Poor Decision, Take One
The Atlanta Braves beat the Philadelphia Phillies this past Sunday, on national television, to start the entire major league season. For all intents and purposes, Derek Lowe dominated the Phillies en route to a 4-1 win. On Tuesday, sophomore Jair Jurrjens toed the rubber and held the defending champs scoreless throughout his 5.2 innings of work. Yesterday, Javier Vazquez made his Braves debut and pitched very well outside of a 2-run homer surrendered to Raul Ibanez. The bullpen that looked sharp in the first two games of the season fell off a cliff in the game, however, which helped turn a 10-3 deficit in the seventh inning into a 12-11 win for the Phils.
The poor decision belongs to Bobby Cox, for using his ace relievers in the first two games when they were not needed at all and then not being able to use them in a game that merited their services.
In the first game of the season, Lowe was cruising and the Phillies looked beaten at the plate. In fact, the Braves led 4-0 at the time of Lowe’s exit, and the Phillies only scored when Gonzalez came in and gave up an RBI single to Jayson Werth. Other than simply wanting to get his designated closer into a game, there was no reason to go with Gonzalez in the ninth inning. Not to say that Lowe needed to complete the game, which he seemed perfectly capable of doing, but a 4-0 lead is usually pretty safe in the last frame of a game, as shown by The Book.
In the Jurrjens game, the Phillies threatened on several occasions but ultimately failed to score a run. After Jurrjens left in the sixth, Jeff Bennett and Eric O’Flaherty held down fort in the sixth and seventh innings, sustaining the shutout. Then, with a 4-0 lead, Cox used Rafael Soriano in the eighth, and Mike Gonzalez once again in the ninth.
I can understand the usage of Soriano especially given the offday between these games and the fact that he had not pitched in a bit, but did Gonzalez really need to pitch again? A closer does not need to follow these age old usage patterns with leads greater than three runs and he probably should not even appear in a lead of exactly three runs. Are you telling me that Blaine Boyer or Peter Moylan could not have handled that ninth inning?
Yesterday, Cox wanted to get the remaining members of his bullpen yet to appear into the game, which ultimately led to an absolutely horrible loss. Boyer, Moylan, and O’Flaherty combined for 0.1 IP, 3 H, 8 ER, 4 BB, 0 K. Jorge Campillo completed the inning but did not get an out in the eighth, meaning this foursome combined for 1+ IP, 5 H, 9 ER, 6 BB, 0 K over the seventh and eighth innings. Bennett came in and relieved Campillo in the eighth, but you might have noticed that neither Soriano nor Gonzalez got into this game.
I can understand the need to get pitchers work over the course of a season and why that would merit relief aces appearing in 4-0 games, but hopefully this situation teaches the Braves a lesson. Relief aces are used for situations that really require top-of-the-line pitchers to get the job done. Moylan or O’Flaherty likely could have handled end of the game work in the first two meetings, and if they got into trouble then you turn to Raffy or Gonzo. Yesterday, when all of these pitchers did get in trouble, both Raffy and Gonzo were nowhere to be seen. Bobby Cox is a tremendous manager but he dropped the ball here by steadfastly sticking to the idea of a specialized bullpen.
22
I hear what you’re saying, but keep in mind this all happened in the 7th inning. How many times do you bring in your closer to get out of trouble in the 7th inning? That’s what they would have done by bringing in Gonzalez. Who pitches the 8th and 9th in this scenario? Its a bit of a stretch to lay this loss at the feet of Bobby Cox. The fact is that the entire bullpen laid a huge egg in this one.
If the Phillies are threatening like they were, you better get someone like Soriano in there to attempt to limit the damage, especially if he is considered one of your two relief aces. Then you go from there. It isn’t important what inning it was, especially if it was the 7th or later. Sticking to these rules that the closer HAS to appear in the 9th and the setup man HAS to appear in the 8th, hurt here. If you have Soriano come in when the score is 10-7, who knows if everything plays out the same way. Then you can either go with Raffy to start the 8th and go to Gonzalez if he gets in trouble, or go with someone else entirely in the 8th. But the point is that they weren’t really available because of unnecessary appearances the previous two games.
Just because Bobby Cox followed the age old usage patterns doesn’t leave him free of criticism. The right decision in this game was to use a relief ace when he was needed most, which was in an inning when the Phillies were on the verge of scoring 8 runs.
