The San Francisco Giants entered play Thursday with the best record in baseball at 38-21, a game ahead of their cross-bay rivals in Oakland and a full 3.5 games better than any other National League team. Given their recent history of World Series success, maybe this doesn’t stand out to you. If you know the team and have been following along, however, this seems more than a little surprising, because the Giants certainly don’t seem like the best team in baseball.
That’s not to disparage them – they’re a good team, to be sure, and their Pythagorean win-loss record of 36-23 doesn’t indicate that they’ve been benefitted too much from run distribution.
But that’s just based on run differential, and there’s certainly a lot of variance and good fortune that can go into how teams produce runs. For the Giants, well, being the clutchiest bunch of clutches who ever clutched is certainly helping.
The Colorado Rockies blog Purple Row wrote a bit about how the Giants have over-performed in high-leverage situations in the early parts of the season, a team-wide “skill” that doesn’t appear to be very repeatable outside of St. Louis. Chalking high-leverage performance up strictly to luck probably isn’t the best approach, but the degree to which San Francisco has outperformed every other National League team in high leverage situations at the dish is striking:
While they rank tied for fifth overall in high leverage wOBA, they’re the top NL team and rank as above-average in terms of the number of high leverage plate appearances they’ve had. That is, not only have they been really good, they’ve done it over more situations than most.
The reason this stands out isn’t just that they’ve been good, though, but rather that on the whole they’re a completely average offense, ranking 16th in wOBA and sixth on the senior circuit. They’ve “saved” their best performance for when it matters, in so much as any team could possibly control that. Have a look at how their high leverage wOBA compares to their wOBA in other situations:
|Team||HighLev PA||HighLev wOBA||MedLev PA||MedLev wOBA||LowLev PA||LowLev wOBA|
That’s a lot to sort through, so what we can do for a rough gauge is compare high leverage performance with low and medium leverage performance, simply looking at the percentage change when the heat gets turned up:
|Team||% of PA hiLev||HiLev/LowLev||HighLev/MedLev||HighLev/Other|
The Giants are performing 11.7 percent better when the leverage is turned up, the biggest gain of any National League team. What’s more, they’ve saved their worst performances for when the leverage is the lowest, ranking only behind Cleveland in the gap between high and low leverage performance at 22 percent.
There are a few ways someone could classify that as “luck,” especially considering the largest “high versus low” gain over a full season in 2013 was just 16.4 percent (Kansas City), and they were the only team to top a 7.1 percent high leverage gain. Again in 2012, an 11.5 percent mark (Cincinnati) topped the list, and in 2011, two teams managed a double-digit percentage gain but none topped 13.9 percent (Milwaukee). In short, recent history suggests a team can’t sustain being 20 percent better in high leverage situations than low leverage ones, so the Giants are surely due to regress some.
How much has this benefited San Francisco on the offensive end? Well, that wOBA in high leverage situations has been worth 5.6 weighted runs above average, and their offense in other situations has been 9.7 runs below average. Inputting their wOBA in medium and low leverage situations in place of their .344 mark in high leverage spots, they would lose 6.7 weighted runs, enough to drop their Pythagorean win-loss record another half-win. That’s some rough back of the envelope math, but needless to say this unexpected performance in the clutch has probably added a win or so to their total based on math alone, but may have manifested itself in specific situations to help lead to more real wins. The fact that their projected winning percentage moving forward is just .524 rather than the .644 they’ve amassed so far gives reason to be skeptical they can keep this up.
To this point, the Giants haven’t been all that fortunate in terms of every player outperforming expectations in high leverage spots, but rather that their best bats have tended to come up in high leverage situations. Hunter Pence has been incredibly “clutch” with a .458 high leverage wOBA, and he leads the team with 32 such plate appearances. Pablo Sandoval has struggled and Brandon Belt hasn’t had nearly as many important appearances as you’d expect, but otherwise the distribution of these appearances has been favorable.
|Name||HighLev PA||HighLev wOBA||Overall wOBA|
In part as a result of this, Pence and Morse both rank in the top-10 in win probability added this season despite ranking 32nd and 18th, respectively, in wOBA.
It’s been more of the same on the pitching side, too, with the Giants locking things down when the leverage is cranked up:
|Team||HighLev ERA||HighLev TBF||HighLev BABIP||HighLev FIP||MedLev FIP||LowLev FIP||High/Low FIP|
Here, the Giants have the lowest high leverage ERA by nearly a full run, trimming their FIP by five percent from low leverage spots. This is a case where “luck” is harder to use as an explanation, because a manager’s bullpen usage plays a key role. Consider how starters and relievers differ, first:
Relievers have lower FIPs than starters and tend to pitch in higher leverage situations, on average, which both make perfect sense, but the previous table clearly shows that pitchers struggle when the leverage is high. What we see with the Giants in particular is that Bruce Bochy has done well to get his best arms into the game when the situation calls for it:
|Pitcher||FIP||gmLI||HighLev TBF||HighLev FIP|
You can quibble with Romo’s usage some but he also has a multi-year track record to suggest he’s beter than he’s performed so far this season. And despite Casilla’s struggles with FIP in the clutch, he’s posted a 2.25 ERA in those spots. Like on the hitting side, the result is several Giants among the leaders in win probability added, with Machi, Casilla and Romo ranking 10th, 17th and 20th, respectively despite none of them ranking in the top-30 for FIP overall.
For the season so far, the Giants have been involved in the single game with the highest average leverage index, three of the top-50, eight of the top-100 and 17 of the 276 games that have had an average leveraged index of 1.5 or greater, not a disproportionate share at any cutoff.
In those games, however, they’re respectively 1-0, 2-1, 4-4 and 9-8, and they’re 13-14 in games with an above-average leverage index over the course of the game. That is to say, while their aggregate numbers in high-leverage spots are extreme, they haven’t made a difference in the highest-leverage games (they’re also 12-9 in one-run games, only 10th in the league). Instead, you get early-inning dominance when the leverage remains high (they have four first-inning home runs when the leverage is above 1.5 and are 17-2 when leading after one) and a bullpen that locks things down if the leverage ratchets up later (they’re 31-0 when leading entering the ninth).
None of this is to say they won’t be regressing over the last two-thirds of the season, because they almost surely will. Players and teams can’t just turn things up when their adrenaline rises, otherwise Brett Lawrie would be batting 1.000. Luckily for the Giants, the wins they’ve already achieved are in the bank, and regression can’t change their record to date. Even with some regression factored in, they remain the NL West favorites by roughly four games. And, of course, should the race come down to a high leverage final week, the Giants would be untouchable.
Print This Post