Ned Yost Is The Worst Manager, Except For All The Other Managers
If you looked at Twitter for even half a second on Sunday, you probably already know that Ned Yost is doing it again. Nearly six years to the day after the Brewers took the nearly unprecedented step of firing a first-place manager in September, at least in part due to some extreme bullpen mismanagement, Yost’s decisions — and bizarre defenses of them — are again being questioned, as the suddenly struggling Royals have lost six of their last nine.
This article isn’t really going to be about Yost’s one decision, but we have to at least explain what happened. On Sunday, with the Royals up 4-3 in the sixth inning and starter Jason Vargas on his way out of the game, Yost brought in Aaron Crow with a man out and two on. Crow walked Yoenis Cespedes to load the bases, struck out Allen Craig, then allowed a grand slam to Daniel Nava to blow the lead and then some. Nava is one of the more extreme platoon bats in baseball — a switch-hitter, he’s got a career 125 wRC+ against righties and merely a 60 against lefties — and he even admitted to being surprised after the game that Yost allowed him to face a righty.
Worse, Yost’s postgame comments defy logic. He chose Crow because he wanted strikeouts, but Crow doesn’t really strike people out, with a K% mark tied for 296th of the 311 pitchers with 50 innings. He found it frustrating that the game was lost before he could bring in Kelvin Herrera, but didn’t actually bring in Herrera because “the sixth inning is Crow’s inning,” whatever that means. Crow’s velocity is way down and he’s having the worst season of his career, yet he was still allowed to face a hitter who had the platoon advantage in the biggest spot of the game, apparently because Yost feared Mike Napoli would pinch hit if he made a move.
And I do mean the biggest spot, of course. The Leverage Index on that home run was 4.95, easily the biggest of the game, nearly five times higher than a regular, normal plate appearance, and also the highest of Crow’s season. Between better righty pitchers and any lefty pitcher, a diminished Crow is arguably the worst possible pitcher that Yost could have had on the mound in that situation, and the team paid dearly for it. This isn’t the first time that Yost has run into complaints about his tactics, and if the Royals end up not beating out Detroit for the AL Central, this is going to be one of many mistakes that critics point to.
But you know this was a mistake. We don’t need to spend a considerable amount of time going over why, because it’s pretty obvious: Put your players in a position to win, and don’t let your worst players lose the game before your best players can win it. Yost didn’t, but he’s not alone in that, because managers do this kind of thing all the time. Just think about how many managers refuse to use their closer in a tie game on the road, preferring to wait for a save situation that never comes. Think back to how angry Craig Kimbrel looked standing out in the Dodger Stadium bullpen in last year’s NLDS as David Carpenter, who had been acquired off waivers the previous winter, gave up a go-ahead homer to Juan Uribe. This kind of thing, and worse, happens all the time.
That being said, I’m somewhat less interested in talking about whether Yost was wrong on Sunday, because clearly he was, and more interested in the following question: Is Yost, for all his infamous decisions, really the worst manager in baseball at this sort of thing?
It’s a harder question to answer than you’d think, because there’s a lot that goes into these decisions, more than the public can ever know. We can’t know if a pitcher has a personal issue that makes him unavailable that everyone would rather not publicize, or if a pitcher is hiding arm pain from the team, or any number of other things. There’s no indication that any of this was the case for Yost and the Royals, but these are real things that happen and can never be accounted for in a stat.
There’s not a one size fits all perfect stat for this. There’s a good-enough stat, which is Leverage Index, specifically pLI, and we’re going to use FIP to measure pitcher performance. Here’s why. pLI is a “player’s average leverage for all game events,” which in this case is more useful than gmLI, which only accounts for leverage when a player enters the game. In the example of Sunday’s Royals mess, it would probably have been defensible had Yost allowed Crow to face Cespedes and Craig before being yanked for Nava, but because Crow had walked Cespedes to load the bases, his pLI for that plate appearance was much higher than it had been when he entered. This needs to be on a hitter-by-hitter basis, which pLI does. For Crow, his pLI for the game was 3.08, because even though the bases-loaded at-bats by Craig and Nava were hugely important, he also had to face (and retire) Will Middlebrooks in what had quickly become a relatively meaningless bases-empty, two out situation.
We’ll use FIP because it’s easily understandable and the best available option for relievers, since I really doubt anyone wants to see this being judged on ERA or saves. This is admittedly imperfect because it doesn’t account for platoon issues that pitchers may have, meaning that a “true FIP” could be different for each plate appearances, but when you go down that rabbit hole, you also have to account for the quality of the hitter and his platoon issues and so on. For a simple overview of “are the best pitchers pitching in the most important situations,” this will do.