I do like what you say, about bringing in a relief ace to shut down a precarious situation. Especially when you have 2 relievers that could be considered “relief aces”. Have we seen many managers doing this yet? I’m sure a manager would be more likely to bring in their setup-man instead of the closer though. Have we seen closers brought in the game in the 7th inning or so?
If you’re on the ropes like that, you bring in your best reliever. This is why the term “closer” is so absurd. You need a guy who can shut down the other team… hanging on to him in case you happen to have a lead of three or fewer runs in the ninth inning is a joke. I cannot fathom how this closer concept continues to linger.
“Are you telling me that Blaine Boyer or Peter Moylan could not have handled that ninth inning?”
If they couldn’t handle a 7 run lead in the 7th, what makes you think they would have handled a 4 run lead in the 9th?
This is monday morning quarterbacking.
I was about to point out the same exact thing. Three different pitchers couldn’t throw strikes. It’s an anomaly. The blame goes on the relief pitchers for not throwing the ball over the plate. Boyer especially, pitched like he was scared to be the one to lose the game by giving up a big hit.
No it isn’t. I thought it was dumb at the time in those first 2 games to use those relievers. In my eyes it is dumb to use your relief aces in 3-4 run leads and there is no way that you can transpose yesterday’s game like that in saying that they couldn’t handle a 7-run lead so they couldn’t handle a 4-run lead. That sounds more like the reaction of someone saying a player stunk because he had a low batting average, who believes the true talent of these pitchers will give up leads at all times.
The games aren’t interchangeable. For all we know, circumstances could have been very different. And, as I said, if one of Moylan or Boyer or whomever started to struggle, then it becomes justifiable to bring in Gonzo or Soriano.
We need to get past these specialized roles and use the best pitchers when they’re needed. I wouldn’t advocate using Raffy or Gonzo in the 5th or 6th, but if it’s the 7th inning and a crucial game, do whatever you can to stop the bleeding.
Eric, I don’t think that Cox’s decisions in the first two games were as clearly bad as you maintain. Lowe was removed from the first game because he had accumulated 97 pitches. The Braves had apparently decided that 100 pitches was the max that they were going to allow for their starters the first time through the rotation. That doesn’t seem like an unreasonable decision. As for bringing in Gonzalez, he was injured last season and coming off of surgery, so I’m sure that it was Cox’s thinking that he wanted to get him in there and let him get his feet wet. That too seemed like a sound decision. I’m sure he was thinking the same thing in the second game when he brought in Soriano and Gonzalez. He wanted them to become accustomed to their roles.
That said, I’m afraid that Cox will learn nothing from this game. He’ll just chalk it up to bad luck. Cox is a players’ manager. Most players are apparently more comfortable with well-defined roles, so that’s what Cox gives them, which means that we Braves fans could be seeing more of this.
Chuck your initial points are very valid relative to the current mindset being put to work by major league managers. My point is that this mindset really needs to go. There should not be “roles” persay. The idea of roles led to Farnsworth pitching against Thome in that game when you had Soria, Mahay, and Cruz all available and better options.
Besides, their “role” is to get hitters out.
I’ve got to agree with Eric here. From a flawed perspective, it probably seemed like a sound decision to manage the bullpen according to established roles, rather than according to what best puts the team in a position to win games. Sure, you don’t want to sacrifice long-term success for the sake of one game, but I find it hard to believe that whatever benefit a team games from having its bullpen pitchers comfortable with their roles is worth the cost of not using your best relievers when they’re most needed.
Just compare the leverage index values for when Soriano & Gonzalez were used in the first two games to the values during the 7th-inning implosion last night. During the first two games, the index was below 1.0 when the two “ace” relievers entered the game, while last night it was above 5. It doesn’t matter how comfortable your pitchers are if your most talented relievers are used to protect safe leads while your least talented relievers are used in the most important situations.
Monday Morning Quarterbacking is about questioning a decision only because the results were bad. What Eric is saying is that the decision-making process is flawed (just like it was in the game that Farnsworth lost), and that this process is more likely to lead to bad results.
Greg,
See it doesn’t matter what we’ve seen.. the idea is that what we’ve seen is wrong. I do agree that babysteps would call for the setup guy used prior to the closer. For instance, in last year’s World Series, Game 5, when it re-started in the 7th, the Phillies used Ryan Madson, arguably the best setup guy in the game right now. Granted this probably cannot happen every single game due to fatigue, but they felt the best option at that point was their best non-closer, who was Madson.
