Last year, the Yankees won 103 games. They did it nearly entirely without Miguel Andújar, Luis Severino or Giancarlo Stanton, and with Aaron Judge and Gary Sánchez each missing considerable chunks of time. With a return to health from that group, you might have thought, plus the addition of Gerrit Cole on a record-shattering contract, plus the Red Sox trading Mookie Betts, the AL East would be theirs for the taking. They'd win 115 games. It's over before it starts.

Maybe so. Every projection system gives them a healthy edge. But ... the Yankees don't start at "103 wins," because that's not how this works. You might also have noticed that New York's spring, so far, has been plagued by serious injuries to talented players, and that lurking within their own division is a very good Rays team that just won 96 games of their own.

So, we thought, it might be interesting to ask the all-important question: What would have to happen for the Rays to actually overtake the Yankees in 2020? How unreasonable would that be?

We started, as we often do, at FanGraphs. They project the Yankees to win 95 games,* and the Rays to win 91. Those numbers are the midpoint of thousands of simulated seasons, or the 50th percentile of all of those simulated outcomes, because obviously the Yankees aren't winning 95 and the Rays 91 every single time. Sometimes, it's a lot more. In the worst-case scenario, it's a lot less. What we're interested in is the version of this where things go really well for the Rays (the 75th percentile, where they would be projected to win 95 games) and at the same time, things go pretty poorly for the Yankees -- the 25th percentile, where they would be projected to win 91 games.

*Why 95, an eight-win drop from last year, even with Cole added? It's because a 100-win season generally is an outlier, not an expectation, and that projection systems work by looking at "this year's collection of players," not "starting with last year's win total." It's also that you can't expect everyone to be healthy and for all of last year's out-of-nowhere success stories like Mike Tauchman and Gio Urshela to repeat themselves. You probably also shouldn't assume the Yankees will go 17-2 against Baltimore again, though if you want to joke "yes, because it will be 19-0" we wouldn't blame you. And it says a lot about how stacked the Yankees are, we think, that the "things go poorly" scenario for them is "still winning 91 games."

You can see what we mean on the distribution chart below, which shows that the Yankees are expected to win more games most times, but that there are versions of this timeline where they don't. That's what we want to get to.

OK, so: How do we get there? What happens for the Rays to get to that better-than-projected year and for the Yankees to get to that worse-than-projected year? We should note that this isn't fantasy-land stuff, because a four-win gap isn't that much, and the FanGraphs odds have the Yankees with a 55% chance to win the division, with the Rays at a healthy 28%. (The remainder is with unlikely runs from Boston or Toronto.)

That 28% is roughly one-in-four. Let's find that one.

The simple answer is: The Rays stay healthy. The Yankees don't. Then again, the Yankees didn't stay healthy last year, and then what happened? Let's try to be a little more interesting here, and instead look into the Steamer projection system -- one of the most respected projection systems, and part of what fuels the FanGraphs projections section -- to try to find what might bring us to this point. In order to try to match the chart above as closely as we can, we'll look at the 20th (bad) and 80th (good) percentiles of the Steamer projections.

What's that look like? Like this.

How things go poorly for the Yankees

We're more interested here in showing how the Rays could be really, really good this year, but a quick refresher on the year-to-date for the Yankees:

• James Paxton (back surgery) out through May/June
• Domingo Germán (suspension) unavailable through June
• Aaron Hicks (Tommy John) out through midseason
• Severino (Tommy John) lost for the year
• Stanton (calf strain) no timetable for return
• Judge (shoulder) no timetable for return

Both Stanton and Judge are expected to miss Opening Day, though they might each be available shortly thereafter.

Injuries aside, here's a few ways to get the Yankees to the bottom end of their projected win scale, looking at those 20th percentile projections. Do any of these feel that unrealistic?

• Urshela (.369 wOBA in 2019) posts a .289, better than the .261 he actually had between 2015-'18
• Brett Gardner (.344 in 2019) posts a .296 at 36, only slightly worse than his real-world .306 from 2018
• Hicks (.325 in 2019) returns to post a .300, struggling post-Tommy John as Didi Gregorius did last year
• J.A. Happ (4.91 ERA in 2019) hits that 20th-percentile 5.96 ERA in his age-37 season

... and so on. You get the idea. There's a version of this for literally every player.

It's not likely all of these things happen, but it's pretty likely that at least one of them does; remember, we didn't even hit on Tauchman and his own breakout here. While 91 wins seems like a total disaster -- and it would be, given the expectations -- remember that the team very recently went five years, from 2013 through '17, without topping that number. These things do happen.

How things go well for the Rays

Here's a four-year win trajectory for Tampa Bay, dating back to the 2016 season: 68, 80, 90, 96. By that trend, the Rays are guaranteed to win over 100 this year.

We're kidding, obviously, but the point is that after posting their worst record in nearly a decade in 2016 -- Kevin Cash's second as manager -- the needle has been steadily pointed upward since. (Not just at the Major League level, too; both The Athletic and FanGraphs have the Rays with baseball's best farm system, led by super-prospect Wander Franco, MLB Pipeline's No. 1 overall prospect.)

The Rays are projected to have the third-best starting pitching, a tick ahead of the Yankees, and the fourth-best relievers, just behind the Yankees. (We cannot stress enough how good somewhat-anonymous relievers like Nick Anderson, Colin Poche, José Alvarado, Diego Castillo, and Oliver Drake are or can be. It's fair to note that Blake Snell is dealing with some elbow discomfort, however.)

Combine those, and the Yankees and Rays are essentially tied with the Mets for the best projected pitching in baseball. Let's do the same exercise but in a more positive lens; what does it look like for some Rays who hit their 80th percentile projections?

• Kevin Kiermaier (.284 wOBA in 2019) posts a .325, or almost exactly what he did from 2015-'17
• Hunter Renfroe (.320 in 2019) posts a .336, or basically what he did in 2018 (.338)
• Mike Zunino (.235 in 2019) rebounds to a mere .311, actually worse than the .325 he had from 2016-'18
• Charlie Morton (3.05 ERA in 2019) posts the same 3.05, holding off age at 36 for one more year
• Tyler Glasnow (1.78 ERA in 60 1/3 innings in 2019) posts a 3.11, but he does it in 154 innings

Again: Not all of those things will happen. Several definitely won't. Not everyone is going to be the best version of themselves at the same time half of the Yankees are the worst versions of themselves. But the point here stands, which is that on an individual basis, none of these things are terribly unlikely or unrealistic.

A handful of games over a full season, after all, is somewhat of a narrow gap. If anything, the focus ought to be on the 19 head-to-head matchups between the Yankees and Rays. Last year, the Rays dropped 12 of the 19. Just flip a few of those games, and suddenly there's a real race here. That's underselling, it actually; there's already a real race here.