That's undoubtedly impressive, and so the question is obvious: How is Lincecum trying to work around his lost velocity to find success? There's a definite change in approach, but there's also some very clear signs that we're already seeing the appearance of the regression that was always inevitable.
You could ski down the decline in Lincecum's average velocity, which has annually lost approximately 1 mph from his peak of 94.6 mph when he was winning back-to-back National League Cy Young Awards in 2008-09. Sure, 88 mph might work if you're a DeLorean, but as far as fastballs go, well, it's slower than Chris Archer's slider.
Despite efforts to regain the lost velocity, it hasn't happened, so in response, Lincecum has done the only thing he could do: he's stopped throwing them as much. Once regularly responsible for over 65 percent of his pitches, he's throwing his two fastballs only 47.4 percent of the time these days. Lincecum has particularly diminished that four-seamer, which appears a career-low 22.5 percent this season.
Instead, Lincecum is relying more than ever on his low-spin split-change creation, which comes in at 1,412 rpm, well below the Major League average of 1,751 rpm for changes. Normally, that's a recipe for plenty of grounders, but his 45.5 percent grounder rate on the change is both down from last year and below his career averages.
The difference is that at Lincecum's peak, he was regularly seeing a 10 mph difference between his fastball and his change, and that's now down to only 6 mph, not enough to throw hitters off balance. Unlike many pitchers, he's not gaining or losing anything with perceived velocity, either, even though his average fastball extension of 6.57 mph is better than the average of 6.12. What you see is what you get.
Still, Lincecum did get off to a good start this year, despite his issues. After throwing seven scoreless innings against the Dodgers on May 20, Lincecum's ERA stood at an excellent 2.08. Three consecutive starts of allowing four earned runs apiece (7.04 ERA in that span) have since pushed that ERA over 3, but there's something very interesting to be found in those two groups of starts.
First eight starts: 18.7 percent strikeout rate, 10.9 percent walk rate
Last three starts: 18.6 percent strikeout rate, 10.0 percent walk rate
The results couldn't have been more different, but as far as strikeouts and walks, Lincecum was always the exact same pitcher. (The 8 percent total difference in his strikeout and walk rate are the lowest of his career.) Even his spin rate hasn't changed much; the change, for example, was at 1,426 rpm through those first eight starts.
So what did change? Partially, it's that Lincecum moved away from the sinker (down 9 percent) in order to use his other pitches more, and that helped his ground-ball rate drop by 10 percent. It's a little that he was just allowing harder contact, as Statcast™ shows us that his batted-ball velocity went up by 3.18 mph. But mostly it's that in that first group of eight starts, Lincecum had allowed only a single homer in 47 2/3 innings, which was never, ever sustainable. In 15 1/3 innings since, he's allowed five homers.
Now, that's not likely to continue either, because no big league starter lasts with a rate like that. Instead, what's likely is that it settles somewhere in between. Based on xFIP -- that's Fielding Independent Pitching assuming a league-average home run to fly ball rate -- Lincecum's ERA would be 4.33 right now. Last year, his actual ERA was 4.74. The year before, it was 4.37. The respected Steamer projection system pegs Lincecum for 4.28 the rest of the way.
That sounds just about right, actually. It's about what Lincecum has done over the last two years, and we haven't seen much this year to indicate that his hard work has actually allowed him to overcome the velocity drop. He's still a decent enough back-end starter. It's just not the Lincecum that San Francisco fans hoped they'd see again.