Madison Bumgarner Is Back to October Bumgarner

Madison Bumgarner has been having quite a month of August. He’s posted a 53:4 strikeout-to-walk ratio while allowing only six earned runs in 37.2 innings, and he’s looked almost as unhittable as he did last October, when he relentlessly took the ball for the Giants in high pressure situations during the playoffs. That’s not a coincidence, it turns out, as Bumgarner is currently exhibiting tendencies that are quite similar to the 2014 postseason version of himself.

Before we go down that particular road, let’s have a brief introduction to Bumgarner, 2015 starting pitcher. On these digital pages, we’ve featured an article on how well he hits — which is quite well indeed — but not much else in the way of analysis this season; I will remedy that fact in a brief, limited manner. We could spend an entire article about the minute changes Bumgarner has made in 2015. Instead, here’s a cliffs notes version:

• He’s throwing more fastballs than at any point since 2011.
• He’s basically ditched his changeup, as he’s relied almost strictly on a three-pitch combination of fastball, slider (referred to at times as a cutter), and curveball in 2015. Here is his pitch usage since 2010, his first semi-full season in the majors (courtesy of Brooks Baseball, which calls his slider a cutter):

• Finally, his command improvements from last season have stuck, as he’s posting a career-low walk rate (4.3%) and career-high strikeout rate (27.4%).

These are all good things. By the numbers, Bumgarner is perhaps the best version of himself that he’s ever been, and it shouldn’t be a surprise that he’s posting the best xFIP of his career while challenging his best in ERA and FIP.

That’s why this month has been extra interesting; because, in the midst of one of his best seasons, Bumgarner seems to be up to something. And that something just happens to be intentionally throwing slower.

That sentence is a bit of a setup, as it’s not like Bumgarner is now throwing his fastball 80 miles per hour and lobbing eephus pitches (even though he does occasionally throw a very slow curve). This is a more subtle change with how he throws his breaking pitches, but it’s one that is nonetheless tied to times of high success for the Giants left-hander.

First, take a look at the speed difference between his fastball and slider from May of 2011 up to today:

Notice the peak right before the valley for this year; that was October of last year. He was slowly stretching the speed differential between his fastball and main breaking pitch over the course of the 2014 season, and it all culminated in the fall of 2014, when he was virtually unhittable. To be clear, this is not Bumgarner throwing his fastball harder; this is him taking something off his breaking pitches (even when he was throwing harder during last year’s postseason — because adrenaline — the velocity gain was seen across the board, so that wasn’t the reason for the increased difference). Now, during his starts this August, he’s stretched that gap again, and we’re seeing the same sort of results.

Before we get down to the exact benefits of how this approach is working, we should take a look at the difference between his fastball and curveball:

The difference with the curveball is just as large as with the slider. Again, notice the most recent peak before this season: it’s October of 2014, and he’s currently matching that. The other peak I’ve labeled is August of 2012, which was another very good (but not incredibly great) month for him (2.97 ERA/3.25 FIP). Then again, he was a different pitcher in 2012 (with a lower fastball velocity and a lower rate of curveballs thrown), so we should take that into account.

This leads to how the speed difference actually impacts Bumgarner’s performance. As much as we might like to point at these graphs, yell “he’s Bumgarnering!” and not look back at the amazing explosion that we created and are subsequently walking away from, let’s perform our due diligence. Is the greater speed difference between his fastball and breaking pitches a main driver in his improved performance?

That question can be answered best, in my estimation, by looking at the whiff rates of his various pitches to see if they match up with the increases in speed differences. Let’s take a look at his whiff/swing over time on his three pitches, with the three data points we’ve discussed highlighted (August, 2012; October, 2014; and August, 2015):

The first takeaway is that the slider has above average whiffs in months of greater difference (his career rate from 2011 onward is 23.5% whiffs/swing). It might not be eye-popping, but it’s definitely there. There’s always going to be noise with looking at monthly whiff rates due to the size of the samples, but the curveball is really where we can actually see some noticeable spikes. In months of greatest difference, we see peaks in his curveball whiffs/swing, far above his average rate of 37.6%. The fastball doesn’t seem to respond much to the differences, even though Bumgarner is posting close to a career-high whiff rate on his fastball this August.

I ran a couple linear regression models for the whiff rates and speed differences between each pitch, and though we get “significant” (or very close to significant) p values associated with the relationship (.05 for the curve, .06 for the slider), the r-squared values were very low, as we might expect. There’s a lot of variation in baseball, and it’s not too much of a surprise that speed differences between different pitches can’t explain the entirety of why a pitcher is effective.

What we do see, however, is a big uptick in how effective his curveball becomes when he stretches this speed difference. In the months that we’ve highlighted, when the speed difference has been greatest, most of his strikeouts have come via the curve. Last October, 46.7% of his strikeouts came on his curveball; this August, it’s been 36.5%. Maybe that’s why he’s throwing the pitch more than he ever has, and maybe that’s why he’s slowing the curve down, hoping to improve its effectiveness at getting swings and misses.

This isn’t the only reason behind Bumgarner’s hot streaks. When you have impeccable control, command over three plus pitches, and swing-and-miss stuff, you’re going to go on runs in which you’re almost untouchable. But, as we see, there are things that even established pitchers do to try and make themselves even better. During one of Bumgarner’s best years, he’s slowing some things down, and it’s working — all without an appreciable change in pitch movement. Not many pitchers can say they intentionally do that.