The Arizona-Florida Plan Creates a Solvable Scheduling Pickle

Last week, two competing plans for an alternate-site baseball season were leaked. The first was the so-called Arizona Plan: Send all 30 teams to Arizona, rotate games between the available fields, and play an abbreviated major league season with no in-person audience. That plan has its logistical pitfalls, but one of the few things the plan doesn’t alter is the existing divisional structure of baseball. Aside from a shorter season and its attendant complications, baseball would mostly work the way it always has: the Red Sox, Rays, and Yankees would attempt to club each other into submission, the AL Central would be full of rebuilding teams, and so on.

The second plan, the so-called Arizona-Florida plan, would be something else entirely. Instead of recreating the exact structure of the league in one city, this plan would place each team at their spring training facility. Many of the logistical issues from before would still need to be answered. Assuming those can be handled, however, there’s still one major twist: instead of existing divisions, the teams would be grouped by geographic proximity — and, of course, given that the existing setup isn’t 15 AL teams in one location and 15 NL teams in the other, the leagues would be scrambled.

Per Bob Nightengale, the divisions would look like this:

That’s quite the scramble. Dan Szymborski is running the new divisions through ZiPS to get an idea of what it does to teams’ playoff odds, but I thought I’d consider the mechanics of playing with 15-team leagues, as well as highlight some interesting matchups, to give you some sizzle to go with your steak, as it were.

First things first: how would an odd-numbered league work? Baseball has gone to interleague play in part to avoid this question: in a game where most every team plays most every day, having odd-numbered leagues won’t work. That’s why expansion was always done with two teams joining a single league at once; before the Rockies and Marlins were added, the AL had 14 teams to the NL’s 12. When the Diamondbacks and Devil Rays joined the majors in 1998, the Brewers were shifted to the National League to keep the numbers even — 16 NL teams and 14 AL.

When, in 2013, interleague play became a daily occurrence instead of a specific chunk of calendar, the Astros shifted to the AL, setting up our current 15-team leagues. But of course, the daily interleague play plan won’t work anymore; no one’s flying back and forth across the country in this plan, which is specifically designed to avoid flying back and forth across the country.

What other options do we have? One choice would be to have some non-traditional double headers. A team could play a day game and then a night game against two different opponents, serving as a de facto 16th team for game balance. Of course, that schedule is about as unfair to that team as you would expect — playing two games in a day is usually a mutual disadvantage, and while it’s hard to estimate the exact cost, it’s not likely to be low.

How can that cost be offset? Most likely, it can’t be. An extra pitcher or always playing at home can’t offset the cost of playing twice a day. But if we give each team an equal number of split double headers, the cost can at least be shared among all teams. Nightengale’s article references a 108-game season. You could handle this issue by having 13 days where a team has an off night and 90 days where a team plays a split doubleheader, with each team playing six of them.

That plan is probably too radical. A team playing its second game of the day against a fresh opponent might not be compelling baseball. An alternative, then, would be to give one team an off day every day, but have some of the remaining teams play double headers. Those double headers aren’t strictly necessary, but the season needs to be compressed, and off days don’t exactly jive with fast seasons.

If baseball is sticking with the 108-game schedule, they could finish it in just over 100 days by having every day feature a double header and one team with an off day. That’s aggressively compressed, though: if you have 110 days to use, you could have double headers every two out of three days and still get through the entire schedule even with one team off every single day.

Of course, that’s on paper, and neither Florida nor Arizona come without challenges. In Arizona, the main challenge is heat; every double header would have to be played at Chase Field. That works out okay at first, but on the 30th consecutive day of double headers, the field is going to be pretty beat up, no matter how much work the grounds crew puts in. If baseball has the aforementioned 110 days, that would help; using every third day to refresh the field is a lot better than almost never having off days.

