PECOTA 2025: Minnesota Twins Take the American League Central Crown

Hang the PECOTA division title banner. 

It's easy to forget in the underwhelming offseason the Twins have had because of the self-imposed payroll limitations dictated from the Pohlads that their AL Central rivals haven't exactly set the world on fire either.

I'm mildly skeptical about projection systems, but part of that is in my head I want to use them for the wrong things and treat them as being more predictive than they're intended to be.

one thing I fine interesting about systems like PECOTA is they're not as infected by recency bias like fans are. The late-season implosion in 2024 by many players isn't going to be seen as predictive for 2025, just another 6 weeks of the total season. There's some good in that. (There may also be some bad in that as it's possible that players may have gotten exposed down the stretch and a projection system can miss that)

I do think the Twins roster is still a good one, with hitting strengths, quality starting pitching (and depth) as well as a bullpen that is very good at it's top end with potential to be great with decent performances/health from some of the guys further down the line. Biggest weakness is defense right now; while there are some quality defenders, there's also some areas where it's tough.

I will say I'm not surprised to see Castro rated relatively poorly. His Tigers track record will be taken into account, but he also had a poor second half, belying his all-star status. I think he was probably a bit overused last season myself.

48 minutes ago, Vanimal46 said:

PECOTA was off on projecting the AL Central by an average of 13 games last season. Forgive me for not putting any value into their projections. 

PECOTA 2024 AL Central:

MIN 88 wins, actual 82 wins

CLE 83 wins, actual 92 wins

DET 75 wins, actual 86 wins

KC 70 wins, actual 86 wins

CWS 66 wins, actual 41 wins 

All projection systems are wrong, every year. That doesn't mean they are useless. It is more interesting to figure out where the assumptions diverged from what actually happened.

FYI - No projection system is going to pick a team to win just 41 games.

Unfortunately...this means nothing. 

There is utility in looking at statistics and analytics, which is why teams hire a raft of employees to dig and scape for any advantage or edge possible that improves any player to increase potential win totals. 

For those of us not employed by an MLB team crunching data, these sites and minutiae of collected numbers are good entertainment. When i'm working specifically with a pitcher (in the past) on refining one pitch and changing the shape of another pitch based on some information on hand, the close proximity to the task at hand has consequences and verifiable successes and failures. Reading the data and not having any personal hands on knowledge means the data is just numbers. When Joe Ryan works specifically with one of the many businesses/people tracking his delivery and pitches and then using information in conversation with Joe to find subtle improvements, there is a close knowledge of how that works. Looking over the data without seeing the guy pitch is an incomplete picture, which is what all of these sites give us. They are mere entertainment and I enjoy them too but it is best to put it into context. 

Thus ______ (fill in the name of the site/ prognosticator) publishes a tome of information telling us where our players and teams should reside. Guess what? Sometimes they are close because numbers do tell a good story. 

The Twins benefitted from their residence in the AL Central for a few years as Chicago, Detroit, and Kansas City faltered and Cleveland struggled with money. Last year we began to see a natural rise of a couple of teams (Detroit and Kansas City), which should continue in 2025. The AL Central will almost certainly be a close race between four teams and the White Sox will be more competitive while losing 100 games. Don't bet on any of the teams at this time to win 90 games. Expect 82-88 wins for all of Detroit, Cleveland, Minnesota, and Kansas City. 

This Just In: Projection system ran by spreadsheets adores team put together and ran via spreadsheets!

Twins had a pretty good team in '23. In '24 we still have pretty much the same team, especially on paper. Last year we were projected to win the AL Central, afterward, they made some changes but still looks good on paper but the difference is in the intangible of chemistry & unity. '25 we are projected to win the AL Central, we still look good on paper, but what'll boil down to is how much chemistry, leadership & unity from the core are maintained & health.

33 minutes ago, bean5302 said:

I think what @Vanimal46 is getting at is PECOTA is consistently wildly inaccurate. Their model is projecting 86 wins as enough to win a division when 86 wins hasn't worked in 16 years, and it's even very rare less than 90 wins gets it done. Their model is forecasting extreme outliers to start with.

FGDC + 45 wins baseline at least projects the Twins as a 90 win team for the lead in the division. That provides a some modeling credibility where PECOTA has little to none.

More likely they're heavily regressing the model than forecasting extreme outliers. 86 wins is not an outlier season.

Every projection system should have to publish their error bars. Median projections are not that interesting.

Very interesting.  Thanks for bringing it to us Matt.

Spending six weeks out here in the California desert where my wife's sister and husband live.  They/she are huge Dodger fans who were season ticket holders until they left LA a few years ago.  Sure didn't need to read anything about the Dodgers being most likely to win it all, I hear that crap every day.

Said several weeks ago that I believe Castro has played his best half season of his career and that it was likely that one of Helman, Keaschal. Martin or Keirsey would be as valuable to the Twins in 2025 as Castro.  Thus, I remain hopeful that he will be included in a trade before spring training is over.  Just might be some GM out there who believes in the player he was the first half of last year.

As for the projections, should I be ordering playoff tickets? 

24 minutes ago, Doctor Gast said:

Twins had a pretty good team in '23. In '24 we still have pretty much the same team, especially on paper. Last year we were projected to win the AL Central, afterward, they made some changes but still looks good on paper but the difference is in the intangible of chemistry & unity. '25 we are projected to win the AL Central, we still look good on paper, but what'll boil down to is how much chemistry, leadership & unity from the core are maintained & health.

I don't know why their intangibles would change. None of their personnel is different.

I'll go by what I've seen on the field lately. Cle, Det, and KC all made the PO's. Det and KC have both made some helpful upgrades on the field. Cle did some reshuffling which was far from retooling. MN has added Gasper and Mike Ford to an 82-80 team that was that good thanks to a 12-1 record against a team that lost 121 games. In addition feasted on the losing A's and Angels. These three organizations are also improved in different ways. 

We are sold that we will be good because? Buxton and Correa will be healthier in spite of being older? That recipe didn't work for me. Foregive me for not being more excited about our 1st place finish in PECOTA

One piece of information about PECOTA that would be useful to know.  PECOTA was originally designed by Nate Silver (remember FiveThirtyEight, political predictions?).  Now that I know that tidbit of information, I have no faith in this model as my first assumption is that it will be as far off as anything Nate Silver has ever developed.

3 hours ago, Vanimal46 said:

PECOTA was off on projecting the AL Central by an average of 13 games last season. Forgive me for not putting any value into their projections. 

PECOTA 2024 AL Central:

MIN 88 wins, actual 82 wins

CLE 83 wins, actual 92 wins

DET 75 wins, actual 86 wins

KC 70 wins, actual 86 wins

CWS 66 wins, actual 41 wins 

That is why we watch the game, isn't it? If it was all based upon what the computer models say, it wouldn't be any fun. Plus, who would have bet on the White Sox to lose 120 games a year ago? That was definitely fun to watch.

Personally, I think this is going to be a competitive division. Outside of the White Sox, who are still going to be terrible this year (probably not 120 loss terrible), all four of the other teams have a realistic chance of winning the division.