About this report
With the most recent MLB owners' proposal that showed a salary cap, it seems very likely that Major League Baseball will lockout its players on December 1, 2026. The last time that owners proposed a salary cap was in the 1994. That resulted in the World Series being cancelled for the first time in over 85 years and when baseball came back, fans didn't come back right away.
The standard story is simple and would be familiar to any baseball fans. The 1994 - 1995 strike hurt baseball attendance but the 1998 home run record chase between Sammy Sosa and Mark McGwire brought back the fans.
That may be true in spirit and it was really exciting watching those two greats chase down Roger Maris, but the numbers deserve a little more care. Raw attendance can be misleading because 1994 was not a full season. If the season stops early, total attendance falls even if fans were still showing up at a strong rate when games were actually played.
So this report looks at attendance in two main ways:
- Total attendance, which tells us how many fans attended across the season.
- Attendance per opening, which tells us how many fans attended per paid gate or attendance event.
The second number is especially important because doubleheaders can put two games into one attendance event. Lahman calls those attendance events openings. That is a really confusing term because to me it speaks of opening day. But when you start thinking about a doubleheader where one ticket gets fans admission to both games it makes sense. The doubleheader is two games, but there was only one attendance event.
This all gets important when we look at the rebound in attendance and also consider some other things that happened in baseball. Attendance did rebound in 1998, but Major League Baseball added two new teams. We need a solid number if we want to account for expansion and figure out what really happened on the revenue side.
Why openings matter
In this report, total openings means the sum of all team-level openings from Lahman HomeGames.csv in a season.
An opening is a separate paid attendance event or gate. If a team plays a doubleheader, there can be two games but only one opening. That means total openings can be lower than total home games.
This matters because attendance per opening answers a different question than attendance per game.
Attendance per game asks:
How many fans were there if we spread attendance across all home games?
Attendance per opening asks:
How many fans showed up per paid attendance event?
For this report, attendance per opening is usually the cleaner number. But that's based on what I do for a living. I am a software developer who consults and sells products. I get happy when I actually sell something, everything else is just theoretical. If you would like to run this report with different numbers all the data is available on this page or you can open up developer tools and copy the file called data.json. You don't have to credit me for it because it's just a hack job on Lahman's data.
The strike-window problem
The 1994 season is pretty screwed up. Raw attendance fell hard because the season stopped in August. But attendance per opening did not collapse in 1994.
The first full post-strike season is where you see attendance per opening get smuckered.
In 1995, baseball came back with a shorter 144-game schedule. Total attendance was still far below 1993, but the more important number is attendance per opening. That is where the fan backlash really shows up.
The rough shape is:
- 1993: the pre-strike reference season.
- 1994: the strike-shortened season.
- 1995: the first full post-strike season.
- 1998: the expansion and home run chase reference season.
That 1998 year is tricky too. MLB added two teams, so raw total attendance had a built-in expansion boost. The raw number matters, but it should not be treated as a clean one-to-one comparison with 1993.
How to read the summary cards
The summary cards focus on a few key comparisons:
- 1995 attendance per opening vs 1993: the cleanest measure of the post-strike attendance wound.
- 1995 total attendance vs 1993: the raw attendance comparison.
- 1994 attendance per opening vs 1993: a reminder that the strike-shortened season is schedule-distorted.
- 1998 total attendance vs 1993: a useful recovery number, but one affected by expansion.
- Team-seasons analyzed: the size of the underlying team-season dataset.
- Highest total attendance season: the strongest raw attendance season in the report window.
The chart uses attendance per opening because that is the metric most directly connected to the question of whether fans were showing up and paying when gates were open.
How to read the season table
The season table has one row per year from 1985 through 2005.
The main fields are:
- Season: the MLB season.
- Team count: the number of MLB teams represented in that year.
- Total attendance: summed home attendance from Lahman HomeGames.csv.
- Attendance per opening: total attendance divided by total openings.
- Attendance per game: total attendance divided by total home games.
- Total openings: summed paid attendance events from HomeGames.csv.
- Total home games: summed home games from HomeGames.csv.
- Multi-park teams: team-seasons where a club used more than one park.
- Note: context such as strike year, post-strike year, expansion year, or multi-park weirdness.
