A Few Pieces of Advice for Ned Yost
Tonight, the Royals and A’s play a single-game elimination to determine the winner of the American League Wild Card, with the winner going on to face the Angels starting on Thursday. Both teams have their best starters on the mound, and Jon Lester versus James Shields is about as good a match-up as you can hope for in a winner-take-all contest. With these two starters, strong defenses, and a pitcher’s park as the venue, we shouldn’t expect a ton of runs to be scored in tonight’s game.
And so the managers for each squad are likely going to feel the pressure to try and steal a run here or there, knowing that in a low run environment, every little advantage could turn out to be the difference between advancement or the end of the team’s season. So, against that background, let’s offer Ned Yost some friendly pieces of advice.
Don’t Ask James Shields To Go Too Deep
Yes, his nickname is “Big Game”, but that’s mostly because his first name happens to rhyme with game, not because he has a stellar track record of performances in the postseason. In fact, Shields has been actually quite mediocre in the postseaosn, running a 104/101 FIP-/xFIP- along with a .327 BABIP that pushes his ERA- to 123. The sample is tiny and the numbers aren’t useful in any real way, but let’s at least understand that the origin of the nickname isn’t based on Shields having any kind of great run of success in October.
Of course, he’s still a very good pitcher, and one of his hallmarks is going deep into games; he ranked 4th in the majors in total batters faced this year, averaging 28 batters faced per start. But tonight is an elimination game, and James Shields shouldn’t face anywhere close to 28 batters. He maybe shouldn’t face more than 18.
The negative effects on a pitcher facing a hitter multiple times throughout the same game have been well documented, and the effect is both real and substantial. Here are the average lines against a pitcher this year the first, second, and third time they face a starter within the same game:
| PA | BA | OBP | SLG | OPS |
|---|---|---|---|---|
| 1st | 0.246 | 0.304 | 0.377 | 0.681 |
| 2nd | 0.256 | 0.313 | 0.395 | 0.708 |
| 3rd | 0.268 | 0.327 | 0.421 | 0.748 |
The gap between a pitcher’s average performance the first time they face a hitter and the third time they face a hitter is 22 points of batting average, 23 points of on base percentage, and 48 points of slugging percentage. This is a big difference, and the times through the order penalty means that by the time a pitcher has gone through the line-up twice, he probably should be expected to be less effective than almost any reliever in the bullpen, even a very good pitcher and a pretty mediocre reliever.
And the Royals have a lot of relievers who don’t qualify as mediocre. We think of James Shields as an excellent starter because he’s held batters to a career .313 wOBA against him, and he’s right in line with that at .310 this year. Here are the career wOBAs as allowed out of the bullpen by the Royals primary relievers:
Aaron Crow: .310
Jason Frasor: .301
Kelvin Herrera: .282
Wade Davis: .221
Greg Holland: .249
On a per-batter-faced basis, each of these relievers have performed as well or better than Shields; he’s essentially equivalent in offensive production allowed to a guy who is widely considered to be a busted prospect and the team’s fifth best reliever. And that’s Shields’ overall performance, which includes the batters he faces the first time through the order. By the time he’s facing a hitter the third time, his expected wOBA against is closer to .320 or .325; worse than just about any pitcher the Royals might bring out of the bullpen tonight.
The Royals biggest advantage over other teams is their relief corps, and they’ve done a good enough job building depth that they don’t need to ask Shields to pitch through the order more than twice tonight. Given an expected .300 on base percentage against him, that would put him in line to get something like 12 or 13 outs, leaving the remaining 14 or 15 for the bullpen.
That’s three outs apiece for Frasor and Herrera to get the team through the sixth inning, and then Davis and Holland can split the final three innings between them, with Crow and Scott Downs or Francisco Bueno around for situational match-up work if it’s needed. The team had a day off yesterday and will have a day off tomorrow, so there’s no reason to hold back on the bullpen usage. Even with “Big Game James” on the mound, the Royals best chance to win is to get him out of the game after 18 batters faced.
