In a recent post (Exactly who are the DHs in the AL?) coolhandbane suggested an article on shortening the MLB schedule. And Laura Peebles mentioned double-headers and problems that MLB has in scheduling make-up games for rain-outs given that many teams don’t have common off-days.
Constructing schedules for any sports league is incredibly complicated. There are analytical techniques and lots of heavy-duty software programs that can do this. They are broadly described as Linear Programming or more specifically Mixed Integer Programming. If you are so inclined, check out A Mixed-Integer Programming Approach to Generating Sports Schedules. It was presumably written by a student/professor at Duke University. Note that it gets pretty advanced/technical pretty quickly.
Lets start with some simple arithmetic. In 2015, the regular season was 26 weeks long. Teams were scheduled to play 162 games. So each team averaged 162/26 = 6.23 games each week. Over the years many folks have proposed a number of changes to shorten the season:
- Going back to the 154 game schedule. As many may know, the 154 game schedule changed to 162 games after the 1960 season. One implication of that is that there are records based on 154 game seasons. One of the arguments against shortening the schedule by just 4 games (for example) is the impact on records. Given the current average of 6.23 games a week, a 154 game schedule shortens the season by a little more than a week. But we all know that this is a non-starter. The owners are not going to agree to 8 less games (4 less home games) a year.
- Mike Scioscia was quoted in Should baseball bring back doubleheaders? as advocating more double headers in order to shorten the season. The article also described how often double headers had been played in the past:
- Ernie Banks played in both games of 318 double headers in his career.
- Quoting from the article: The White Sox played a record 44 doubleheaders in 1943, or 58 percent of their season schedule.
Keeping the arithmetic simple, suppose the MLB schedule could be constructed as follows:
- Every Sunday (or every other Sunday) is a day-night double-header. We all know the owners aren’t going to give us two games for the price of one!
- Every Monday is an off-day. This has two benefits:
- A day off after every double header
- A scheduled off-day for every team that could be used for the occasional make-up games for rain-outs.
- Since Sunday double headers mean that a Friday-Sunday series is a 4 game series, perhaps we have more 2 game series (e.g., some weeks would also have Thursday or Friday as an off-day).
- Maybe get-away games on Thursday are afternoon games. That would avoid an NL/AL East team flying all night to play the next day against an NL/AL West team (and vice versa).
So how long would a 162 game season be if MLB just did items 1 and 2 above?
- If there is a double header every Sunday, the teams would play 7 games each week and so the season would be 162/7 = 23.1 weeks long. In other words the season could start on April 8 and end on/about September 17. So the World Series would be done by the middle of October.
- As a result MLB could expand the Divisional Series from 5 to 7 games.
- With a double header every other Sunday, they teams would play 13 games every two weeks. So the season would be 2*162/13= 25.9 weeks long. Basically one week shorter.
I have to admit that I have no idea if such scheduling is possible given all the constraints in the scheduling. Seems to me that this is something that MLB should investigate.