Meanwhile, in Florida, it rains. In the summer, it rains a lot. In Miami, more than half the days in July, August, and September feature some rain, and those three months will figure heavily into any plan. In theory, the domed stadiums in Miami and Tampa could handle some of these otherwise unplayable games, but darnit, rain doesn’t just happen in one city at a time. Stupid weather! If it rains near Tampa, at least five stadiums will be affected. That’s the biggest concentration of spring facilities, but depending on how big of a storm front is involved, every stadium could easily be unusable for a day.

Of course, scheduling makeup games is far more difficult when there’s no such thing as a mutual off day. Makeup games would need to be doubleheaders, which might tax an already difficult schedule for players. That doesn’t even get into hurricane season, which could be a massive wrench in the works; lose four days of the schedule to a hurricane, and the already-compressed season would be in shambles.

One possible solution? Abandon this facade of making baseball almost like a regular season and simply make both leagues even. The Rays train in Port Charlotte, a 90 minute drive from Tampa. They could use Tampa as their home base, and one of the 10 teams that shares a facility in Arizona could take Port Charlotte. Heck, they could even share Port Charlotte if the league would prefer to reserve Tropicana Field for double headers — it’s hot in Florida too, after all. Sharing a spring facility is nothing new; 14 teams share spring facilities. But most likely, Miami’s newer stadium could handle the twin bills and the Rays could have the Trop.

Of course, that makes 16 teams in Florida and 14 in Arizona. But quite honestly, who cares? Baseball used a 16/14 split for nearly 20 years with no real problems. The NL Central had six teams, the AL West had only four, and it’s not as though anyone looks back on the 1998-2012 era as having unfair competition.

That arrangement would require other scheduling quirks; in the 16-team Florida league, teams not in the six-team division would play six games against each team in the other five-team division, and only five against the six remaining teams. In the division with six teams, instead of playing 12 games against each divisional opponent, it would be a mix of 11’s and 12’s (they’d have an extra 10 intra-division games due to only playing non-division opponents five times). The opposite pattern would hold in Arizona; one division would play a longer divisional series (16 games instead of 12) due to its four-team construction.

For my money, that change is much less disruptive than the various double header-centric plans. You can still schedule two teams at a time for off days to make rest work; but as it’s no longer required, you won’t have the everyday off-day plan that was previously necessary. With two teams at a time scheduled for off days, there’s a much higher chance to schedule makeup games when there are rain delays. And with fewer double headers necessary to make the schedule work, teams could more reasonably play double headers as makeup games.

Heck, if the weird difference in divisional games is too much for your sensibilities, just abandon an unbalanced schedule altogether. Each team could play each other team in the Florida (or Arizona) league an equal number of times, give or take a few games at the end to make the two leagues have an equal number of contests. In a 108-game season, that would mean the Florida league consists of 12 seven-game sets and three eight-game sets. The Arizona league would be nine eight-game sets and four nine-game sets.

To be clear, these theoretical scheduling discussions don’t mean that the plan is flawless. The same quarantine issues that make playing in Arizona difficult would apply. Without massively increased access to testing, the plan won’t work. But if that side of the equation can be handled, I prefer this plan to the Arizona-only plan. Scheduling games for the East Coast is a big problem in Arizona, as is the need for nearly every field to be used every day.

This plan does a good job mitigating those problems. It won’t be perfect — there are two teams who play in Arizona whose home cities are in the Eastern Time Zone, and only one could head to Florida even in the 16/14 plan. There would be some unavoidable time zone issues for one of the Indians or Reds.

But short of that, I think it’s far better. The Florida weather is a smaller problem, in my eyes, than the difficulty of playing so many games in the desert heat. Any of several plans could handle the scheduling requirements, and all of those plans would make for better viewing than the Arizona-only plan. Some of the plans would also have their own weirdness; strange double headers or one team constantly resting or unbalanced leagues. But they’d all deliver on the key thing we need: baseball, and live baseball at times when people could watch it. Solve the non-baseball issues, and this Arizona-Florida plan could handle the rest.