Don’t Give Away Too Many Outs On Purpose
The Royals surprisingly haven’t actually been too bunt happy this year, as their 33 sacrifice bunts rank just seventh in the American League. However, Yost got a little too cute towards the end of the year, including some atrocious displays of giving away outs in a late season match-up against the Tigers. In their 3-2 loss to the tigers on September 20th, we saw Yost do the following against Max Scherzer:
1st inning: leadoff double, bunt runner to third.
3rd inning: leadoff single, single, bunt runner to third.
In both cases, the bunter was Nori Aoki, a mediocre hitter who isn’t necessarily expected to drive guys in from second base. However, in both situations, Scherzer buckled down and dominated the next few hitters, inducing three strikesouts and an infield fly, with the bunts leading to scoreless innings in both cases.
Of course a sample of two bunts doesn’t tell us anything, and the results easily could have gone the other way, but Yost mostly did a good job of not falling into the small-ball trap this year, and he shouldn’t get suckered into it just because he thinks low scoring games against good pitchers are the time to start giving away outs. Bunts don’t have to be universally terrible, especially if they are unexpected bunts or attempts to get on base, but Jon Lester is a good enough pitcher that he doesn’t need free outs. Avoid giving him free outs, Ned. Late in the game, playing for one run might very well be the right call, but don’t have your #2 hitter give himself up in the first inning of a scoreless game. It doesn’t help.
Be Very Aggressive Stealing Bases
Instead of bunting to move guys into scoring position, consider stealing bases instead. Your team led the majors in steals this year, swiping 153 bags at an 81% success rate, also best in the majors. Your team is fast and full of guys who know how to run the bases. That’s a big advantage, especially against the A’s, who allowed 100 stolen bases and only threw out 22% of opposing runners, the fourth worst mark in baseball.
And Jon Lester, even as a left-hander, isn’t great at shutting down the running game. For his career, he’s allowed a 70% success rate on stolen base attempts, and that mark was 75% this year. Especially if Derek Norris (career 78% SB%) ends up behind the plate for tonight, the Royals should be extremely aggressive on the bases. Geovany Soto (career 73% SB%) is a little more of a deterrent to the running game, but even if the A’s go with Soto, the Royals should still force the A’s to try and throw them out.
This is the kind of small ball the Royals are good at and the A’s are bad at defending, and this is the kind of scenario where putting the game in motion could actually be a real benefit.
Get to the bullpen early, don’t bunt before the seventh inning of a close game, and give your runners the green light. If the Royals want to advance to the division series, that’s their best shot of getting there.
Dave is the Managing Editor of FanGraphs.


I just put this out on Twitter as well, but per your last point about being aggressive stealing bases, Jon Lester has not attempted a pickoff even once this year. He seems to have completely moved away from pickoff throws as a strategy in the last three seasons.
Pickoff attempts by season:
2009: 79
2010: 98
2011: 70
2012: 5
2013: 7
2014: 0
Source? Because wow, if true.
http://www.sportingcharts.com/mlb/stats/pitching-pickoffs-leaders/2014/
Great find, Jon. That’s such an unlikely happenstance I had to forward the results to a friends who was a semi-pro player before turning to journalism.
Nice if we can dig up numbers too on attempted/successful SB year-by-year too, and if the note below about problems throwing is correct, then throwing errors too.
Zero attempts? And for a LHP? How is that even possible?
It’s all a big ploy. He’s saving the element of surprise for a playoff game to get that out when he needs it. /s
Lester has a problem throwing to bases, his typical pickoff attempt ends up with the runner on 3rd. Same thing if he has to field, you just have to hope he’s close enough to the base to underhand it over.
Jon that is some good info
thanks
I think it’s important to have things you enjoy. Maybe he just has more fun throwing the ball to his teammate behind the plate.
Would his change in strategy reflect the change in catchers?
I’m unsure, but I think varitek was catching Lester until 2012, and varitek was known to have a noodle arm.
From 2009 to 2011, opposing runners stole 215 bases, and Varitek threw out 37 (14% RTO, I think).
