2025 Predictions

This is not a sabermetric piece (although a bare minimum of “sabermetric” considerations go into it, as described below). This is simply a fun exercise that I have been doing for as long as I’ve been a baseball fan, undertaken with the understanding that:

1) it’s going to be “wrong”

2) point estimates are not really what you would want here (so this year I’m adding equally questionable probabilities!)

3) many readers will completely ignore any such caveats and disclaimers

I start by trying to figure out the regulars at the nine offensive positions, five starting pitchers, and four relievers, using publicly-available projections with my own subjective tweaks (usually to estimated playing time, not performance) sprinkled in, and filling in the remaining plate appearances and innings that a team would need to fill with an estimate of the level of performance that would be needed to reconcile the league totals for runs and runs allowed to where I expect them to fall out. It’s certainly less rigorous than the full-scale projections run by BP, Fangraphs, Clay Davenport, etc., but more rigorous than simply blindly guessing. I also consider the projections produced by the three outlets I named to see if anything jumps out that the more crude approach is missing.

The rank order list below is my own, generally following my crude estimates but overridden if I feel appropriate. The more interesting part is at the bottom, where I’ll show some the crude win total estimates and corresponding playoff probabilities.

NL EAST

1. Atlanta

2. Philadelphia (wildcard)

3. New York (wildcard)

4. Washington

5. Miami

The Braves may be underrated after a season in which much of what could go wrong did (Chris Sale’s Cy Young award the obvious exception). To me they are the only team in MLB that on paper is even within shouting distance of the Dodgers, with the Phillies right on their heels. I actually have the Phillies allowing the fewest runs in all of baseball, which puts me out of step with other prognosticators. One could make the argument that they have as strong a rotation as anyone in terms of having five very solid projected starters, with LA having many more options but also worse bets to make 25 starts whether by design or not. The Mets rotation looks a little scary (Clay Holmes Opening Day starter is kind of wild), but a lot of folks have them ahead of the Phillies, and not just those dazzled by Division Series results. My subjective impression was that the Nationals would be closer to the wildcard chase, but when you put it to paper their pitching looks bleak and the offense has prospect names that suggest more than performance to date justifies. The Marlins just barely edge out the Rockies to avoid the worst projected record in the NL.

NL CENTRAL

1. Chicago

2. Cincinnati

3. St. Louis

4. Pittsburgh

5. Milwaukee

The Cubs and the Reds are separated by a mere .2 games in my projected records, with both under .500, and the Cardinals are less than a game behind. I expected more out of Chicago last year, and both Baseball Prospectus and Clay Davenport really like them rather than the biggest shrimp treatment they get here. .2 games is a thin enough margin that it’s possible I should list the Reds as the favorites given that I know the Cubs started 0-2, but I won’t cheat. I watched Terry Francona manage enough teams in Cleveland to not dismiss the possibility that he could have a very positive impact on a young teams that seemed to lack consistent direction under David Bell, but it’s not enough for me to put my thumb on the scale in their favor. The Cardinals are an interesting team to watch, as it seems they wanted to do more retooling then they did, perhaps willing to throw in the towel on 2025 and focus more on Chaim Bloom’s first season, yet in this division all is possible. I waited until the last minute to “finalize” the projections underlying this; it’s quite possible that I haven’t dinged the Pirates enough for what seemed like it might be their strength (starting pitching depth) becoming shakier as the spring progresses, and it seems foolhardy to pick the Brewers to finish last – none of the other three outlets concur.

NL WEST

1. Los Angeles

2. Arizona (wildcard)

3. San Diego

4. San Francisco

5. Colorado

The Dodgers are the obvious favorite, but even favorites coming off World Series wins don’t always repeat. The core of their offense is older and becoming more injury-prone, and Mookie Betts as a full-time shortstop is an underrated radical move. This is exhibit A for picking nits. The Diamondbacks figure to take step back offensively – they almost have to – but their pitching should be better (read: not atrocious), and if that comes to pass it should be enough to get them into the playoffs. San Diego is a fascinating team to watch as it seems their window is closing; it’s hard to see AJ Preller keeping the magic act going, where you trade Juan Soto but then bring in Dylan Cease and then the next spring flirt with shipping Cease out, seemingly wishing new prospect capital into being as he goes. San Francisco seems to be spinning their wheels with no particular organizational direction – is that a bad thing to say when a new Baseball Ops head was just installed? But in terms of spinning wheels they can’t compete with the Rockies.

