September 27, 2016
The San Francisco Giants, you may have heard, have not had a good second half. At the All-Star break, they were 57-33. That was the best record in baseball, three games ahead of the Cubs. Since then, they are 25-41. That’s not the worst record in baseball—the existence of the Minnesota Twins ensures that—but it’s the worst in the National League. They’ve fallen from first place in the NL West at the break--6.5 games ahead of the Dodgers--into a tie with the Mets for the Wild Card and just half a game ahead of the Cardinals, trailing the division-clinching Dodgers by 8.0 games.
In doing so, they remain in line to set a record for the biggest swing in winning percentage from the first half to the second half. Their drop of .255--from .633 to .378--is the greatest since the first-half/second-half dichotomy was established by the 1933 All-Star game, ahead of the 1943 Philadelphia Athletics, who followed a 34-44 first half with a cover-your-eyes 15-61 second half, a .238 decline. The Giants have to win at least four of their remaining six games to avoid breaking those wartime A’s record.
You might think you’re hearing cries of “Howl! Howl! Howl!” from the Bay Area, King Lear-style, holding their even-year championship dreams in their arms instead of Cordelia. (I know, I know; technically, the Giants’ postseason is not yet dead.) Actually, they’re saying, “How?” as in, “How does a team have such a spectacular first half and such an abysmal second half?”
Let’s start with the hitters. At the break, the Giants were hitting .263/.336/.406--a .742 OPS. Adjusted for park (using Baseball-Reference’s OPS+), they were one percent above average. In the second half, they’re hitting .244/.314/.375--a .688 OPS that’s 12 percent below average, park-adjusted. But the swing in OPS+, while certainly not good, isn’t over-the-top horrible; the largely Giancarlo Stanton-less Marlins have slipped from 101 to 86, a drop of 15, while the still-in-contention Orioles’ offense has fallen off 22 points, from 115 to 93. So it can’t be just the hitting.
How about the pitching? Again, it hasn’t been pretty. The Giants allowed a .245/.301/.394 line in the first half--a .695 OPS and 87 OPS+. In the second half, they’ve surrendered .242/.309/.401--a .710 OPS and 93 OPS+. So, while the pitching hasn’t been really bad, it’s certainly been worse than it was before the break. The six-point decline in OPS+, though, is pretty minor. The Cardinals have remained in the Wild Card hunt despite allowing an OPS of .762 in the second half of the season after giving up only a .687 in the first half, an OPS+ decline of 21, and the division-clinching Dodgers have experienced an 18-point OPS+ decline.
But the Giants have been worse in both hitting and pitching. Let’s combine them. In the second half they’ve been 13 points worse in OPS+ on offense and 6 points worse in pitching. That total of 19 points is pretty bad, right? Well, yeah, but it’s not amazingly bad. It’s not even the worst in the National League. That honor goes to St. Louis, which has a net declinke of 30 points of OPS+. Nor is it the second worst in the National League. That’d be the Diamondbacks and their 23-point OPS+ decline.
Remember, we’re talking a historically precipitous decline in winning percentage for the Giants. Their decline of 19 OPS+ points from the first half to the second half is in a 28-way tie for the 128th-worst in the 1,924 team seasons since 1933. But OPS is a process stat, it’s not an outcome stat, at least in terms of run production. Maybe the Giants aren’t totally hopeless in generating (or preventing) OPS; maybe the problem’s one of sequencing, or, as Joe Peta put it, “cluster luck.”
For example, on September 16, the Angels got eight hits and five walks but scored no runs. The day before, the Diamondbacks got two fewer hits and two fewer walks and scored seven runs. Maybe Giants hitters are like those Angels, spreading out their hits and walks, while their pitchers are like the Diamondbacks, bunching them up in a few high-scoring innings. That’d suggest the problem with the Giants is run production.
They’ve certainly scored fewer runs in the second half. Their 4.71 runs per game in the first half--fifth in the league--has shriveled to a dead last 3.83 per game in the second half. Only the Orioles (-1.05 runs per game) and Cardinals (-0.93 per game) have had a steeper fall than the Giants’ -0.88 per game. And Giants pitchers, one of only five teams to have held opponents to fewer than four runs per game in the first half—3.90, to be exact—have given up 4.08 per game in the second half.
Combining the 0.88 decrease in runs scored per game and the 0.18 increase in runs allowed per game is a net 1.05 reduction in run differential per game, which is ... well, not great, but not as bad as the Cardinals’ 1.46 runs per game slide. The Giants’ reduction from the first half to the second is the 123rd-worst of all time, which certainly doesn’t translate into the largest reduction in winning percentage ever.
So maybe it’s how the runs are being scored? In the first half, the Giants scored 424 runs and allowed 351, a run differential that’s predictive of a record of 52.70-37.30. The Giants were 4.30 games better than that. In the second half, they’ve scored 253 and given up 269, which should yield a 31.15-34.85 record. They’re 6.15 games worse than that. So that’s 10.45 games’ worth of difference just from run differential.
Has that run differential resulted in a lot of one-run losses? That’d suggest misfortune, as I’ve shown that winning percentages in one-run games are almost entirely random and a product of luck. The Giants have an overall 28-27 record in one-run games this year. But they were 20-10 in the first half and are 8-17 in the second half, flipping from pretty lucky to pretty unlucky. Assuming their true talent is equal to their .526 winning percentage to date, they managed to win 4.2 extra games in the first half by winning two-thirds of their one-run contests. In the second half, when they’ve lost 68 percent of one-run games--it’s cost them 5.1 games.
And why have they done poorly in one-run games? Obviously, scoring fewer runs and allowing more has been a contributor. But, rather conspicuously, there’s also been a problem with the bullpen. Giants relievers, to be fair, have a lower ERA in the second half (3.40, fifth in the league), and while they’ve blown 13 saves in the second half, they blew 17 in the first (albeit in more games with more frequent save opportunities).
But that doesn’t really tell the whole story. NL teams have won 91.2 percent of games they’ve led after seven innings. The Giants have won 84.4 percent, last in the league. NL teams have won 95.6 percent of games they’ve led after eight innings, but the Giants have won 88.0 percent, also last. And, as you’d expect, it’s been particularly grim in the second half: 72.0 percent when leading after seven, 75.0 percent when leading after eight. Had the team been league average, they’d have won nearly five more games in the second half.
So why are the Giants, who’ve been merely garden variety bad in terms of OPS and runs scored/allowed since the All-Star break, on pace to be the worst ever in terms of winning percentage decline? Because they’ve been unlucky as well. Their bullpen’s been unreliable at holding leads, which has been a factor in a poor record in one-run games, which has contributed to a record that’s fallen well short of their production. I’m not saying that this has all been due to misfortune—the late-inning meltdowns, the failure to drive in runners, that’s all been real. But there’s been an element of poor timing and bad breaks, and the poor timing and bad breaks haven’t evened out.
This isn’t to say that the Giants have been good. They haven’t been. But take a good, if somewhat lucky team, in the first half of the season, turn the dial on pretty much all of their underlying metrics from good to bad at the All-Star break, and throw in a generous dollop of bad luck—well, you get a team that’s on pace to challenge a war-depleted A’s team that had only one hitter with a better-than-average OPS and whose best pitcher was a 28-year-old rookie.
*Unless they make it to the postseason, and MadBum goes crazy again and the whole even-numbered thing becomes a nightmare for the other 29 teams.