Projecting Wins for All MLB Teams
Will the Rangers win 100? Who wins the NL Central? The A's are going to do what in the second half?
I realize a good chunk of my readers are likely not interested in the info below and that’s OK. There is Astros “stuff” in there, but likely not enough for the average Astros fan to spend 10 minutes.
But I’m a baseball fan and a math fan, so here goes.
Back in 2021 I realized that the Fangraphs’ in-season projections for the San Fransisco Giants were wildly off. Though I don’t recall the particulars, as the Giants ground towards an obvious 100+ win season (finished at 107) Fangraphs had the Giants wins sitting in the upper 80s.
Anecdotally it seemed to me that whatever formula Fangraphs was using was (and still is) conservative. It’s baseball, so that’s not really surprising.
Those that write there are more intelligent than I am by a large factor, of that I have no doubt and this shouldn’t be taken as a swipe. I use their information and statistics daily.
That said, I have often been of thought that sometimes we make things more complicated than they need to be. Sometimes we develop complex formulas for simple questions and I believe projecting baseball wins is one of those things.
The Pythagorean Theorem in baseball is a good starting point for estimating wins. The problem for me is it’s too simple. In short, you’re estimating wins and losses without including the actual wins and losses.
So what about a formula that includes both run differential and actual wins and losses? Seems simple enough and when I back-tested it, it was more accurate than Pythagorean.
Is it “mathematically correct”? I’m sure it’s not, but that’s not as important to me, a non-mathematician.
With that in mind, below are the projected wins for every team at the All-Star Break. Keep in mind all three of these are “works in progress” and not “final answers”, and as we’ll see, I’m more aggressive than Fangraphs and less aggressive than Pythag in general, for the reasons above.
AMERICAN LEAGUE
The biggest difference of any team in MLB between myself (AP), Fangraphs (FG) and Pythag is the Texas Rangers, with the Pythag being even higher than I am.
The second biggest? The A’s. Fangraphs projects the A’s to go 30-40 the rest of the season, while I’m at 19-51 and Pythag says 18-52.
I get forecasting the Rangers decline. I saw their bullpen in action last weekend in person, but it seems to me that without injury or illness projecting a 37-34 finish for a team with that lineup is not a good bet.
There’s also a significant difference between the projections for the Rays, with myself and Pythag 7 and 10 wins above FG, respectively.
Everyone thinks the Twins win the Central.
My All-Star Break Projected AL Playoff Teams:
Rays, Rangers, Twins, Orioles, Astros, Blue Jays
NATIONAL LEAGUE
We all agree the Braves are the class of the league, we just vary on the degree of their dominance.
After that, it gets interesting.
The Marlins are my biggest conundrum, as I see a regression due. I used to say “regression coming”, but they’ve defied the odds, going 21-6 in one-run games and 4-1 in extra innings and have won 8 more games than Pythag says they should have.
The Central is a cluster, with each of us projecting a different winner.
In the West, we all have the Dodgers, but the order below that and therefore the Wild Card teams are not the same.
My All-Star Break Projected NL Playoff Teams:
Braves, Dodgers, Reds, Diamondbacks, Giants, Marlins
Minutiae:
Biggest Difference Positive: Rangers +12
Biggest Difference Negative: A’s -11
Teams with 0 Difference with FG: Angels (81) Nationals (66)
Team All 3 agree on: Nationals (66)
My formula is designed to dovetail into the actual results more and more as the season continues, because that means we have more data to base our projection on.
More than anything, this should be a fun follow as the season continues. I admit I’d like to be more “right” than “wrong”, but since this is now “in writing” and “memorialized”, I have a feeling I’ve unleashed the negative karma.
Questions and comments are welcomed, encouraged and appreciated.