The table is sortable so you can quickly check which seasons had the highest attendance per opening, which years had expansion effects, and which years were most affected by schedule changes.
How to read the team table
The team table has one row per team-season.
The main fields are:
- Season: the MLB season.
- Team: the team name from Lahman Teams.csv.
- League: American League or National League.
- Record: regular season win-loss record.
- Home attendance: aggregated attendance from Lahman HomeGames.csv.
- Per opening: team home attendance divided by team openings.
- Openings: team-level paid attendance events from HomeGames.csv.
- Home games: team-level home games from HomeGames.csv.
- Park count: how many parks were used by that team that season.
- Park: the primary park and any additional park names.
- Note: quick context such as top-five attendance, bottom-five attendance, strike year, expansion year, or multi-park season.
This table is searchable and filterable. You can look for a team, a park, a specific year, top-five attendance seasons, bottom-five attendance seasons, or multi-park team-seasons.
Methodology
This report uses three Lahman CSV files:
- HomeGames.csv for attendance, openings, home games, and park usage.
- Teams.csv for team names, leagues, divisions, wins, and losses.
- Parks.csv for park names and locations.
The most important data decision is this:
HomeGames.csv is the source of truth for attendance.
Teams.csv also has attendance values, but those values are not used as the canonical attendance field in this report. HomeGames.csv gives us the attendance data along with openings and park usage, which makes it the better source for this particular question.
The report window is 1985 through 2005. That gives a pre-strike baseline, the strike-shortened season, the first post-strike seasons, the 1998 expansion/home run chase year, and several later seasons to provide context on how the recovery went.
This report does not attempt to measure ticket revenue, ticket prices, television ratings, local market conditions, ownership decisions, labor sentiment, or even attempt to answer why. It's purely about attendance data.
Accessibility notes
This report is built around text and tables first.
The chart is included to make the trend easier to scan, but the season table contains the same trend data in text form. The tables use captions, column headers, normal buttons for sorting, and normal form controls for filtering. The result counts update after filtering.
This page was scanned for accessibility with Siteimp and tested with NVDA.
I am still learning accessibility, so I make mistakes. You'd maybe think that after 16 years I could have stopped learning by now but it doesn't work like that. If something does not work with your screen reader, keyboard, zoom settings, or browser setup, please contact me so I can fix it.
What did we learn?
The strike attendance story is more complicated than raw attendance alone.
Raw attendance fell sharply in 1994 because the season stopped. But attendance per opening was actually slightly higher than 1993. That does not mean the run up to the strike was good for baseball. It means 1994 is a distorted year to compare because the season was cut off because of the strike.
The 1995 season is where the damage becomes much clearer. Baseball returned, but attendance per opening dropped sharply from the 1993 reference point. By 1998, raw attendance had climbed back above 1993, but MLB had also expanded from 28 teams to 30 teams. So the return of raw attendance should be read with that expansion context in mind.
In other words: the fans did come back, but the wound was real, and you can see that wound most clearly if you consider attendance per opening.
Did baseball learn?
As a fan, this is the part that concerns me the most. I never played professionally and I am in absolutely no danger of ever being able to afford to buy a Major League Baseball team. So, I don't have a pony in the race and am really not qualified to speak about any of this. Except that I love this game and I love everyone else who loves it.
It took at least until 1998 for attendance to recover in an era when people didn't have as much to do as they do now. The sporting landscape was not as great, the media landscape was not as vast, video games were still several years from reaching their technical and artistic zenith, and social media was not even comparable to the dopamine fuelled services we give our attention to today.
What happens if we lose 2027 and this time the fans don't come back? It's not big clubs like the New York Yankees or Los Angeles Dodgers that will get in trouble. It's the small market clubs that this competitive balance stuff claims to be helping whose fans will pay the biggest cost.
One of the most special things about baseball to me is the opportunity to help a new fan see the game the way I see it. When I get them into the mix of lore, tradition, superstition and science of it, I watch their eyes perk up as they start to see how interesting of a hobby the sport is. And while I'm really trying not to be pessimistic, I am genuinely worried that if we lose 2027, I won't have that experience again for a very long time.
Can baseball survive without small market clubs? Most likely. Is that an MLB that I will love and feel passionate about? Most likely not.