When the Braves relievers were incapable of throwing strikes in that 7th inning, that tells me a relief ace needs to come in. EVEN IF Soriano comes in and gives up runs, he would have been a better option than leaving those other guys in, especially given his status as a relief ace. And then you go from there. Say he comes in and escapes the 7th with a 10-8 lead. Okay, then the Braves bat and you make a decision afterwards regarding the 8th and 9th. Maybe Soriano pitches the 8th again and you go to Gonzo in the 9th. Or maybe an entirely different reliever pitches the 8th depending on who the Phils have up… Feliz, Ruiz, pinch-hitter ain’t exactly Gonzalez-worthy.
First, I should say that I always enjoy your work on this site. So thanks for all your writing/insight/analysis. In this case, however, I think you’re wrong about the availability of Soriano. I believe in his post game comments, Cox said he would have brought in Soriano to pitch the eighth or ninth if they held the lead. In the seventh, O’Flaherty got in trouble so he (I think correctly) brought in Moylan, who excels at getting ground ball outs and who is probably the 3rd best guy in the pen. When Moylan didn’t have it, he got Boyer up quickly and brought him in. At that point, the lead was still significant, but Boyer threw 1 strike in 9 pitches. For me, that was the tipping point in going from a bad inning to an out-of-control inning. I don’t really have a problem bringing in Boyer with a 4 run lead in the 7th, but it didn’t work out well. To follow Boyer, I think there’s a good argument that we should have seen Soriano instead of Campillo, due to the leverage of the situation. I’m guessing that’s the decision you find fault with. But, I don’t think it was because he was unavailable due to being used the night before. I think he was betting on Campillo being able to come in and throw strikes. It turned out that didn’t work either, but it took a lot of bad pitching to make this one go down. It does seem a little like monday morning quarterbacking.
Also, at this stage of the season, I think managers are really trying to sort out what they have in their bullpen and the roles that each guy will fill. For the Braves this is certainly true. Just from following the Braves for as long as I have, it seems as though Bobby Cox experiments a little with guys in different situations for the beginning part of the season and then develops a few guys that he trusts and rides them (maybe a little too hard) the rest of the way. Bringing in Campillo in the 7th yesterday does fit that pattern. Last year, Campillo tended to baffle hitters the first time through the lineup and then became more vulnerable. Part of the decision may have been to see how Campillo would handle a late relief role.
If Soriano was available, then Cox made an even bigger mistake. Soriano or Gonzo should have been brought in instead of Boyer, when it was a 10-6 with the bases loaded. I was following the game on Talking Chop and someone even mentioned that then would be a great time to bring in Soriano.
I don’t disagree. I’d like to have seen that move myself, but I’m not too up in arms about it. I appreciate Eric’s larger point about bullpen roles, but I also think there is something to be said for experimenting with different guys in different situations at this stage of the season. I’d like for Boyer to develop into a guy who can come in and shut down a team that is threatening to score runs in the 7th inning. So, I think there is value to putting him in that situation. The Braves don’t really know what mean any of their relievers will regress to. I think they would like for Boyer to be a guy they can turn to in high leverage situations, same with Moylan. But, you’ve got to find out somehow.
ADC, I think agree with you. If this was a playoff game, Soriano is in instead of Campillo or even Boyer. Cox may have been thinking that what he would gain by what he would learn from how Boyer or Campillo handled this situation outweighed the increased risk of a loss. The Braves have playoff aspirations. If Cox does not want his elite relievers burned out in October, he will have to learn which of the second tier guys he can trust. Of course, to Eric’s point, they will also be fresher if Cox would stop using Soriano and Gonzo in low leverage situations. Also, how can Soriano be an elite reliever if he was traded straight up for Horacio Ramirez?
Eric, in principle, I agree with you about bullpen roles. Unfortunately we live in an imperfect world and a manager pretty much has to assign his relievers certain roles. For one thing, there’s the politics of it. Closers expect to get saves. Those who get a lot of saves get the biggest contracts. If you tell your closer that you’re going to use him as a stopper who will be brought in to pitch in high-leverage situations whatever the inning, you’re probably going to meet resistance. He’ll feel like you’re hurting his ability to earn a larger contract in future seasons. There’s also the psychological aspect of it. Some guys may pitch really well in low leverage situations but tighten up when the leverage is high. It’s hard to find such guys just by looking at the numbers, mainly because of small sample size considerations. Hence, the manager has to go by his understanding of the mentality of the individual.
