Going into this season, ZiPS projected **Madison Bumgarner** to have a 3.93 ERA and a 1.32 WHIP built on the back of a 5.84 K/9 and a 2.78 BB/9. It’s safe to say that he outdid those conservative numbers this year. Was this more about improvement or consolidation of talent? In other words, did Bumgarner say the same in the face of regression, or did he take fundamental steps forward this season?

First, let’s examine the role of luck in his season to date. His BABIP is currently .326, which looks high for a starter with a 8.15 K/9 and 46.1% groundball rate. Indeed, using Matt Swartz’s xBABIP for SIERA calculations, he should have a .293 BABIP. He’s stranding 70.9% of his runners and the league’s been stranding 72.4%. This is not some extended run of luck by the standards set by these two stats.

So we are back to our original question. It’s clear that, in at least one area, Bumgarner has taken a step forward. His swinging strike rate this year is 8.8%. That means he’s above average (8.5%) this year after being below-average (7.6%) last year. This is responsible for his leap in strikeout rate from below-average to above-average (6.97 K/9 to 8.15 K/9).

He’s still throwing the same pitches, but the mix is slightly different. The slider has become a center piece, going from 20.9% to 31.9% usage. The changeup has suffered, dropping from 10.3% to 3.2%. It’s a bit surprising, given the fact that his changeup was a better pitch by linear weights pitch types last year (5.2 runs to 0.2 runs above average). Now he uses the pitch almost exclusively against right-handed batters (he’s thrown four against lefties all year), taking advantage of the fact that it’s a platoon-neutral pitch by most research. Maybe his coaches noticed that he gets more whiffs and ground balls from his slider (true over 1500+ pitches in two seasons), or maybe he just personally decided to use his better pitches more.

Other than a consolidation in his pitching mix, there’s been a slight step forward in velocity. Most Pitch F/x tracking systems have his fastball a half MPH faster this year, and his slider almost three MPH faster. Other than his slider getting a little straighter (one-plus inch of horizontal movement less this year), the pitches are moving the same. Obviously, this amount of velocity is not the explanation for his newfound success, but it works hand in hand with the change in pitching mix to help put his new swinging strikes in context.

With a groundball rate that’s a mere percentage point above last year’s number, and an almost identical walk rate, it’s tempting to say that the swinging strikes explain all of his improvement. But we do have to give Bumgarner some credit for fighting off regression. He’s always had great control in the minor leagues (1.9 MiLB BB/9), so maybe that wasn’t so surprising. But he only had above-average ground-ball rates in the minors (45.2% at Triple-A in 2010), and most pitchers would expect to show a worse ground-ball rate in the major leagues. Avoiding regression in that category was a particularly heartening fact for his future.

Madison Bumgarner is ninth in the league in SIERA among qualified starters (eighth in xFIP and fourth in FIP). That’s an elite place to be, and it’s remarkable that he’s done it without being elite in any of the main three factors. He’s 21st in K%, 19th in BB% and 46th in GB%. Put it all together, though, and he’s a well-rounded, and yes, elite pitcher.

And he got there with a little mix of getting better and staying the same.