“… but Yost mostly did a good job of falling into the small-ball trap this year, …”
Should this be “… good job of “not” falling into…?”
Well if you’re going to fall into a trap, might as well be one with small balls, rather than big balls?
We’ve got big balls, we’ve got big balls. Such big balls. Dirty big balls.
While I agree with the premise of your article, I don’t see any way that Yost deviates from his regular season regimen and will not go to Davis before the 8th or Holland before the 9th. If Shields gives up 2 or 3 earned runs, Yost will let him work through the 6th inning.
This was correct until Crow blew a 6th inning lead against the Red Sox, and “only one out until we could get to Herrera”. After that, Yost used Herrera and Davis early on multiple occasions over the last two weeks.
They have quite a nice piece in Brandon Finnegan now. Crow should only be working blowouts at this point.
is SB% off a pitcher that good of a stat? It doesn’t take into account the runners who don’t even attempt to steal on a pitcher with a good move or quick to the plate. It’s like outfield assists. You build their arm into the decision to try for an extra base
Good point.
For that matter, doesn’t SB% off a catcher have the same issue?
You probably wrote this before the rosters were announced, but it should be noted that Aaron Crow, Francisley Bueno (not Francisco), and Scott Downs are not on the wild card roster. The only two lefties are Duffy and Finnegan.
I bet those three are sad that they were left out.
Not bad advice, but I would imagine the Royals quit taking advice from Fangraphs writers around the time they suggested they trade James Shields in July, because, playoff odds!
Unless they get bounced this round. Then whoo hoo for keeping Shields for one game. Royals fans will look back on 2013-2015 as the time they didn’t completely suck and had that one playoff game…
Now can they revel in their decision to keep “Big Game” James?
Get to the bullpen early, don’t bunt before the seventh inning of a close game, and give your runners the green light.
As a Tigers fan who has to watch Ned Yost manage 18 times a year, this advice will make Ned Yosts head explode with confusion
Well done, the twins are hiring, interested ?
would the Royals be better of going with relievers for the first few innings to try to get some platoon advantages or force the A’s to sub, then put in Shields, then close it out with a lefty and righty?
I know fangraphs wrote about it last year, but do you think a team would shy away from it, thinking it’s “cheap”?
Frank Firke advocated a similar strategy against the A’s
http://www.hardballtimes.com/playing-the-as-grow-a-beard./
Another note about those Aoki bunts: Leading up to that 9/20 game he was 13-16 in his previous 4 games. This is the kind of small sample that might justifiably convince you that a guy is seeing the ball well. All that aside, Aoki is both difficult to double up (5 GIDP on the season), and he’s very difficult to strike out, which is how Scherzer would normally get out of a jam.
I was overjoyed by these bunts. As a Tigers fan, in moments of frustration I sometimes think forget it, my team doesn’t deserve to go to the playoffs if this is all they can do with $160 million. When I saw those bunts, however, I could only imagine how annoyed I’d be if I were a diehard Royals fan. If you do that you do not deserve to win the game, period.
In Yost’s defense, I’m told that that first bunt was Aoki’s idea. Okay fine, but Yost went ahead and called for the bunt in the 3rd inning, same situation.
Of course a sample of two bunts doesn’t tell us anything, and the results easily could have gone the other way…
I disagree. The sample tells us that the manager is probably clueless. Also, the fact that they could have gone either way has very little or nothing to do with whether or not the play makes sense. Playing the lottery could work out, too. Actually, that’s a bad example because playing the lottery probably makes more sense than Aoki bunting in that situation, since outs are worth a lot more than a dollar or whatever it is.
It’s misleading to look at Crow’s career numbers since he has completely gone off a cliff this season. He was even demoted to AA a month ago, and that was before he gave up a game winning grand slam to the Red Sox.
Fortunately, the Royals have at least five guys better than him. Holland, Davis, Herrera, Frasor, and Finnegan.
As for Scott Downs, the closest that he should get to this game is the parking log.