I’m not defending Bobby Cox here. I think he’s too rigid in the way he uses his relievers. What I am saying is that bullpen usage involves human factors that must be taken into consideration. The Red Sox, who are probably the most “saber-literate” organization in the majors are a perfect example of that. A few seasons ago, they tried to use their bullpen in the fashion suggested by sabermetric studies, but they met such resistance that they’ve pretty much reverted to traditional roles.
I mostly agree with Eric on this one. Once the outcome of the game became in question, Cox should have employed his better relievers. To allow the bullpen to choke up that kind of lead was malpractice from Bobby.
At the beginning of the bottom of the 7th, the Braves win probability was over 99%. I am curious the last time a team lost a 7 run lead that late in the game. I honestly cannot remember something like that happening to the Braves in recent years although they did blow a 6 run lead last year to the Phillies.
As a long time watcher of Bobby Cox, you can count on all of those relievers getting into a game very soon to regain their confidence. It will be interested to see how they respond. I would hope that Boyer and Moylan aren’t put into a high leverage situation until they prove they can throw a few strikes.
Great write up. It’s interesting because I am used to seeing the opposite, where a manager doesn’t use his closer for a week because no “save” situation comes up. Do managers make decisions based strictly on gut and memory recall?
It may be April, but a game is a game. The NL East will be tight this year and they had a chance to go up 3-0 on the Phillies to start the season. This might haunt them even if no mainstream writer will refer back to this game in late October.
Cox is not a good game manager and never has been. See game six of the 1991 WS and game 4 of the 1996 WS. He is excellent at putting a team together and can tell what kids can play and which can’t. But he has no idea how to manage a game. If you could have him build the team and Spark Anderson run it, they’d have won about six WS.
I agree strongly both with the initial post and with you Mike. Bobby Cox is absolutely dreadful at managing a game (despite the fact that he is beloved by the players). This goes double for the bullpen. Any Braves fan will notice that Bobby finds one reliever and beats them to death over the course of the season, seemingly regardless of their how they have been performing. Reitsma, Boyer (last season), the list goes on. I am baffled at his choices to bring in certain relievers in close games when they have shown no capability of handling the situation.
The decision to bring in Boyer after Moylan becomes even worse when you consider Boyer’s 2008. This is a reliever that struggled mightily in the second half, often times with his control. Bobby plugs in a guy coming off a 5.88 ERA year who needs to build confidence with the bases loaded and one out. If you are arguing purely on statistics it shouldn’t matter the situation, but it seemed apparent that Boyer was rattled (throwing 1 strike in 9 pitches). Gonzalez and Soriano are much more used to the high pressure situations and probably would have fared better at getting the ball over the plate.
Either way, with Campillo’s relatively strong performance last year I would have started off the 7th with him in the game, regardless of the lefty-lefty match up that Bobby values so much. Campillo proved his worth last season, so why throw O’Flaherty out there on back to back days?
@ Eric
I agree the closer should have come in in the 7th as that is when the game was on the line. Billy Martin once used Gossage in the fourth inning of a game. When asked why he said, “Because that’s when the save was.” I am not advocating that tactic, but why does the save only have to happen in the ninth?
OK, Case One:
You use Blaine Boyer and Peter Moylan in games 1 and/or 2, sparing Gonzalez for other games. If the first two succeed, then you have Gonzo in reserve for game 3 and win it.
Case Two
If the first two fail, then you bring in Gonzo to save the first two games, meaning that he is not available for Game 3. In that case, you’re no worse off than you actually were.
That’s certainly the right management sequence. But who, least of all Bobby Cox, would have guessed that he would need Gonzo in game 3 to preserve a 10-3 lead? Losing one of the first two games (without Gonzalez) seemed more plausible at the time.
