Stuff

We saw earlier that Nelson took the cake in terms of velocity, while Pena’s movement paced the team. On the other end of the spectrum, there was no such divide — one pitch had both the lowest velocity z-score and the lowest total movement z-score:

For a curveball, below-average vertical movement isn’t necessarily a bad thing, since the average such pitch already moves negative. Indeed, the fact that Fiers led all starters in curveball drop last season would seem to work in his favor.

Of course, all the dive in the world won’t count for anything if the pitch has no velocity. Only Jered Weaver, Mark Buehrle, and Julio Teheran threw their curveballs slower in 2015, and it showed. Fiers’s curve went for strikes just 43.0 percent of the time in Milwaukee, and when he put it in the zone, hitters such as Tyler Moore made him pay:

View post on imgur.com

Those kinds of dingers made Fiers’s curve worth -1.61 runs below average on a rate basis. The Brewers will miss Fiers overall in 2016, but I for one am glad that his curveball will stay with the Astros.

Run Values

Interestingly enough, though, Fiers’s curve didn’t finish last when it came to Linear Weights. Another starter, who remains on the team, held that distinction:

Peralta’s 2015 regression fell squarely on the shoulders of his sinker, which traveled nearly a mile and a half slower than it did in 2014. That meant the pitch often resulted in this sort of treatment, courtesy here of Curtis Granderson:

View post on imgur.com

But a run value alone doesn’t really do Peralta justice. For his case, we’ll look at something else.

TAv

Using the same samples from the velocity and whiff rate z-scores, I found the average and standard deviation for each pitch’s resulting TAv. These created TAv z-scores, to better reflect how hard batters hit an offering. They certainly clobbered Peralta’s heater:

How bad was this? The next-worst pitch, Tyler Cravy’s four-seamer, had a TAv 1.75 standard deviations over the mean. Only Vidal Nuno and Sean O’Sullivan allowed opponents to abuse them more than this. Let’s throw in another GIF of this atrocity, for good measure:

View post on imgur.com

I discussed Peralta’s meltdown in December, noting that the movement of his pitches actually improved in 2015. Let’s hope for his sake that he regains his velocity, because if he doesn’t, the 2016 campaign will bring more of the same kind of pain.

Whiffs

When it comes to swinging strikes, things get complicated. Unlike Fiers’s curveball and Peralta’s four-seamer, this pitch actually held its own in 2015:

Blazek didn’t earn many whiffs with the curveball, yet it still gave him 1.99 runs above an average pitch (per 100 appearances). That’s because it accrued its strikes a different way:

View post on imgur.com

Matt Carpenter and his fellow batters took the curveball for a called strike 30.7 percent of the time, an incredible amount. An exercise like this has limits, as Blazek demonstrates — no one metric can really capture all the value of a pitch.

That isn’t to say, however, that Peralta’s four-seamer didn’t struggle, or that Fiers performed well with his curveball. We can pretty conclusively deem one of those two the worst Brewers pitch of 2015. Hopefully, 2016 will bring more pitches like the ones we saw on Thursday (although, if the team continues this rebuild, we won’t witness great pitching for a few years).

Click here for a complete spreadsheet of all 39 pitches.

Beyond that, pitchers make a better target for sabermetricians, as we can quantify so many elements of their game. Thanks to PITCHf/x, we can now look at pitch usage, velocity, movement, release points, locations — and, most importantly, individual result breakdowns for each offering. This can lead us down any number of analytic rabbit holes, which is where our story begins.

Because we still have a few weeks until baseball returns, I’ve decided to pass the time by constructing arbitrary “best-of” lists. Most of them wouldn’t appeal to any sane person, but I feel that many fans of the Brew Crew would like to know the answer to this question. Of the many pitches we had the pleasure of viewing last season, which stood out above the rest? A question that broad doesn’t have one right answer, but it’s an entertaining exercise regardless.

To accomplish it, I looked at pitches that appeared at least 200 times, isolating myself to those that occurred when the player was in Milwaukee. (In other words, this won’t include Mike Fiers’s time in Houston or Jonathan Broxton’s work for St. Louis.) This gave me a sample of 39 pitches to work with:

Whose quiver contained the deadliest arrow? Well, we can look at the issue a few different ways, each of which has its own merit. I’ll run through them all, with fun GIFs and full explanations, then return with a final summary.

