The San Francisco Giants have never been an easy team to predict. In the early 2010s, they built a dynasty just a few years after parting ways with maybe the greatest player ever. During that dynastic run, they sandwiched seasons in which they didn’t even make the playoffs between each World Series victory. And just last season, they resurrected themselves out of nowhere with 107 wins, the most in the 140-year history of the franchise.
So it’s not surprising that the Giants might be the toughest team to pin down in 2022 as well. After a disappointing (and controversial) loss to the rival Los Angeles Dodgers in the National League division series last October, San Francisco is back at it this season with one of the best records in baseball through the first few weeks of play, even after dropping three of four to the New York Mets. But they have more to contend with than merely their opponents this season: The Giants are facing a wave of potential regression as large as practically any team has faced in MLB history. Just how hard it hits could very well determine the eventual outcome of the NL West race, and perhaps even the World Series.
One of the oldest maxims in sabermetrics is the plexiglass principle — the notion (as popularized by Bill James) that teams that improve drastically from one season to the next tend to fall back the following year. That truism goes hand in hand with the so-called Law of Competitive Balance — which states that teams with exceptionally good (or bad) records get pulled back toward .500 in the future with ruthless inevitability. And at the most basic level, the 2022 Giants are on the negative side of both precepts, to a historically extreme degree.
For one thing, teams with at least 107 wins are incredibly rare — there have been only 17 in MLB history — and their win totals tended to decline by a shade over 13 percent in the following season. Similarly, AL and NL teams since 1901 that improved their records by at least 25 wins (per 162 games) year over year1 gave back more than a quarter of their improvement the next year, on average. The Giants fit into both regression-heavy categories, plus another one — they were below .500 not just the season before their big breakout, but also the season before that (and before that… and before that).
All of that means San Francisco was only the fifth team in AL or NL history to win more than 100 games per 162 after at least two consecutive sub-.500 seasons (joining the 1918 Chicago Cubs, 1945 Cubs, 1961 Detroit Tigers and 1993 Giants). Those teams all felt the pull of gravity the following season, and if we drill down to the components of what made them unique, we can see exactly how much each factor ratchets up the probability of regression:
|Wins per 162|
|Teams that …||Teams||Season 1||Season 2|
|Won more than 100 games …||160||105.1||96.2|
|… and were below .500 the year prior…||11||103.2||91.4|
|… and were below .500 at least two years prior||4||103.1||84.2|
San Francisco’s leadership is well aware of the forces this year’s squad is fighting against.
“We hear the rhetoric out there, the narrative about the 2021 Giants that it’s unsustainable, that there’s regression, that guys aren’t going to be able to back it up this year,” Farhan Zaidi, president of baseball operations for the Giants, told MLB Network in March. “Look, 107 wins is tough to back up.”
Zaidi went on to say that he believed in the breakout players who enabled the Giants’ turnaround, and elsewhere he stated that there were valid reasons why his team could defy history — as well as the statistical projection systems (like ours) that doubt San Francisco’s ability to repeat in the division or even return to the postseason at all.
“What those projection systems have a [tough] time at is processing and identifying what are true breakthroughs for players who are coming off strong seasons versus just performance flukes,” Zaidi told Tim Kawakami of The Athletic. “And that’s where you get this notion of a regression to the mean. If a player just happens to get lucky with a bunch of hits falling in or a pitcher gets a lot of at-’em balls over the course of the season, you’d expect some regression in their performance if they’re coming off a strong season. But for players that made a swing change, that made a meaningful adjustment, that picked up velocity, that’s one that projection systems struggle a little bit more [with].”
The yearning to rise above regression cuts to the core of what actually causes it in the first place. Recently, no lesser authority than James himself wrote an essay reconsidering the Law of Competitive Balance, or at least reframing it as a series of adjustments that are (or aren’t) made in response to the perceived need to make them. Great teams fall back when their luck cools off, to be sure, but they also may be less likely to make beneficial changes to what had been a successful formula in the past — “if it ain’t broke, don’t fix it,” as the saying goes. Meanwhile, bad teams improve in part because there is a greater urgency to address problems, as well as a greater clarity as to what those problems are (something that winning, by contrast, can mask).
It is probably not accurate to say San Francisco is just working harder to maintain a high level of performance than did the other teams on the list above. Surely the 1919 Cubs wanted badly to return to the World Series, for instance. But Zaidi and the Giants do have the advantage of greater information, particularly as it pertains to which players had lucky seasons and whose production will likely be sustainable going forward. (Not that we wouldn’t want to know what the exit velo and xwOBA were for Charlie Hollocher in 1918!) This fact alone could go a ways toward helping San Francisco resist the regression that wrecked similar teams throughout baseball history.
As for tinkering with a winning formula (or not), the Giants certainly had personnel losses to address simply to avoid falling behind. Still-productive catcher Buster Posey retired, while starting pitcher Kevin Gausman and do-everything deadline acquisition Kris Bryant departed in free agency. San Francisco is hoping to replace their value through a mix of in-house promotions (touted catching prospect Joey Bart) and external pickups (Carlos Rodón, Joc Pederson) — though Zaidi also mentioned the expectation for other, less-heralded young players to step into bigger roles, and he stressed that an offseason spent searching for even more rotation depth could offset any decline to a pitching staff that ranked third in wins above replacement2 a year ago.
So far, the plan to defeat regression has been working better than history would have predicted. The Giants are 8-5 this season — yes, it’s absurdly early, but that translates to a nearly 100-win pace over 162 games — and they’ve come by that record honestly. Their Pythagorean winning percentage (.685) is better than both their actual 2022 winning percentage (.615) and their Pythagorean mark from last season (.635), while they also boast MLB’s 11th-best BaseRuns expected winning percentage (.563).3 If you didn’t know about the plexiglass principle and the history of similar teams, you would see San Francisco’s start to 2022 as simply a continuation of its 2021 success.
There are potential pitfalls still, from the absences of Evan Longoria and LaMonte Wade Jr. to newly acquired starter Alex Cobb’s recent trip to the injured list. The Giants are the fourth-oldest team in MLB (weighted by playing time), and they’re relying heavily on repeat production from 30-something hitters Brandon Belt, Brandon Crawford, Mike Yastrzemski, Darin Ruf and Wilmer Flores. It’s tough to win when betting against the aging process, even if you have better information than ever about how well your veteran players are holding up (from both a health and performance perspective). As such, our forecast still expects the Giants to have declined by 20 wins when this season is all said and done, and we give them only a 59 percent chance to get back to the playoffs, with reversion to the mean playing a major role in that prediction.
Every game the Giants win, however, represents another data point in favor of them beating those odds — and another victory for the theory that the regression wave can be avoided if you know how to outmaneuver it.
Check out our latest MLB predictions.