How Ken Giles Became a Minor Leaguer

Ken Giles did not have a good night on Tuesday. Called upon in the top of the ninth inning to protect the Astros’ 4-0 lead over the A’s, he served up three straight singles over the course of eight pitches, allowing one run and bringing the tying run to the plate in the form of Matt Olson. On the heels of a visit from pitching coach Brent Strom after back-to-back hits by Mark Canha and Jed Lowrie, manager A.J. Hinch didn’t like what he saw, and after a first-pitch single by Khris Davis, came out to get Giles, who… didn’t like what he saw either. The closer appeared to have some choice words as he handed over the ball.

The A’s wound up tying the game in the ninth, with all three runs charged to Giles’ room, and they went ahead, 5-4, in the top of the 11th. The Astros nonetheless rallied for two runs in the bottom of the inning, scoring the winning run in bizarre fashion, when A’s catcher Jonathan Lucroy made a hash of Alex Bregman’s swinging bunt:

Giles was reportedly not in the clubhouse after the game, and by Wednesday afternoon, he was a Fresno Grizzly. He’d been optioned to the team’s Triple-A affiliate in what general manager Jeff Luhnow called “a baseball decision.”

The GM added:

“I don’t know what he said (on Tuesday). Things are said between players and players and staff, and that’s a clubhouse issue. That’s dealt with separately. This was a baseball decision. Ken has had success in certain situations and he hasn’t been successful in other situations the way he should be.”

Luhnow steered clear of singling out Giles’ displays of anger, which included punching himself in the face after serving up a three-run homer to the Yankees’ Gary Sanchez on May 1. Hinch, however, expressed concern, saying, “He pitches with a lot of emotion and when things are going well, we love it, and when things aren’t going well we have our concerns just because of the volatility of the end of the game type stuff.”

Giles hasn’t been to the minors since he was called up by the Phillies in June 2014, but the 27-year-old righty has pitched erratically this year and hasn’t come close to the lofty expectations that come with being acquired in exchange for five prospects, as he was in December 2015. Though he saved 34 games in the regular season last year, he quickly became the invisible closer during the Astros’ championship run, getting lit up for 10 runs in 7.2 postseason innings. While he netted a save in closing out the Division Series against the Red Sox and another in the ALCS opener against the Yankees, he turned two-run leads into one-run leads both times. After he blew the save and took the loss in ALCS Game Four, he didn’t get another save chance, as Hinch kept the ball in the hands of starters-turned-relievers Charlie Morton, Lance McCullers Jr., and Brad Peacock for multiple innings rather than turn the ball over to Giles — a departure from orthodoxy that played a huge role in the team’s ALCS and World Series wins.

While Giles has more saves (12) than any other Astro this year, has converted each of save opportunity successfully, and has pitched the ninth inning in 27 of his 34 appearances (compared to just five in the eighth), Hinch hasn’t hesitated to use other pitchers to close games as well based upon matchups. Hector Rondon has seven saves and has outpitched Giles in terms of both ERA and FIP (1.62 and 1.78 for Rondon, 4.99 and 2.25 for Giles). Chris Devenski has two saves and has generally been more effective than Giles, as well (2.45 ERA, 2.88 FIP). Brad Peacock (3.12 ERA, 4.12 FIP) has two saves, too.

In the wake of last fall’s diminished role, it seems clear that Hinch doesn’t trust Giles to the same extent that he once did. On April 14, Hinch conceded to the Houston Chronicle that he had not “done a good enough job” of finding work for Giles in non-save situations. Though Giles notched his first save of the year on April 9, it was really only from April 25 through the end of May when he was consistently the ninth-inning go-to guy. His failure to protect an 8-3 lead against the Indians on May 27 — another three-hits-and-out stint — this one including a 17-pitch battle that he lost to José Ramírez — kicked off a five-run rally in a game the Astros lost in 14 innings, and more or less ended Giles’ hold on regular closing duty. Eight of his 12 saves came in that five-week span, but he’s had only three save opportunities since, with Rondon receiving seven and converting five in that span, Devenski blowing his only chance, and Peacock converting a two-inning save.

Despite all of the juggling, the system has worked well for the Astros, if not for Giles himself. Though they were victimized by the A’s twice while losing three out of four games in Houston this week, the Astros’ bullpen has the AL’s second-lowest ERA (2.85) and FIP (2.88), and the team has the league’s third-highest winning percentage (.902, 46-5) in games in which they’ve led after six innings, behind the Red Sox (.946, 53-3) and Yankees (.940, 47-3).