Stuff

When evaluating the quality of a pitcher, we can focus, broadly speaking, on two things: the process (think cFIP) or the results (think DRA). The same general logic applies to the pitches themselves. We’ll begin with a few metrics that will tell us how well the pitches theoretically should have performed in 2015; from there, we’ll then move to the measures of how well they actually performed.

For a pitch to blow away the opponent, it generally has to have either velocity or movement. We’ll thus begin our journey with these two categories. For this, I used the BP PITCHf/x leaderboards to find pitchers (separating starters and relievers) with 200 of each pitch type in 2015. I then found the average and standard deviations of each sample, from which I constructed velocity, horizontal movement, and vertical movement z-scores for all 39 offerings.

The pitch with the most power behind it might seem familiar:

Nelson’s curveball, which ranked above Sonny Gray and Jacob deGrom in terms of velocity, left hitters such as Aramis Ramirez guessing:

View post on imgur.com

For movement, I took the sum of vertical and horizontal z-scores, to get a rough measure of overall bite. Although the top offering here might not be your first guess, it certainly deserved its spot:

Pena’s four-seam fastball didn’t have much heat, but man, could it dance. Watch how it rises and tails away from Kyle Schwarber:

View post on imgur.com

Nelson saw more action than Pena did in 2015, which may explain why his curveball feels more familiar. Still, each of these offerings stood out in its own regard. As the Brewers experiment with their rotation in 2016, we’ll probably see a lot more of both the heater and the curve.

Run Values

With that said, velocity and movement alone don’t make a pitch. A better line of thinking would look at the actual production of a pitch, and for that, we have a unique metric.

Pitch Type Linear Weights measure the count changes by each pitch, as well as the result when the offering ends a plate appearance, and expresses it as runs above or below average. This metric generally achieves its goal, and its selection for top Brewers pitch (on a per-100 pitch scale) probably wouldn’t get too much blowback:

Rodriguez won this contest by a wide margin — Michael Blazek’s curveball came in second, at 1.99 runs above average. Darin Ruf would probably agree with that verdict:

View post on imgur.com

This shouldn’t come as a surprise, since Rodriguez’s cambio made his 2015 resurrection possible. In fact, the changeup was worth more (by this metric) on a rate basis than any other pitch, of any kind, in all of baseball. Taking note of that a few months back, Mark Simon crowned it the best pitch in the majors. Sadly, K-Rod will pitch in Detroit this season, so this changeup will no longer work in Milwaukee’s favor.

Whiff rate

But we shouldn’t stop there, because Linear Weights aren’t perfect. Aside from the fact that they don’t necessarily reflect true talent (which we’ll discuss in due time), they treat all changes of count the same way. In the eyes of Linear Weights, a swinging strike to begin an at-bat is the same as a foul ball, since each puts the pitcher ahead 0-1. That doesn’t testify to the quality of the offering, though — whiffs are clearly better than fouls, meaning the best pitches will usually maximize the former.

For that reason, we’ll move to swinging-strike rate. Here, it’s important to take into account the different baselines, as each pitch fools hitters to different extents. To level the playing field, I created some more z-scores, using the averages and standard deviations of the whiff rates from the aforementioned velocity samples. As with the Linear Weights, the winner here won’t shock anyone:

Here, too, the leader dominated the competition: The runner-up — Neal Cott’s cutter and its 15.7 percent whiff rate— only topped the mean by 1.24 standard deviations. Based on this hilarious A.J. Pierzynski swing, I can’t argue with that:

View post on imgur.com

I covered Smith’s transcendent slider back in June, and although Smith himself faded a bit down the stretch, this pitch didn’t miss a beat. Unlike Rodriguez, Smith should stick around, so this glorious breaking ball will continue dominating for the Brewers.

In the end, Nelson’s curveball, Pena’s four-seamer, Rodriguez’s changeup, and Smith’s slider disrupted plenty of hitters last year. These standout offerings gave us something to look forward to in an otherwise dismal campaign. (On that note: Later in the week, I’ll use this methodology to find the worst Brewers pitches of the 2015 season. Until then, we’ll simply have these masterpieces to keep us warm.)

