04 April 2016

Expansion To Japan: How The Schedule Would Look

Previously, I wrote an article discussing that MLB should consider expansion into Japan. I proposed that MLB should relocate the Tampa Bay Rays and Oakland Athletics to Japan while adding teams in Tokyo, Osaka, Nagoya and Yokohama for a total of 32 teams. Here’s some more detail about this might work.

I propose a structure of 8 divisions consisting of 4 teams each. Such a plan would eliminate the AL and NL.
The logistics only work if all four Japanese teams are in a single division and having a Japanese division in a specific league would put that league at a significant disadvantage. There were two ways that I found that made sense to organize the divisions.

I don't think it matters if MLB chooses these exact four Japanese cities. MLB certainly needs a presence in Tokyo. If they can't build a team there, then this project has limited value. Aside from that, it would be nice if MLB could put teams in Osaka or Nagoya but other solutions could probably be found. Yokohama isn't a large city and it may make more sense to put a team in Kobe or Nishinomiya instead.

The first structure, and the one that I developed a schedule for consists of the following:

Division #1: Braves, Reds, Nationals and Marlins
Division #2: Mets, Indians, Tigers and Pirates
Division #3: Red Sox, Orioles, Phillies, Yankees
Division #4: San Diego, Dodgers, Diamondbacks, Giants
Division #5: Cubs, Brewers, Cardinals, Blue Jays
Division #6: White Sox, Twins, Rangers, Astros
Division #7: Mariners, Angels, Rockies, Royals
Division #8: Osaka, Nagoya, Yokohama, Tokyo

The idea behind this grouping was to attempt to minimize intra-division travel while maintaining some familiar rivalries. One major challenge is that there are only seven West Coast teams if the Athletics are relocated and this requiring a Midwestern team to play on the West Coast. Kansas City is the best option to be placed in such a division and therefore gets put into a tough situation. Each new division has at least two teams that were in a previous division together. I presume that the New York, Chicago and Los Angeles teams would need to play in different divisions as is currently the case.

The second way, and probably the better one, consists of the following:

Division #1: Braves, Rangers, Astros and Marlins
Division #2: Reds, Cubs, Cardinals, Brewers
Division #3: Red Sox, Orioles, Blue Jays, Yankees
Division #4: San Diego, Dodgers, Diamondbacks, Giants
Division #5: Mets, Nationals, Phillies, Pirates
Division #6: White Sox, Twins, Tigers, Indians
Division #7: Mariners, Angels, Rockies, Royals
Division #8: Osaka, Nagoya, Yokohama, Tokyo

This set of divisions has less internal change than the first set and also requires less intra-division travel.  Teams like the Mets, Twins, Reds, White Sox and Nationals will be forced to travel less under such a structure while only the Braves and Indians have to travel more often. I probably would have a more optimal situation had I developed a schedule for this second set of teams rather than the first.

This schedule is a proof of concept and shouldn’t be considered even a rough draft. For starters, I didn’t do an adequate job ensuring that each team has the same amount of home weekend games. I changed a few decision rules about schedule structure midway through developing this attempt, and therefore would need to tweak the schedule to ensure better consistency. This schedule also does a poor job ensuring that teams don’t need to travel over the weekend, but rather can play only one opponent in one location. Better optimization tools and better thought-out schedule criteria would allow me to address these problems, but at the cost of a significant amount of my time.

The basic schedule structure is like this. The season is 182 days long, or about the length of the standard MLB season, and there are 152 games played by each team. Those 152 games are split into 66 in division games (22 games against each division opponent) and 86 out of division games. For the out-of-division games, teams play between three to five games against opponents in six of the seven divisions. They face opponents in three of the divisions at home and in the other three on the road.

For simplicity’s sake, I have divisions play all of their games against each other over the same period. For example, if one team in Division 1 is playing a team in Division 4, then all teams in Division 1 and 4 are playing each other. Furthermore, once teams in Divisions 1 and 4, start playing a team from a different division, teams from those two divisions won’t play against each other for the rest of the season.