Shields does better than the average pitcher against a lineup multiple times through. Just looked at baseball reference and found these splits for time facing an opponent in a game (maybe not the same split as used here, but it is what I found quickly):
1st PA: 258/305/405/710 (2555 career PA)
2nd PA: 257/308/416/723 (2540 career PA)
3rd PA: 256/301/425/726 (2356 career PA)
Maybe aggressively using the bullpen is still a good idea, but Shields loses less value than average when facing a lineup a second and third time.
Accidental downvote, sorry! Good post.
Novel idea…using actual stats from the player instead of generic league wide stats to make your point. Seems like the type of thing you’d expect from the Fangraphs editor vs a commenter, right?
“James” and “game” don’t rhyme. I’m just sayin’.
That’s precisely the reason you yank him after 4-5 innings.
Wow, you’re such a negative Nancy.
Really it’s just a rip-off from James Worthy, anyways.
You are being way to extreme with the bullpen usage.
Actually the numbers support the strategy. What is your rational rebuttal? Just because teams don’t utilize the strategy does not mean it does not work. As good as Davis and Holland have been, I would not hesitate to give them 10 of the 27 outs and play the matchups around Herrera in the middle innings.
Considering that both Holland and Davis have battled dead arms this year and considering both have been extremely effective starting innings and going flat out through three outs, changing now makes no sense. They’ld have to dial back the first inning, then get warm and loose for a second. Herrera can do that, so if a fireman is needed in the 6th he’s the one, but he has been know to blow up when going back out. Shields for at least 6 if he’s got it, then the usual finish.
If Dave’s idea was correct Billy Beane would have been using it for ten years.
play it as you see it?
“Back in my day pitchers used to throw all 9 innings in both games in a doubleheader. Why are pitchers now so terrible?” -Jose
In a single elimination game, way too extreme is the perfect choice. You pick a spot for Shields to get through, and then the next hit he gives up he’s gone or if he gets through that inning he’s gone. It’s the smart move.
While I’m all for aggressive bullpen usage in an elimination game, I do think Dave can get a little carried away. A couple of years ago, he was calling for Showalter to pull Joe Saunders after four against Texas in the WC game. And I’m not talking four innings. Dave wanted him pulled, no matter what, after four BATTERS. Even a guy like Saunders, who is LOOGY with stamina masquerading as a starting pitcher, deserved a long leash than that. Even mediocre pitchers can shut down a great lineup on their best days.
This bullpen strategy is vastly under-utilized across MLB…there are complex interpersonal reasons why it would be resisted, but with today’s expanded bullpen, it would be a huge boon to a team’s run prevention, and by extension winning percentage! #marketinefficiency
re:complex interpersonal reasons …
If some big-ego pitcher gets his panties in a bunch because his usage in a do-or-die game doesn’t match up with his normal pattern or his image of his own worth, then he should spend his unwanted time on the bench considering real estate agents.
On bunts, one of the biggest factors is positioning of the infield. If 3B is back a bunt is a fine idea for a left handed hitter with a little speed. If 3B is in try to slap it by him. Aoki’s bunt hit percentage is 26.4 this year, so not an automatic giveaway of an out.
“Yes, his nickname is “Big Game”, but that’s mostly because his first name happens to rhyme with game”
Does Jeff Sullivan know that you stole his witticism?
“Big Yields” Shields?
Of course these numbers don’t really mean anything and the samples are way too small, but I’m going to use them anyway bc … why?
because it fits his long standing narrative?
As explained, the nickname Big Game implies some sort of success in big games which the facts dispute. Other than exposing the misnomer, the little else should be read into the sample size.
Dear Ned:
Please be certain to put your relief pitchers in by situation, and not by inning.
Sincerely,
Statistics and logic
This was a very polite letter.
And here comes another nonsense report from a writer/blogger/whatever of how he knows what to do instead of an experienced manager. Yost has been in the game for years, has seen in all. All. ALL. Yost has guided the Royals to their first post season dance in 29 years. He knows what’s best.
