The ZiPS Midseason Update for the American League

As we reach the mathematical halfway point of the season and approach the trade deadline, this is an opportune moment to run an update of the ZiPS projected standings. The standings are based on projections from the most robust version of ZiPS rather than the simpler one, which is more practical to run daily during the regular season, implementing things like the Statcast-aided zStats and up-to-date minor league translations.

The process that ZiPS uses is the typical one, but I’ll run it down quickly for those who may be new to how these projections work. ZiPS starts with a modified version of our depth chart and applies a generalized probabilistic model of available playing time for the players listed. So instead of a team’s roster strength being a simple sum of everyone’s projected WAR pro-rated to a fixed expected number of plate appearances, we end up with a whole distribution of possible roster strength. As an example: While Jacob deGrom still has a median of 55 innings in the roster sims I run for each team, sometimes he’ll be at 65 or 70 innings, sometimes he’ll be at 30 or 45 innings, and occasionally, it’ll be much worse than that. ZiPS will then “fill in” playing time based on the next players available on the depth chart and their probabilistic measure of availability. Just to stay with the Mets: When the outfield is healthy, the depth chart is mostly Mark Canha, Brandon Nimmo, and Starling Marte. But on the particularly bad rolls, the team’s estimated roster strength will have a lot more Ender Inciarte, Nick Plummer, Mark Vientos, and even players like Daniel Palka and Terrance Gore.

After ZiPS gets a distribution of each team’s roster strength, it “draws” one each year and sims out the rest of the season, team versus team, a million times and sees what happens. Is this a perfect methodology? Absolutely not! But I think we get closer to our goal of trying to evaluate team uncertainty and team depth, something which is harder to do using a less time-consuming scheme.

For today, let’s check in on the American League.

Can the Yankees win 116 games? If the projections are to be believed, it’s unlikely, but it’s at least plausible, with the team reaching that mark in 2.9% of the simulations. The Yanks are currently on a 116-win pace, but even if not everything has gone right — hello, Joey Gallo and Aroldis Chapman — quite a lot has to for them to win games at such a torrid pace. The rest of the division is tough, and even the Orioles are showing a pulse, which is something that works against them. What trades might be made is beyond the jurisdiction of ZiPS, but I don’t expect the Yankees to be aggressive at the deadline, considering that the division and the bye are likely going to be locked up in four weeks. Winning 117 games would be cool, but I’m not sure it’s sufficient incentive to go for broke. If anything, the Yankees might go lighter on the innings than ZiPS expects, given the injury history of much of the rotation.

The Red Sox have recovered enough from their poor start that ZiPS now sees them finishing right where it projected them coming into the season, at 88 wins. The Rays have essentially played up to their prediction as well. Toronto is the team other than the Yankees that has seen the biggest boost in expectations, as the increase in the computer’s confidence in Alejandro Kirk and Santiago Espinal is larger than the disappointment in Matt Chapman and the injury to Hyun Jin Ryu. Kevin Gausman’s status is a bit uncertain right now; he has avoided a severe ankle injury, but just how quickly he’ll return and how effective he’ll be are very much up in the air. The Jays also project with the easiest schedule remaining of any team in the division.

Despite the trials and tribulations, ZiPS still sees the White Sox as the most dangerous team in the AL Central and a potential postseason headache in the playoffs thanks to a rotation with a high front end. Also working in the team’s favor is that ZiPS projects the Sox to have the easiest schedule of any AL team remaining. These are significant advantages, but the Twins get a very slight edge in the overall divisional probability because of their big advantage: a 6.5-game lead that’s currently in the books. With the Pale Hose currently 0–5 against the Twins, Minnesota also gets the tiebreaker in the vast majority of tie situations. Contrast that with a game 163 — the superior tiebreak format for anyone who was raised with a modicum of taste and moral character — in which the White Sox project with a 64/36 edge in a Lucas Giolito/Dylan Cease versus Sonny Gray matchup.

Cleveland is also currently ahead of the White Sox, but ZiPS sees the odds of the Guardians holding them off as much longer than that of the Twins. One thing worth noting is that Andrés Giménez has now cracked the top 50 in baseball for ZiPS WAR over the next five years. José Ramírez projects to finish the 2022 season with the most WAR of any AL position player who isn’t also a dynamite pitcher (he loses to Shohei Ohtani, 6.2 to 5.9).

Detroit’s offense still figures to be better than the embarrassing showing so far this year, but it’s a case of too little, too late. Complicating things is that some of the players who would be most likely to be traded this month, such as Jonathan Schoop and Robbie Grossman, have played so poorly that I’m not sure there will be much of a market for either. As for the Royals, they were projected to play like the Royals, they’ve played like the Royals, and they’re run, both on the field and in the front office, like the Royals, so I doubt you need ZiPS to run those numbers.

ZiPS projects the Astros to have the strongest current roster in baseball. They’re a bit behind the Yankees but have the second-weakest projected schedule in the AL and, at 1.2%, have their own shot at winning 116 games. Their stranglehold on the divisional crown is also likely stronger than New York’s; if the wheels came off this particular apple cart, the AL West does not have another team projected to be as strong as the Red Sox, Rays, or Blue Jays.

The Mariners are demonstrating the cruelty of run differential. While it certainly feels like a significantly worse season — standings are tallied by actual wins, not a reasoned evaluation of a team’s strength — the team has made some progress and is arguably stronger than last year’s overachieving unit. Our confidence in the quality of Logan Gilbert should be stronger than at the start of the year, and Robbie Ray has turned things around quickly over the last six weeks or so. This team still has a realistic shot at the last wild card spot or two.

Imagine having Shohei Ohtani and Mike Trout healthy and having fewer wins than the Orioles in the toughest division in baseball. Los Angeles’ early-season surprise standouts like Brandon Marsh and Michael Lorenzen have collapsed over the last month, and the roster has 16 players below replacement level over the last 30 days. Contrast that with the Yankees, who have only two, with one of them, Miguel Andújar, playing in just two games.

At the bottom of the division, Texas has shown progress this season but is probably too weak for a serious playoff run. The A’s, meanwhile, are ZiPS’ favorite to have the fewest wins in baseball at the end of the season (47.2%). They could at least bust out the sweet gold jerseys!