As the Astros’ broadcast pointed out on Tuesday night while Giles dug himself a hole, he’s pitched brilliantly in those save opportunities this year, not allowing a single run (though his only inherited runner in those situations scored). Here’s a FanGraphs-flavored version of the split:

Ken Giles in Save and Non-Save Sitautions, 2018
Situation G IP H BB% K% ERA FIP
Save 12 11.0 7 2.5% 37.5% 0.00 0.68
Non-Save 22 19.2 29 2.2% 18.0% 7.78 3.13
Total 34 30.2 36 2.3% 24.0% 4.99 2.25

Giles has been untouchable in save spots, holding batters to a .179/.200/.179 line, but in all of his other appearances, he’s been beaten at a .333/.348/.517 clip. It’s not uncommon to hear complaints about closers not pitching well in non-save situations, perhaps due to a lack of intensity or focus. Certain closers seem to wind up as targets of the ire of bloggers or sports talk radio spittle-sprayers, complaints that tend to hinge upon selective memory and small sample sizes within a given season. For what it’s worth, in Giles’ five-year career, his numbers are better in non-save situations (2.54 ERA, 2.10 FIP in 132.2 IP) than in save situations (2.92 ERA, 2.43 FIP, 142 IP).

Both Hinch and Luhnow cited Giles’ problems with his slider, which he throws 42.3% of the time, with his four-seam fastball accounting for virtually all of the balance of his pitches. His overall numbers on the pitch look quite healthy: a 47.0% out-of-zone swing rate, 27.6% whiff rate, and .128/.128/.277 line (15 wRC+) on the 47 at-bats that ended with a slider. Those may not be at their peaks within his five-season career, but they’re within the ranges of what he’s done in the past:

Ken Giles’ Slider, 2014-18
Year Pitches O-Swing% SwStr% AVG OBP SLG wRC+ wOBA xwOBA
2014 277 47.1% 24.9% .099 .135 .113 -23 .118 n/a
2015 446 46.3% 25.1% .171 .205 .225 28 .192 .196
2016 529 54.9% 34.0% .094 .141 .188 -1 .158 .131
2017 450 46.5% 23.1% .130 .151 .211 5 .157 .163
2018 196 47.0% 27.6% .128 .128 .277 15 .168 .225
Total 1898 49.2% 27.3% .125 .156 .200 5 .168 .169
wOBA and xwOBA via Baseball Savant. Data for 2014 not included in total calculations for those categories.

The erosion in Giles’ performance stands out a bit more in terms of his relative rankings among qualifiers — pitchers who’ve thrown at least 200 sliders over a full season, 100 this year, and/or 500 over the three-season span — in wRC+, O-Swing% and SwStr%.:

Relative Rankings of Giles’ Slider Among Qualifiers
Year Qualifiers wRC+ O-Swing% SwStr%
2016 202 6 1 1
2017 211 12 22 24
2018 260 37 24 10
Total 195 4 3 2
Number of qualifiers is based on those who threw at least 200 sliders in 2016 and -17, 100 in 2018, and 500 from 2016 to -18.

Across the 2016-18 span, Giles’ slider is in the top four of all three of these measures, but his standing in 2018 isn’t what it once was. In wRC+, he’s fallen from the 97th percentile to the 86th, though the drops in the other category aren’t as steep.

One thing that stands out in that table of raw data above, beyond those insane 2016 numbers, is the elevated slugging percentage and xwOBA against him this year; when his slider has been hit, it’s been hit a bit harder. Where he allowed five homers on sliders in 2016-17, one for every 196 thrown, he’s allowed two this year, one for every 98 thrown — twice the frequency! But still, that’s pretty subtle, and in fact, his Statcast average exit velocity on the pitch is down from last year (82.3 mph vs. 84.9), with the average launch angle unchanged (13 degrees).

It’s actually Giles’ fastball that’s getting mercilessly pummeled (.380/.402/.506, 164 wRC+) this year relative to last (.283/.392/.404, 131 wRC+). His velocity is down slightly (from 98.4 to 97.9), his swinging-strike rate has dropped substantially (from 10.7% to 8.2%), and his average exit velo has soared from 87.8 mph to 93.6 mph, with his xwOBA rising from .344 to .392. He’s throwing the fastball more (57.7%, up from 52.8%), which is probably exacerbating his problems with both pitches.

According to Hinch, Giles’ slider lacks the shape or depth it had previously. Indeed, via Brooks Baseball, here’s a game-by-game graphs of the pitch’s horizontal movement:

In March and April, the pitch averaged 1.27 inches of horizontal movement to the glove side, but that dropped to 0.87 inches in May and -0.2 inches — in other words, more towards the arm side — in June. I asked FanGraphs alum and pitch-shapes expert Eno Sarris (now at The Athletic) to describe what he saw in Giles’ monthly variance in movement. “Crazy weird,” he said, adding, “It looks like he had too much cut early [this season] and then tried to correct it and went too far.”

Meanwhile, here’s the game-by-game graph of the vertical movement of Giles’ slider:

Two years ago, Giles’ slider had an average drop of 0.18 inches — slightly less than expected due to gravity. Last year, that increased to 0.47 and this year to 1.20, so his slider is averaging an inch less drop than in 2016. Couple that with the horizontal inconsistency and, yes, that’s a different shape with less depth.

It’s fair to wonder if Giles’ erratic performance owes anything to the Astros’ unorthodox approach. Maybe he feels the need to be The Man and is lost in other roles. It’s certainly possible that the relationship with Hinch is strained. With two years of club control remaining after this one, he could be an appealing asset to trade come July 31, at least if the Astros can get him pointed back in the right direction. Ultimately, however, the team would probably be better off if he could just rediscover his dominant ninth-inning form.





