25 July 2016

Should Brach Have Been An All-Star?

Brad Brach has had some of the best results of any reliever so far this season. As of July 5th, he had an ERA below 1, with opposing batters having just a .462 OPS against him. Even now, he has an ERA of 1.17 while opposing batters have a .495 OPS against him. In return, he was rewarded by being selected to pitch in the All-Star game. One note, the data in this post in from early July unless mentioned otherwise. Sorry, life got busy and it doesn’t seem like things have changed much.

Jeff Long, a writer for Baseball Prospectus, defended this decision by noting that Brach has been the best relief pitcher in baseball this season. Long noted that while Brach can’t boast the best FIP, Brach has the best RE/24 for a reliever in the majors.  RE/24 is context-dependent and measures the number of runs that a pitcher allowed or prevented in a given inning compared to an average pitcher. Meanwhile, Keith Law wrote that the Brach silliness reminds him of Neshek in 2014 who had a .70 ERA in the first half, a 3.41 ERA in the second half and has had a strong ERA but a poor FIP for the Astros the past two years. He also wrote that Brach is a dime for dozen pitcher who was an AAA arm 18 months ago. Law is clearly using an FIP based approach to judge Brach. Brach has been largely worthless according to fWAR before 2015. Long is using a context-dependent approach to judge Brach and believes that Brach has control over balls put into play. Should Brach be considered an all-star?

Brach really put it together in 2016. In 2014, Brach had ranked 66th out of 203 relievers in wOBA on pitches put into play, but 141st in wOBA on at bats ending in a walk, strikeout or hit by pitch with an ERA ranking 103rd. Brach may have had a 3.13 ERA, but that was still only about average. In 2015, Brach ranked 78th out of 218 relievers in wOBA on pitches put into play, but 129th out of 218 relievers in wOBA on at bats ending in a walk, strikeout or hit by pitch with an ERA ranking 65th. This year, Brach ranks 8th out of 135 relievers in wOBA on pitchers put into play and 24th in wOBA on at bats ending in a walk, strikeout or hit by pitch and 4th in ERA. He’s gone from average, to above average, to elite.

One reason he became successful starting in 2015 is because he improved significantly against left handed batters. His K% increased from 17.3% to 22% from 2014 to 2015 while opposing batters had only a .239 BABIP and .259 wOBA against him. He continued his success in 2016 by reducing his walk rate from 11.5% to 7.5% while increasing his strikeout rate from 22% to 24%. His wOBA against lefties actually increased from 2015 to 2016, due to him allowing a more sustainable .737 OPS when lefties put a pitch into play compared to a .594 OPS in 2015.

However, the reason why he’s so successful in 2016 is because he’s been absolutely butchering right handed hitting. Right handed batters have only a .164 wOBA against him and he has a K% of 37.8% and a BB% of 7.8%. Opposing right-handed batters have only a .198 wOBA and .104 BABIP when they put a pitch into play.

The reason Brach improved in 2015 is because he started to throw a changeup instead of a splitter. Prior to 2015, he primarily used just a fastball and a slider against right handed batters and a fastball and splitter against left handed batters. With the introduction of the changeup, Brach throws only 60% fastballs to lefties and righties compared to the 70% he threw beforehand. He throws the changeup and slider about the same amount of time against right handed batters, but tends to use a slider early in the count while using the changeup to get strike three. Against lefties, he tends to primarily use a fastball and changeup although he does use a slider 8% of the time.

Brach has thrown 786 splitters and sliders from 2013 to 2016 and picked up 39 strikeouts using those pitches. That’s good for a 5% rate. Brach has also thrown 551 changeups over that same period and picked up 54 strikeouts or roughly a 10% rate. Brooks Baseball tells a similar story. Brach has thrown 511 changeups in 2015 and 2016 resulting in 49 strikeouts or a rate of 9.6%. In contrast, the slider has about a 4.6% strikeout rate. Brach had a tough time before 2015 because he didn’t have an off speed pitch that he could use to get a strikeout. Adding his changeup has largely fixed that problem. Being able to implement a third pitch to use against left handed batters could turn him into a legitimate closing option or perhaps even a potential starter if he has enough stamina.