This format allows for not only an all-star break, but an extended period of rest in June, July, August and September for most clubs. I do think that having only 152 games results in a large amount of off days and that 154 or 156 games makes more sense. If I was revising my first draft, I would look into a 156 game schedule, in which 78 games are played against division foes (6 three game series and 2 four game series against each team in the division), 72 games played against out-of-division foes (teams play a three game home series and a three game away series against opponents in three divisions) and 6 games played against special opponents (for example the Orioles would play the Nationals for three games at home and for three games away). The teams in Japan would play against other teams in Japan and not in the States to reduce travel distance.

This format presumes that teams need an off day when traveling from the East to West Coast and vice versa, and need two off days when traveling from the United States to Japan. Schedules were developed so that teams on the East Coast don’t need to travel to Japan AND travel to the West Coast.

Teams are allowed to play at most 17 games without an off day and doubleheaders are allowed only in cases where games are cancelled. This schedule presumes a four day all-star break and also a few breaks during the season. The schedule can be found for each team in a spreadsheet format by clicking here. Below is how the Orioles schedule would look in a schedule format.

According to Baseball Savant, MLB teams traveled 810,000 miles in 2015. According to my calculations, they traveled 827,000 miles. There are some slight differences between our calculations, but nothing major. In this new schedule, the 28 teams that exist in the old and new format would be projected to travel just 632,000 miles. When taking the fact that teams are playing fewer games in this new schedule and thus travel less into account, this results in a reduction of 75,000 miles or roughly 2,900 per team. The chart found here shows how things change.

The four Japanese teams are projected to travel roughly 46,600 miles. This is the highest allowed in my sample by far, but it’s only 3,500 miles more than the Mariners traveled in 2015. It isn’t ideal, but it certainly is reasonable. With better optimization software, I could likely reduce the distance traveled from 46,000 miles.

There are challenges that may prevent MLB from expanding into Japan. It will be hard to come to an acceptable deal with the NPB. It may also hard to convince Japanese cities to provide necessary support for MLB teams to ensure that their financial success. However, despite the many challenges, it is clear that an optimized schedule will prevent distance from being one of them.


Anonymous said...

Great article and very interesting perspective. I like the idea of my first future visit to Tokyo including an O's game.

Matt Perez said...


Anonymous said...


Great to see the follow up to the first article on this. I do think there is some incentive for this to happen in Japan. First, they have the potential to keep their best "home grown" players in Japan. They follow baseball like crazy and it is by far the coolest environment that I have ever experienced.

My question is, with the expansion and the TV deal that would no doubt have to go with this (which I have no clue how that would work with Japan), would the owners pockets be filled enough to give away the home games that 152/156 games takes away.

Second question is, how would the playoffs work?

Love the idea!

Matt Perez said...

If MLB can't make it extremely profitable, then they'll have no interest in expanding to Japan. I've suggested that the loss of games would cost MLB around $250-$500M in revenue. I expect the lower end of that spectrum.

It's hard to tell how all of this would work with Japan. I think you'd need the four Japanese teams to earn at least $1.5 billion in revenue per year in order to make this worthwhile. I think MLB should be able to get there. Realistically, that means the four teams will need to average 12 million fans total in a season and likely bring in $1 billion in media revenues. The Rays and Athletics require significant revenue sharing cash.

The playoffs would be a challenge. Winners of each division get a playoff spot and there are no wildcards. I don't care how the seeding is determined. It can be done via regular season record or by a lottery.

The first round would be five games. Two games would be at one stadium (team with best record). There would be a two game break. Then the next three games would be at the second stadium.

The second and third round would have two games at the first stadium (team with better record), a two day break, three games at the second stadium (team with worst record), a two day break and then two games back at the first stadium.

bordomsdeadly said...

Just a heads up. Astros fans hate being in the same division as the Rangers. Back before crane owned the team most fans rooted for both teams. We consider it mlb's manufactured rivalry. If you do this again could you please format the old 6 team national league central. It would feel much better as an Astros fan to see this, and I imagine rangers fans probably feel the same way.