Tom, its not about predicting that Gonzo will be needed in a 10-3 game (how could Cox know that in game 1/2 anyway?), its about not wasting the limited IP from your top guys in low leverage situations (I know they need to get work in occationally, but why two days in a row for Gonzo with a 4 run lead to start the 9th?). Gonzo was not available because he was used in games he wasn’t needed. And Cox wasn’t going to use Soriano that early in game 3 because Gonzo wasn’t available. Now, if Gonzo or Soriano were ultimitely needed in games 1/2 (which is highly unlikely), you can no longer critisize Cox because he leveraged his best relievers effectively in games 1/2. And if that means they are unavailable in game 3 so be it, you aren’t always going to have your best guys available when they can only combine for 120-160 IP. But what you can do is make sure those 120-160 IP are of the most value. Cox didn’t do that. Being up by 4 in 9th is going to be a win about 99% of the time, you don’t need your best guy for that. What would be nice to have your best guy for is a bases loaded up by 4, 3 or 2 runs, with just 1 out…. If, say Soriano, squashes that rally and keeps a 3 run lead, its much more likely the Braves kepp the lead for the next 6 outs with the bases starting empty. Compare that letting poor relievers try to finish the bases loaded one out jam, and using your good reliever at the start of a frame, maybe when the lead has already been surrendered.
Bottom line, Cox managed conservatively to win games 1 and 2, but by doing so he reduced his chances to win in game 3. Done correctly (as has been outlined), he could have increased his chances to win all games.
Pardon my joining the discussion late.
Eric, I wholeheartedly agree with you regarding your analysis of the decisions regarding bullpen usage.
I quibble with your assertion that Bobby Cox is a tremedous manager. I’ll certainly maintain that he WAS a great manager, but managers have a career aging curve, just as players do, and Cox would be expected to be well into his decline. It’s tough finding a metric for that, but Clay Davenport’s third-order win differential (D3) had Bobby Cox and his Braves at -7.1 wins last season, for what it’s worth. One could, instead, point to how the Braves have struggled more to reach expectations in recent years, and use that to suggest that Cox might be well past his peak.
A manager past his peak might be expected to make suboptimal decisions more frequently. Blowing Wednesday’s game might be included as a suboptimal decision–not a terrible decision, because so many have supported Cox here–but still a decision less good than Cox might have made 15 years ago.
i do not wish to subscribe to your newsletter.
you’re shorting the phillies offense and how well it beats up bad pitching. if moylan or boyer or campillo goes into the games sunday or tuesday there’s a good chance the phillies put up runs on them then as well.
this isn’t bad managing by cox as its more of a turn of bad luck. i think the fangraph said that there was a 1% chance of the braves losing, and it happened. its a statistical anomaly that won’t happen again. but its not cox’s fault that 3 different pitchers couldn’t throw a strike or that the phillies are just a very good offensive baseball team regardless of what the SP of the braves had done to them the past 3 games.
i’ve said on this site in the past that this offense is going to be more like 2007 than 2008. the 2007 squad was one of the best national league offenses of all time. their OBP was i think second in the league and they pounded the hell out of the ball.
give credit where credit is due.
“this isn’t bad managing by cox as its more of a turn of bad luck. i think the fangraph said that there was a 1% chance of the braves losing, and it happened. its a statistical anomaly that won’t happen again.”
I think this is actually making the opposite point from what you’re trying to argue. If a highly unlikely event takes place, then it’s less likely that it occurred as random chance and more likely that it was — at least in part — the result of poor decision-making. The 1% probability that the WPA chart provides is based on a large collection of similar situations in other ballgames, so it essentially assumes an average level of managerial competence. If, on the other hand, you have a less competent manager, then you would expect the odds of losing in that situation to be higher than 1%. In other words, it’s more likely that it was due to a combination of bad luck and incompetence than to bad luck alone.
Your other points are valid — the Phillies are a good offensive team and it’s the responsibilities of the pitchers themselves to throw strikes, but that doesn’t mean that Cox is blameless for making decisions that kept his best relievers from contributing when they were needed most.
Jamie, I’m not Bill James. I’m a KU graduate living in a coastal state far from other Kansas grads who was called “Jayhawk Bill” before anybody knew that Bill James had the same nickname. Don’t hold anything I post against Bill James, please.
Regarding our differences, thanks for commenting!
that wasn’t a response to you…!
i think if anything, the shellacing that the phils offense gave the braves on wednesday reaffirms cox’s decision to go with his good relief pitchers in the ‘close games.’
further proof these past 2 games for the phils. they destroy mediocre bullpens.