The Brewers continue an impressive, albeit unsexy, reload. Parra was a cheap grab at last year’s trade deadline, and while the team went 66-90 during his tenure, he can hardly be faulted. Parra batted .312/.355/.482 in 146 games in a Brewers uniform, and again, that versatility in the outfield was as valuable as the actual quality of his play out there.

Now Parra has done the most valuable thing yet for the Brewers, though, by bringing more in return than he cost to acquire. Davies is no future star, but he could pitch in the big leagues this season, and as Doug Melvin continues spinning off his spare parts, he can look ahead to a rebuild free of some of the limitations he faced when he built this team in the first place.

[To read about the remainder of the Gerardo Parra trade, please visit Baseball Prospectus’ main site and read the entire article, written by Matthew Trueblood and Tucker Blair. The article can be found here.]

Broxton Goes to St. Louis for Malik Collymore

Raw because of spending his amateur days playing in the Great White North, Collymore does show some refinement with the bat for a player with limited experience. He has struggled this year in his first experience off of the Cardinals complex, but he does show the skill set to be a potential above-average hitter, including a short swing path and strong wrists. He’s likely limited to gap power at best, but he does understand how to drive the ball. His bat would profile better at second base where the Cardinals played him last year, but the lack of power could hinder him if he remains in a corner outfield position.

[To read more about Malik Collymore and how Jonathan Broxton fits in St. Louis, visit Baseball Prospectus’ main site and read the entire article, written by R.J. Anderson and Jeff Moore. The article can be found here.]

Jonathan Broxton isn’t too bad. Most fans probably wouldn’t pick up on that, but BP doesn’t exactly cater to the masses. He’s struck out 24.1 percent of the batters he’s faced, and walked 6.2 percent—both of which handily beat the major-league averages for relief pitchers. Those marks, together with a solid serving of ground balls, have given him a 2.94 xFIP, 21 percent better than average. Prefer something a bit more advanced? He owns a 2015 cFIP of 90, better than most other bullpen arms. He’s improved notably since last year, when his mediocre 3.96 xFIP and 99 cFIP suggested he might not have much left. At this rate, though, he could go another few strong seasons.

Which version of Broxton is real? Should we (or the fans of a contending, relief-needy team) expect him to play to the level of his ERA and DRA, or to the level of his xFIP and cFIP? There’s a lot to sort through here, so strap in.

By several metrics, Broxton has taken legitimate steps forward since last season. Hitters have swung at 51.5 percent of his pitches, up from 49.1 percent in 2014, and have made contact on 76.3 percent of those swings, a fall from the previous 78.5 percent clip. Those have combined to give him a 12.2 percent swinging-strike rate, which not only tops last year’s 10.6 percent, but also bests his career mark of 12.0 percent, albeit not by much. Plus, although he’s thrown fewer pitches in the zone (46.5 percent) than he did last season (51.6 percent) or for his major-league tenure (52.2 percent), he’s elicited a swing on 35.5 percent of the pitches that miss, much higher than his 2014 rate of 25.2 percent and his overall rate of 29.7 percent. The product of that: He currently sports a 66.4 percent strike rate, an upgrade from 64.4 percent last year and 65.4 percent overall. Thus, he’s certainly earned his aforementioned strikeouts and walks.

But, of course, those don’t comprise the entirety of Broxton’s story. When hitters have put the ball in play against him, they’ve smashed a .365 BABIP, and a warranted one too: He owns a 35.6 percent line-drive rate and (per FanGraphs) a 31.7 percent hard-hit rate. Additionally, the balls they’ve put in the air have traveled rather far—315.6 feet, on average—which accounts for the spike in home runs. In those areas, he’s never done this poorly; nevertheless, we shouldn’t necessarily brush them off as flukes. Something is amiss here.

Let’s look at his arsenal. Broxton uses three main pitches: a four-seam fastball, a sinker, and a slider. The latter hasn’t slipped at all this year—its current TAv against of .204 syncs up with last year’s .194, and with his career .209 mark. It’s also garnered swings-and-misses (19.0 percent SwStr%) and gone for strikes (62.7 percent Str%) to the same extent that it always has, 2014 included.