Yost ran the Brewers from 2003 until being fired on Sept. 15, 2008, and has managed the Royals since Trey Hillman was fired on May 13, 2010. For the purposes of this exercise, we’ll use 2003-2008 and 2011-2014. The way I’ve gone about this is to set an innings minimum of 20 per season and identified the three pitchers on each of those clubs who had the highest pLI, then noted where their FIP ranks within the bullpen. This is one of those times where I honestly don’t know where the data will lead us. I don’t expect Yost to shock us with hidden genius, but is it possible he’s not a total disaster?
|Valerio de los Santos||1.28||5.36||10|
So what do we see here, other than a fantastic collection of the ghosts of relievers past? (Derrick Turnbow! Ricky Bottalico!) There’s some ugliness, to be sure, in certain spots, starting with his first season, when de los Santos — who was so bad that he was “purchased” by the Phillies, i.e., “given away for absolutely nothing,” in September of that year — was routinely put into high-leverage situations despite a massive homer problem. You can see that he’s relied on some big-name former closers to a fault, notably Bottalico in the final year of his career in 2005 and Gagne in 2008. It’s not like any of those guys had some unexpected run prevention luck either, since their ERAs were also pretty lousy.
In his Kansas City years, you can see that last season, a big issue was over-reliance on Crow, who had already begun his decline, but at least there the argument can be made that Crow had been valuable in each of Yost’s first two seasons. And this year? Sunday’s debacle aside, it’s hard to argue that Yost isn’t generally doing it close to right, at least over the full season. He has three relievers who clearly stand out among the rest, and they’re the only three, even if you drop the innings down to 10, with pLI above 1.00, which is neutral. Perhaps it doesn’t take the brightest manager in the world to know that Holland, Davis and Herrera are more valuable than Francisley Bueno, Louis Coleman and Michael Mariot. For the most part, at least Yost is using them in the biggest spots. The problem appears to be not understanding that the sixth inning on Sunday was the biggest spot.
It’s just not an issue that’s limited to Yost, unfortunately. Back in May, when Yost had to fend off complaints about his refusal to use Holland in tie games on the road, Jonah Keri did some research and came away with this:
Since the start of last year, non-Yost managers have used their closer in the ninth inning of tied road games 19 times out of 307 opportunities, a rate of just 6 percent. Even taking mitigating factors like closer fatigue and closer committees into account, that’s still an astoundingly low number. It’s about half the closer usage rate we’d see if managers simply rolled a die to pick which reliever to throw into the fire.
It takes every manager in baseball making that kind of call to get to a 6 percent usage number. In 2012, Dan Lependorf, then writing for The Hardball Times and now working for the Oakland A’s, attempted to come up with a way to come away with “manager’s WPA” in terms of bullpen decisions. While it’s possible to question the methodology, Yost’s 2012 was ranked as the tenth-best overall. That seems crazy, but maybe it’s not. You can see above that he used three of his four best relievers in the highest-leverage situations, and it’s not at all difficult to find examples of other managers doing the same things that drive Royals fans nuts.
Just a few weeks ago, Don Mattingly allowed Brandon League to face lefty Lyle Overbay rather than calling on J.P. Howell, who was being “saved” since Kenley Jansen and Brian Wilson were unavailable. Overbay doubled in three in a 6-3 Milwaukee victory. There was an endless litany of Charlie Manuel frustrations from deflated Phillies followers. Just say the name “Dusty Baker” around Reds fans and see what the reaction is. We haven’t even made it to sacrifice bunts, either. It’s just a discussion about bullpen choices. They all do these things.
It’s easy to find complaints about every single manager’s bullpen decisions, and maybe that’s the point. Yost, it seems, has some obvious issues with bullpen management. He’s also very likely not really an outlier here, and he’s perhaps just an easier target because he’s got a history as being a manager fired from a first place team in September, and because he recently (and inexplicably) called out his own fans. In the unlikely scenario that Yost were to be fired again before the end of the season, the likely successor may be bench coach and former Mariners manager Don Wakamatsu, who Dave once referred to by saying he had “bullpen management [that] has been curious at best, disastrous at worst.” Is that really better? Even Joe Maddon, generally respected, does things we can’t understand.
If the only true measure of a manager is how many wins he has, and how well his team is playing above what they’re expected to, then Yost might be doing it right… even when he is almost certainly doing it wrong. So, so wrong.
Mike Petriello used to write here, and now he does not. Find him at @mike_petriello or MLB.com.
What is so hard for these people to understand about using your best available reliever in the most critical relief situation?
I wish there was more optimal use too, but you have to remember that it takes more than one hitter for relievers to get ready. In this case, Herrera should have been the one getting up, not Crow, but it’s not like you know it’s going to be bases loaded to put Davis in. Unless you want your best relievers warming up 130 times a year, you need some sort of rigidity.
Lee, you’re making it much, much harder for us to feel intellectually self-righteous here. Watch that, OK?
I think that’s certainly true for most of the season – figuring out how to keep your best relievers healthy and effective is the hardest part of managing, and over a long season it makes sense to err on the side of caution in use.
But when it’s September, and the Royals are fighting to win their division for the first time in ages, and there are September call-ups to absorb some of the lower-leverage innings…it seems like more aggressiveness with getting match-up advantages is called for. Obviously wins and losses count no matter which month they’re in, but in September you have a lot more clarity about the situation the team is in and how critical each one will be.
