afternoon only in hopes of allowing timely conversion of the Metrodome floor to football use, for a Broncos-Vikings game scheduled for the next day. A Twins win mooted that plan by extending the Series and forcing the postponement of the Vikings game to Monday night—a boon for ABC, which was carrying the World Series as well as Monday Night Football.

MLB might yield on night World Series games to the NFL, but not to the Chicago Cubs. The last holdout against night baseball, faced with the ignominy of having prospective World Series home games played in St. Louis or—horrors!—Comiskey Park, finally gave way and installed Mr. Wrigley’s lights in 1988, a mere 47 years behind schedule. The Chicago City Council voted to permit a maximum of 18 night games a year, not counting playoffs, in consideration of neighbors of Wrigley worried about the effect on the neighborhood of thousands of milling fans 60 nights a year.

That consideration has lessened. In early 2004, the council voted unanimously to expand the night game schedule at Wrigley Field, setting a ceiling of 30 night games by the year 2006, to last until 2015. There is pressure to raise that number even higher, and it may not be long before this picturesque holdout succumbs, and is playing more night games than day games, like every other team in the league.

(This article won’t speak to the question of whether the heavy daytime schedule at Wrigley Field has harmed the Cubs’ performance. For those interested, this paper delivered at the 2011 SABR Conference in Long Beach by noted researcher Mark Pankin gives his answer.)

This change would scarcely affect attendance, as Wrigley Field stands at or near capacity almost every game these days. It would affect revenues, though, as television ad rates are higher in the evenings than the afternoons. From the first, money has been a prime motivator in the advance of night baseball, so this is nothing new. One could almost say it’s traditional.

Would baseball lose something if day games at Wrigley became no more common than at any other park? Yes, if only some of that diversity in venues that has made baseball unique in a field of otherwise highly standardized professional sports. But it would not end up being a crushing blow to the game we love. As the players have already shown, baseball can adapt.

References & Resources
Retrosheet
Baseball-Reference
The Sporting News, online archives
Philip J. Lowry, Green Cathedrals, 2006 edition
Craig Robinson, Flip Flop Fly Ball
G. H. Fleming, The Unforgettable Season
Cait Murphy, Crazy ’08
Robert K. Adair, PhD, The Physics of Baseball, third edition
Crosley-Field.com
Digitalballparks.com