AL EAST

1. Boston

2. New York (wildcard)

3. Baltimore (wildcard)

4. Toronto

5. Tampa Bay

A general comment on the AL – on paper there is no daylight to be had between the top nine teams separated by six games. The Red Sox starting rotation is shaky headed into the season, but if Garrett Crochet can handle a full workload and the injuries are what they appear to be, it could be very good by the end of the campaign, and the offense is pretty exciting. As I write this Kristian Campbell has just been given the early nod at second base, and they might have more where that came from. I’m picking a team that PECOTA has last in their division to win the pennant, but I’d argue that says more about the state of the AL race in 2025 than it does about our respective evaluations of Boston’s roster. New York’s spring training has been a litany of bad news and Aaron Judge’s playoff performance is starting to get him heat despite coming off his best regular season, so it feels like a zoo waiting to happen, but I think the most likely outcome is a good yet flawed Yankee team, which has been their norm in most recent campaigns. The Orioles no longer feel inevitable as a long-term force after less than stirring debuts from Jackson Holliday and Coby Mayo, but that’s more the cruel expectations created by a fluky good 2023 than a trap I fell into. In retrospect maybe they took the plunge for a Corbin Burnes a year too early. I am finally swearing off the Blue Jays, which means it may well be the year they get their act together, and by “swearing off” I mean giving them a 47% chance to make the playoffs, illustrating the artificial compression of point estimates. As it feels foolhardy to pick Milwaukee last in their division, that much more so the Rays, but somebody has to finish last in this division and I’m still giving them a 12% chance to win it.

AL CENTRAL

1. Minnesota

2. Kansas City

3. Cleveland

4. Detroit

5. Chicago

The Twins are both the weakest of the projected three AL division winners and, by reckoning, the AL team most likely to win a division title (just edging out the Astros), despite playing in a division that somehow shepherded three other teams into the playoffs in 2024. This of course says more about the division than about them, but a little better health would go a long way. Don’t let the second-place prediction for the Royals fool you, I see this as a team that should brace for big regression from Bobby Witt and their rotation.

The Guardians cannot expect anything like a replication of last year’s bullpen performance. At the risk of making proclamations about relievers, which is always a recipe for egg on one’s face, I think that Emmanuel Clase and Cade Smith are among the very best relievers in the game. I do not think the same of Hunter Gaddis and Tim Herrin, and last year it got to the point that almost everyone they were putting into the pen was more effective than one had any reason to expect (Pedro Avila and Eric Sabrowski being two of the most notable examples). The rotation is shaky, with the primary addition (Luis Ortiz) being shaky enough in spring training to raise concerns. Ben Lively is a poor bet to repeat his performance, much will be expected of Gavin Williams, and with Chase DeLauter not going to be an option as early as might have been expected, it’s difficult to expect much improvement from the offense. Perhaps players like Brayan Rocchio and Bo Naylor can take a step forward, but I hesitate to point out that Jose Ramirez has been pretty durable which has helped keep the offense afloat, yet that is unlikely to continue as he enters his mid-30s. Second base could be a black hole, and not the defensive black hole it was in a good way when Andres Gimenez manned the position. Despite this bearish appraisal of their 2025 chances, I understand the rationale from trading Gimenez and Josh Naylor, for signing John Means to try to be this year’s Matt Boyd or Alex Cobb rather than paying through the nose for...Matt Boyd or Alex Cobb. But 2025 feels like a retrenchment season.

Tarik Skubal is great (but is anyone that great?), and “pitching chaos” to support a weak offense for six months is a tough ask. I may list the White Sox as having a 2% chance to make the playoffs but I would not put any money on them at 50-1.