He puts relief pitchers in based on what inning it is and you dare write something like that? LOL
#WINNING
I regret to inform you that your computer or device appears to be infected with some sort of website redirect malware. Please click on this link to reach your desired web destination:
http://bleacherreport.com/
And in the 6th, Yost left Shields in the game too long, went with someone other than 1 of his best relievers, and the A’s blew the game apart.
But, yeah, Dave’s dumb because all he does is write for Fangraphs.
#WINNING, 2.0 Bro
You’re giving me a bad name. ME. That ain’t easy, son.
Aoki bunted on his own in the Sept 20 game against the Tigers as he has done a couple times this year.
The first time that is.
I’m theorizing but he probably earned the nickname “Big Game” after the 2008 postseason when he had a 2.88 ERA in 25 innings. It’s probably easier to just leave his nickname the same than to say to him “hmm well you haven’t done well in the postseason since then, so now we’re going to call you ‘Slappy'”.
The TBS broadcasters were calling him that at the start of the 2008 ALCS, when he had pitched in exactly one playoff game in his career (6.1 IP, 3 ER). I remember clearly being bewildered by the name because the Rays had played about 5 big games in their franchise history at the time.
It was always about the sound, not the results.
What a prophetic piece. This sixth inning is so horrifying to watch. Can’t use one of your good relievers in the sixth. Just can’t. God Yost is a dumbass.
I thought of this article as soon as Yordano stepped out of the pen, excellent stuff
And now Herrera is in, in the 6th inning down 2!
ho ho ho Merry Yostmas
GOOD THING HE GOT SHIELDS OUT
Ctrl+F: “Yordano”. No results. Yost must’ve TLDNR’d this.
Cool internet speak big boy, I have no idea what it means though
More likely to manage the Royals next year:
A) Ned Yost
B) Ron Gardenhire
Dave absolutely nailed this, just needs to predict the extra innings next time somehow
They bunted a lot. Also, they got lucky with Soto being injured as he handles the running game much better than Norris
Uh, except for the low scoring game part. Steal bases? Yep, I agree. Low scoring? no way.
Only 3 teams in the modern era have given up 7 SBs in a post-season game – & two were the A’s!
Norris is now linked forever to Shrek — rather, Ossee Shrecongost, a.k.a. “Shreck”, catcher for Philadelphia, 3rd game, 1905 World Series, won by the NY Giants 4 to 1. The winning score was 9:0. All 5 games were shut-outs; Christy Mathewson won 3.
Marking my calendar for 2123: will the A’s repeat?
Yordano? Really? In that situation? Wow.
I, for one, am a BIG Ned Yost fan. It makes it seem as though any of us could possibly be a Major League manager.
Even if you concede the points on bunts, Shields’ effectiveness the third time through, etc. etc., there’s just no way that anyone with a brain would even think to use Yordano there. None. Pedro’s post game criticism was spot on.
If this is Yost’s usual mo, one can only imagine where the Royals would have finished the season with an actual manager in that spot. I’m not talking another Major League manager, or even another professional manager. I’m talking the guy who manages my softball team. He’d have won 100 games with this bunch, if Yost can get them to a wild card game, which he mistakenly won despite his best efforts not to.
Can we get Lou Brown?… Joe Riggins?… Jimmy Dugan?…………UCHIYAMA????
Don’t tell me what to do Mr. Column Writer. How many play off games have *YOU* won?
I remember your last piece of advice for the Royals–trade James Shields cause they weren’t making the playoffs lol
I can’t believe the Royals won DESPITE getting so badly Yosted.
1: putting in Ventura (trying to lose the game?)
2: double steal (I’m surprised Billy Butler didn’t pull a hammy)
3: Jason Nix (WHY IS HE ON THE ROSTER)
4: and all the small ball. I get some of it – but why have Dyson bunting over a base stealer when he could steal OR when Dyson isn’t all that likely to be doubled off? Or having Aioki bunt when he’s been on fire swinging the bat. He recently broke the all time Royals record for hits in a series. HOT. He had a great sac fly at bat last night.