This year, Brach has taken a major step forward against right handed batters by giving up fewer called balls while collecting a similar amount of called and swinging strikes. He has a called ball rate of 31.8% and a called or swinging strike rate of 30.8%. There’s only a one percent difference between these numbers, and therefore it’s no surprise that he’s collecting a large amount of strikeouts with a relatively low amount of walks. In addition, batters have a 25% foul rate, but only a 12.5% in play rate. In the rare cases where batters do make contact, they’re not hitting the ball fair. In many cases, this means that the batter is one strike closer to striking out. In every case, Brach receives another chance to get a strikeout.

Right handed batters are especially having problems dealing with Brach’s fastball. They’re only putting 9.9% in play, while fouling off 28%, having 29% result in balls and a whopping 32.6% are resulting in either called or swinging strikes. Even better, opposing batters have only put 23 into play and have just a .095 BABIP and a .280 wOBA.

He’s also taken a step forward against left-handed batters. His called ball percentage has dropped from 40% to 36%, while his called plus swinging strike percentage has increased from 25% to 29%. Again, it’s largely because of improvements in his fastball. His fastball had a called ball rate of 39% in 2015, and it’s down to 33% in 2016. In addition, batters are missing at 12% of fastballs this year compared to 5% in 2015. I think the problem that Brach faces against lefties is that he throws his changeup roughly a third of the time, but throws his changeup in the strike zone only 23% of the time. Otherwise, he probably needs a fourth pitch that he can throw early in the count for strikes. But then, relievers typically don’t use four pitches.

Brach has also gotten pretty lucky in 2016. Opposing batters have a .236 wOBA against him with the bases empty and a line of .156/.221/.302. Part of it is due to a 27.9% K-rate and a 7.7% BB rate but a large part is due to a .188 BABIP. Batters have a .192 wOBA against him with one runner on base and a .105 wOBA against him with two or more runners on base. With two or more runners on base, opposing batters have a .077/.133/.077 line – or 4 strikeouts, 1 walk, 1 single and 9 outs in play. As of July 10th, opposing batters have a .099 wOBA against him, and have a .071/.176/.071 line against him – or 4 strikeouts, 2 walks, 1 single and 10 outs in play. It’s highly unlikely that such a line is due to skill.

Adding a changeup to Brach’s repertoire has definitely made him a better pitcher and taken him from average to good. Keith doesn’t give Brach credit for this and therefore significantly underestimates Brach’s’ value. If Brach can add another pitch to use against lefties, then he’d likely be good enough to become a closer. In 2016, there’s no question that Brach has been lucky. He’s gotten extremely good fortune with multiple runners on base. When that luck evens out, his ERA will go up. He isn’t pitching well enough to deserve an ERA before 1.00. Realistically, he’s good enough to be an  setup man. Not bad for a guy picked up in AAA.

In July, Brach is struggling as he "only" has a 1.69 ERA. This is because opposing batters have had an .804 OPS against him with two runners on base or with the bases loaded. He's allowed 1 strikeout, 1 walk, 4 outs in play, a single and a double. He's not getting as lucky with two or more runners on base as he had been previously and therefore it's hurting him.

I feel that Brach should be an all-star. Unlike Keith Law, I tend to think that 2016 performance should be the dominant factor in whether a pitcher is sent to the all-star game. I’m willing to consider previous performance when it comes to tiebreakers, but otherwise I want the best pitcher this year. Even if Brach has been lucky, there’s no question he’s also performed. But I also think that looking at RE24 just gives Brach credit for having an excellent BABIP with runners on base. Is that a skill or luck?

Brach is having a good year and has shown significant growth. If he can develop another pitch to use against lefties, he can legitimately become an elite reliever.


Roger said...

Good article. I would add that this year's increasing responsibility due to the O'Day injury also counts toward being an All Star, although I don't think it's a measurable element.

Jon Shepherd said...

The metric you are looking for is called leverage index.

Matt Perez said...

It is true that the metric you're looking for is leverage index. But realistically, humans are pretty good at determining which relievers pitched in the most crucial situations. The voters knew that Brach has been important this year.