AL WEST

1. Houston

2. Texas (wildcard)

3. Seattle

4. Los Angeles

5. Sacramento

The whole AL is desperately close on paper, and some outlets have the top three here stacked on top of one another as they were in 2023, but I give the two Texas squads a little bit of daylight. The Astros shed two more of their stars, yet their offense remains solid on paper, their infield defense should improve despite the loss of Bregman by shipping Altuve to the outfield and bringing in Christian Walker, and they have enough arms that they should be able to keep a decent rotation through the uncertainties of the season. The Rangers are a good bounceback candidate, with the hopes of a full season of late season Wyatt Langford, a return to last season 2023 Evan Carter, and anything out of Jacob deGrom providing some extra upside beyond what I put on paper. The Mariners are held in higher esteem by the three outlets, but the offense remains a liability (they’re comparable on paper to Cleveland’s) and they need their rotation to remain healthy. The Angels are just drifting off in space; it’s interesting that the Minasian brothers are both in high leadership roles of aimless franchises. I’m calling the A’s Sacramento because that’s where they’re playing, this isn’t a personal protest like my embarrassing refusal to change “California” to “Anaheim” for many years. Their second half performance last year may have them slightly overrated in the mainstream.

WORLD SERIES

Los Angeles over Boston

Below, I offer some very crude playoff odds that are fed by crude win estimates and developed by a Monte Carlo simulation. To make it easy, I just worried about team’s relative positions and not forcing the league win total for a given “simulation” to balance. I split it up by leagues, fifteen teams at a time, and each team got 10,000 random numbers to map to a number of wins using the normal distribution. Even though team wins should follow a binomial distribution, the central limit theorem gives us license for this. I somewhat arbitrarily set the standard deviation to 8 wins. Theoretically, for a .500 team the standard deviation of team wins is SQRT(162*.5*(1-.5)) = 6.36. But we are dealing with uncertainty around the team’s actual strength, exacerbated by what injuries and other unknowns would do even if we had perfect estimates of each player’s rate of performance. If anything I should have made it higher, and in the future I might use actual errors of these preseason estimates to come up with a value to use.

I also added another kludge which is less justifiable, which to apply a normal distribution with standard deviation = 4, with the mean equal to the projected wins for each team, for each of the 10,000 runs. I meant this to represent some of the uncertainty around the true team quality, but it’s not quite right – teams should in general be drawn closer to .500. This actually results in more extremes than it other wise might – across my 10,000 draws Houston maxed out at 124 wins. But I do think that standard playoff odds error in treating each team’s strength as a constant and then simulating X seasons based on those strengths, rather than recognizing that those estimates are highly variable themselves. I’m not claiming these estimates are better – the win totals they start from are way too crude to even consider that, and the pennant/World Series estimates are even rougher – but they do try to account for uncertainty in the starting point of estimating each team’s strength.

For each of the 10,000 “seasons” thus simulated, I simply picked the team with the highest win total as division champ, ranked the division champs 1-3, and ranked the three highest win totals for non-division champs as 4-6 to give me playoff seedings.

Then I took a big shortcut to get pennant and World Series odds that understates differences across teams, which was to assume that each playoff seed had a fixed probability to win the pennant based solely on the playoff structure itself, assuming that all teams were .500 teams entering the playoffs. I just used my playoff odds model and set all team ratings to 100:

For the World Series, I assumed that the higher seed in their league would have home field, and then averaged the probability of winning across the two #1 seeds (these weren’t identical as I arbitrarily assumed that one league’s seed would always have home field advantage over their same-seeded opposite). I just uniformly applied these percentages to the probability that each team would earn a particular seed and called it a day.

One thing I’m showing in these estimates that you don’t usually see is the probability of each team having the best record in their league. Since I actually in theory value this more than a playoff pennant, why not act like it by paying attention to it? Of course, you can also see how those are much more likely to be achieved by the assumed “actual” best team than the pennant (with the caveat that the pennant methodology does artificially flatten differences in team quality).

I showed a decimal place on the percentages only because I didn’t want to show 0% for any probability; I certainly do not claim these are accurate to anything approaching that level of precision.