Brooklyn-based Jay Jaffe is a senior writer for FanGraphs, the author of The Cooperstown Casebook (Thomas Dunne Books, 2017) and the creator of the JAWS (Jaffe WAR Score) metric for Hall of Fame analysis. He founded the Futility Infielder website (2001), was a columnist for Baseball Prospectus (2005-2012) and a contributing writer for Sports Illustrated (2012-2018). He has been a recurring guest on MLB Network and a member of the BBWAA since 2011, and a Hall of Fame voter since 2021. Follow him on Twitter @jay_jaffe... and BlueSky @jayjaffe.bsky.social.

44 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Dmjn53
5 years ago

His peripherals are mostly fine, and you can bet that Houston’s front office is well aware of that. I don’t think this was a baseball move.

Free Clay Zavada
5 years ago
Reply to  Dmjn53

Eh, I’m not so sure. I’d agree with you if not for his postseason performance last year. That puts off a front office quick.

Dmjn53
5 years ago

Small sample size of 7.2 innings. It’s like saying that the Indians are put off by Jose Ramirez because he had 22 bad PA in the NLDS

Free Clay Zavada
5 years ago
Reply to  Dmjn53

Different for pitchers, especially closers. Yu Darvish is a starter, but his world series performance made it essentially a 0% chance he was resigned by the Dodgers.

Jimmy Duganmember
5 years ago

The fact that he was signed by the Cubs made it essentially a 0% chance he was resigned by the Dodgers. You can’t describe the likelihood of an event happening by stating whether or not it actually happened.

Free Clay Zavada
5 years ago
Reply to  Jimmy Dugan

So we can never speculate on the likelihood of a past event occurring? Good to know.

It’s silly to look at this as a black and white issue. We can look to rumors of the Dodgers interest in Darvish in free agency, which was, like the likelihood the Dodgers would sign him, essentially zero.

Jimmy Duganmember
5 years ago

Sure we can, but the fact that you say the chance of the Dodgers signing him was 0%, combined with the fact that he didn’t actually sign with the Dodgers, doesn’t mean that his chance of resigning was actually 0%.

A simple internet search of articles from last winter indicates that, to the best of public knowledge, the Dodgers did indeed have at least some interest in signing him. Erego, the chance wasn’t nil.

For example… Jan 29, 2018 – MLB Trade Rumors
“The fact that [Darvish has] waited this long indicates at least some level of mutual interest [between Darvish and the Dodgers]”

Your argument goes:

Hypothesis – Teams will react to small sample sizes of pitcher performances in different ways than teams will react to small sample sizes of hitter performances

Evidence – The Indians chose to give Jose Ramirez a starting job in 2018 after a poor 2017 playoff performance, but the Dodgers chose not to give Darvish a starting job in 2018 after a poor 2017 playoff performance.

But your evidence is flawed, because regardless of whether or not the Dodgers gave Darvish a starting job in 2018, they didn’t necessarily CHOOSE not to give Darvish a starting job in 2018. This is not the only assumptive flaw btw, just the one I pointed out in my previous comment.

Jimmy Duganmember
5 years ago
Reply to  Jimmy Dugan

To elaborate…
I believe that you were attempting to use the fact that the Dodgers didn’t resign Darvish as evidence that the chances a team in general would want to resign a theoretical pitcher who had performed poorly in the prior postseason would drop more significantly than a comparable hitter. So I read your comment as basically saying.

“If Darvish had performed well, there would have been a 67% chance that the Dodgers would have resigned him, but given his poor performance, that chance dropped to 4%. If, however, Jose Ramirez had been a free agent, his poor performance would have only dropped his odds from 84% to 81%.”

This obviously sounds like something a crazy person would say, but it’s also just a more fleshed out version of the same argument.

oozyalbies1
5 years ago
Reply to  Dmjn53

Except this isn’t a hypothetical. They clearly, actually were put off. They “benched” him from Game 5 of the ALCS through the WS. They implemented a closer committee coming into the season, at least effectively if not nominally. Now they’ve demoted him.

Is it easier to believe that they demoted him for cursing at the manager than that they demoted him based off of one October + half a season’s worth of poor performance?

Joe Joemember
5 years ago
Reply to  Dmjn53

On his peripherals, his Ks are down. He’s kept his BBs down by throwing fastballs down the middle. I would not say his peripherals are good based on his xwOBA on his balls on contact. His fastballs have always been hit hard and throwing them more often and more often down middle is not a recipe for success even of it keeps his FIP, xFIP down.

RonnieDobbs
5 years ago
Reply to  Dmjn53

yeah, its not like that at all. If they benched him in the world series and optioned him to the minors recently, then they would be more similar!

frangipard
5 years ago
Reply to  Dmjn53

You can also bet Houston’s front office has their own proprietary analytics that they based their decision on.

RonnieDobbs
5 years ago
Reply to  Dmjn53

I am kind of thinking that you don’t follow the games much… remember when he lost the job in the playoffs last year? The don’t like him in the role. As he has proven several times, there is more to it than the peripherals. Giles is very prone to blowups and that doesn’t show up in the metrics.