The issue, then, lies with his harder offerings. Broxton’s 2015 sinker rate of 18.8 percent eclipses his 8.4 percent all-time mark, and the pitch has done little to justify its place in his arsenal. While its career TAv of .301 suggests that it’s never served him all that well, its present .509 (!) TAv simply won’t stand for a major-league pitch. However, the usual culprits—velocity or movement loss—haven’t played a role:

If anything, the sinker has packed on a bit more power. And, perhaps uncoincidentally, it’s caused about the same amount of whiffs (8.7 percent) and strikes (58.2 percent) that it did before this year. So what gives?

Broxton has always located the sinker off the plate, to the catcher’s left side:

Recently, though, he’s become more predictable in doing that, and in placing it lower in that area of the zone:

Picking up on that tendency, hitters have only swung at the sinkers down and in, and have made better contact (i.e., have had a higher BABIP and ISO) on those swings. Here, the issue doesn’t appear to be stuff, but the usage thereof—which works in Broxton’s favor, since the latter would theoretically represent an easier fix than the former.

The sinker still doesn’t make up most of his pitches, though. That honor falls on the four-seamer’s shoulders, and at one point, it earned it. It’s held hitters to a .239 TAv throughout its existence, riding 96.4 mph of heat and a 9.3-inch drop to greatness. Now that those have fallen to 95.3 and 8.5, respectively, the opposition has upped its game to a .278 TAv. Unlike his secondary pitches, Broxton’s four-seamer simply doesn’t possess the clout of yesterday.

Still, most hitters wouldn’t excel against a pitch that fast—especially one up in the zone, where Broxton’s four-seamer has traditionally resided:

Know what would help them to fight back? The knowledge that the pitcher won’t spread the pitch across the plate; on the contrary, it’ll only come down the middle:

Broxton has kept the fastball high, but it hardly touches the outer reaches of the strike zone. Instead, he’s laid it down the middle—including a stupefying 10.2 percent groove rate—and hitters have pounced. Even with all the heat in the world behind it, a meatball will still come back to bite the man who throws it.

Both the sinker and the slider have fallen from their perches; looking solely at their output would certainly justify any pessimism regarding Broxton’s future. Then again, the stuff remains (for the most part); if Broxton can simply return to the location pattern he once used, the results could come back as well.

There’s one other thing of note here, and it relates to Broxton’s approach. After Broxton allowed the go-ahead home run in a May loss to the Diamondbacks, he and manager Craig Counsell had this to say:

“I’m fine. The ball is coming out fine,” he said. “I’m just getting behind guys I shouldn’t be, especially 1-0 pitches like tonight. It cost me.”

Said Counsell: “It feels like there’s a little bit of a lack of put-away for him. He got behind and you get behind like that, you feel like he should be able to challenge him, get a good fastball and get an out with it.”

Recovering after falling behind has severely plagued Broxton this season. For his career, when he’s gotten a first-pitch strike, batters have a .501 OPS off him; they’ve posted a .517 OPS in such situations during 2015. The problem comes after 1-0 counts: Those have resulted in a .787 OPS against for Broxton all-time, but have led to a staggering 1.212 OPS this year. Together with a 58.1 percent first-pitch strike rate, a slight drop from the 60.0 percent standard he’s established for himself, that has created disaster.

In terms of contextual pitch usage, Broxton has definitely switched something up. Broxton, unsurprisingly, has always used his slider more when ahead in the count:

For whatever reason, he’s now taken that to the extreme, and abandoned the slider when the hitter has the advantage:

Again, his slider rate overall hasn’t budged from prior campaigns—but the scenarios in which he’s implemented them has. Broxton just doesn’t use the slider as a comeback pitch, and its absence has cost him.

Maybe Counsell has told Broxton to feed the hitters fastballs when behind (I don’t base that theory on anything other than Counsell’s quote above, so don’t take it too seriously). The cause for this certainly isn’t obvious, so it might regress away over the second half of the season. Should it remain, Broxton will likely continue to struggle after a first-pitch ball—and, thus, to struggle overall.

Broxton’s right arm has its fair share of mileage on it, and his career doesn’t lack blemishes; despite all that, he only turned 31 in June, so he can presumably play for a few years beyond this one. The Brewers would like to unload his contract onto another club, which would have the option of bringing him back for 2016. In that year, maybe Broxton can rediscover his dominance. While his future is anything but clear, he has more hope for a comeback than you might suspect.

All data as of Wednesday, July 22nd.

Lead photo courtesy of David Kohl-USA TODAY Sports