Were they facing the same Daniel Nava that has a .355 slugging % this year (and that’s after the HR)? And he has a sub-.400 slugging % against righties? I can see why they wouldn’t bring in their best available reliever to face Nava in the 6th inning.
Two things on the Nintendo usage model:
1) Managers don’t have ESP – it’s easy to look at a game chart and find the highest leverage spot, but it’s not always clear without hindsight. Some cases though are obvious, but it’s not always the case. This problem also ties into the next item…
2) As someone else mentioned, pitchers don’t have an instant ready button. Guy on 1st in the 7th? Not so high leverage, and you probably don’t have your relief ace warming. Next guy doubles, and suddenly it is potentially the highest leverage spot in the game. If that next guy GIDP instead of doubling? Not so much leverage. How often do you have a good reliever warming just in case before he is burned out in the pen?
1) No one here is going to criticize you if you put your best guy in spot like the 6th inning with the bases juiced and clinging to a one run lead.
OK, that situation is extreme, but even for more mediocre leverage points, you should still pull the trigger while you can. There might be other high leverage points in a game, but you know this one is happening now.
Look at the Mets-Marlins game today. In the 7th deGrom let the first two batter reach base and the LI was at 3.1. Now I know these teams are out of it, so who cares, let deGrom get some innings. But that’s absolutely a spot you can bring in your best reliever. The 8th inning ended up having the games highest LI at 5.44, but does that even happen if the Mets prevent the Marlins from scoring in the 7th? If you assume the Mets still tack on three runs in the bottom half of the 7th, it would have been 5-3 rather than 5-5 when the bases were loaded with one out in the top 8. It still would have been a high LI, and that’s why you hopefully still have another competent reliever rather than just one guy, or that your best guy could just go two innings.
2) Actually, starting the 7th in a 2 run game the LI like the Mets-Marlins game the LI was already 1.25 (above average). The first batter gets on and its 2.4 (already worthy of a damned good reliever, but maybe not your closer in the 7th). In close games and during the 3rd time through the order (who cares of inning really), I personally am always calling for someone to be ready in the pen. Except for maybe with a true ace on the mound when he’s clearly on his game.
I disagree with #1. If the Braves brought in Kimbrell in the 6th inning against a hitter like Nava and then lost the lead late, the manager would absolutely be criticized. If Trout was up, he can rationalize it. For someone like Nava, the manager would absolutely be criticized if it didn’t work out.
I also disagree with #2. I just looked at the box score, and DeGrom had given up 3 hits, 1 walk, no runs, and struck out 13 batters through 6 innings. The first batter gets on, and you want him out of the game? Even with two hitters come up with OPS’s under .700 (and the third hitter had an OPS under .600). It’s an awful way to manage a bullpen over the course of a season because you’d burn everybody out, and I also don’t know if it’s good bullpen usage in a do-or-die game considering how great DeGrom was pitching.
1) The batter matters a lot less than the situation there. The manager might get criticized in other places, but I would doubt you’d see it here.
2) No you’re not going to burn everyone out. You’re talking about 1 extra inning here and this isn’t something that’s going to happen every game. Maybe more like an extra inning or two a week. And a lot of times the problem isn’t from not bringing in a reliever, its also from bringing in the wrong reliever. So the bullpen innings don’t even change. It just shifts who they go to and when.
Also deGrom is a good pitcher, but not it’s pretty clear from what happened that he wasn’t really capable of maintaining that 13K performance, because he gave up 3 hard hit balls that inning and one of the outs he got was given to him on a bunt. This is why you have a guy ready as your pitcher is going through the order the 3rd time, because the wheels tend to come off and you can lose the game in the 6th if you’re not ready for it.
It’s only clear in retrospect that he was losing it. The inning before, against better batters, he pitched a clean 1-2-3 inning with a pop-up, a ground out, and a strikeout. He entered the inning under 90 pitches.
You seem to be suggesting that you pull the starter when he gives up a lead-off hit after the 6th inning if the score is fairly close. That absolutely impacts bullpen usage.
The other thing to mention is that he didn’t get any deep counts in the 7th inning. They scored two runs within 6 pitches of the inning starting (and the inning was over 4 pitches later). The way he was cruising through six innings, you shouldn’t have your best relief pitcher or two warming up at the start of the seventh with the bottom of the order coming up. Even if you warm him up after the first batter reaches base, he wouldn’t be ready until after they knocked in two runs anyway.
vivalajeter, lead off hit in the 6th or 7th is certainly reason to get some activity in the bullpen if its a 2 run difference or less. The LI against deGrom was 2.4 already after the lead off hit. The next guy comes up, gets a hit. You get someone throwing quickly in the pen for sure. The tying run is now on 1st base with no one out in the 7th inning. What the heck are you waiting for at this point?
I agree, that it happened quickly, and probably would prevent anyone from getting into the game by the 3rd batter, but after a coaching visit to the mound after the double, then a bunt, to get a guy to 3rd. What’s the excuse there? You still have only 1 out with now the possible wining run at 3rd…um….?