The Pirates’ rotation situation will probably work itself out. The discussion about Gerrit Cole potentially returning to Triple-A so the Pirates will have six starters available is just this week’s version of Pittsburgh’s least exciting game show: What Should The Pirates Do With Their Rotation?
In previous episodes, the correct answer was: Someone Gets Hurt. Wandy Rodriguez and Jeanmar Gomez leave a game after one inning or fewer. A.J. Burnett pulls up while running in the outfield. Jeff Karstens suffered a Karstens. Injuries happen to pitchers.
What we don’t know exactly is why pitchers get injured so often or if the rate of injuries can be reduced. In 2002, Rany Jazayerli studied the differences between four-man and five-man rotations, which became popular in the 1970s. One of his conclusions: “pitching in a four-man rotation is less damaging than pitching in a five-man rotation,” even though the difference between the two was not statistically significant.
[Full disclosure: I wrote about the subject of four-man rotations last year for Pittsburgh Sports Report. The Pirates had a 3.50 season ERA when I wrote the piece and a 4.23 ERA after that. I have made every attempt to not plagiarise myself.]
The Pirates find themselves with a problem, albeit a good one. Rodriguez, Gomez and Burnett are all set to return to the active roster soon, and the Pirates need to activate James McDonald from his rehab or pack his bags for another team. At the same time, Jeff Locke and Francisco Liriano are pitching like All-Stars while Charlie Morton and Gerrit Cole each started his Major League season with three good starts. It’s possible someone will get hurt or start struggling to make the decision easier, but what if that does not happen? The Pirates are now considering the idea of demoting Gerrit Cole to be able to keep everyone in the organization.
Any decision will either maintain a five-man rotation or (God forbid) create a six-man rotation, which would serve no purpose other than keeping everyone smiling and giving fewer starts to the better pitchers. Instead, the Pirates should consider a four-man piggyback rotation. It would not only maintain all the best pitchers on the Major League team but also utilize those pitchers in an optimal fashion.
The Four-Starter Plan
Let’s lay out the plan, which I’ll call the Four-Starter Piggyback since every starter has a “piggyback” partner except Rodriguez:
- Rotation Spot 1: Jeff Locke, Gerrit Cole
- Rotation Spot 2: Francisco Liriano, Charlie Morton
- Rotation Spot 3: A.J. Burnett, Jeanmar Gomez
- Rotation Spot 4: Wandy Rodriguez
- Long Reliever: Vin Mazzaro
- Swingman: James McDonald
- Lefty Reliever: Justin Wilson
- Setup Man: Mark Melancon
- Relief Ace: Jason Grilli
The Four-Starter Piggyback would option Tony Watson, and demote Bryan Morris (who has already been optioned this year), Ryan Reid and Duke Welker, but would maintain every pitcher that does not have options. The pitcher who starts the game could aim to throw about 75 pitches or five innings while his partner prepares to throw about 50 pitches or three to four innings.
Is it at all possible that the many pitching injuries the Pirates have suffered have at least something to do with subscribing to a rotation plan out of habit? Every team in baseball now uses a five-man rotation, even though it is not proven that throwing 100 pitches every five days leads to better performance or overall health than 80 pitches every four days. Pitchers actually throw a little better on three days’ rest as opposed to four days, Jazayerli found. Trying something new could be helpful not just for run prevention but also keeping players on the mound.
Who Would it Help?
I’ll rank the starting pitchers that would benefit most from a Four-Starter Piggyback, from one to seven:
1. Charlie Morton would get the best results for two reasons.
He is coming off reconstructive elbow surgery, putting him at a more prominent injury risk. The Pirates need to monitor Morton’s innings in his starts carefully to ensure his health through next season, his last under Pittsburgh control. Morton will not hit a 160-inning limit that Stephen Strasburg hit last year following Tommy John Surgery, but a normal workload of 100 pitches every five days may not suit him with the elbow scar still fresh.
A new role in which Morton only pitches through the opposing lineup twice (4-6 innings) would best utilize his skillset. In 2012, Morton had the 10th biggest difference between his xFIP in the first three innings (3.64) and later innings (4.71). And look at his numbers in 2011:
- 1st time through lineup – .706 OPS, 42 K, 28 BB
- 2nd time through lineup – .662 OPS, 43 K, 25 BB
- 3rd time through lineup – .897 OPS, 21 K, 24 BB
One should not construct a rotation around Charlie Morton, but he can be very effective under a Four-Starter Piggyback Plan.
2. Gerrit Cole‘s situation is similar to Morton’s. First, hitters are starting to crack Cole later in games (1st time – .630 OPS, 2nd time – .264 OPS, 3rd time – 1.032 OPS). It’s small sample but becoming significant. Second, the Pirates are going to want to monitor his innings. Here are Cole’s innings pitched going back to college:
- 2009 – 85.0 innings
- 2010 – 123.0 innings
- 2011 – 114.1 innings
- 2012 – 132.0 innings
Cole is currently pitching incredibly efficiently in the Majors, as his 13.6 pitches per inning would be second-best among starters if he had enough innings, and he is doing it with fastballs instead of potentially more stressful offspeed pitches.
Be that as it may, Cole has pitched 86.1 innings this year, on pace for more than 170 by the end of the regular season, and Major League innings will be tougher to work through than Minor League ones. If the Pirates want to win in September and possibly onward, Cole needs to be healthy and not pitching on fumes. Limiting his innings would be a lot easier if he is pitching four innings every four days instead of six innings every five.
Pirates GM Neal Huntington told Travis Sawchik of the Pittsburgh Tribune-Review: “Part of our development plan is to build guys accordingly so that when they get to the big leagues they are hopefully in a position to log the innings without the media attention that some have gotten. We’ll let you know if he ever gets to his workload.”
The Washington Nationals did not get creative with Strasburg. Instead, their ace was shut down after Sept. 7 and the Nationals were dramatically eliminated in the first round of the playoffs. The Pirates cannot let the same fate befall their top draft pick.
3. Francisco Liriano is more effective his first time through the lineup as well. Over his career, opposing batters hit .217/.304/.339 (26% K rate) against Liriano their first time facing him in a game but .262/.337/.398 (22% K rate) their second and third times. Even though his numbers are actually pretty good the third time through the lineup the last few years, this year’s results are pretty stark, even though it is a small sample of nine starts.
- 1st time through lineup – .423 OPS, 27 K, 10 BB – 3rd among MLB starters
- 2nd time through lineup – .606 OPS, 17 K, 7 BB – 45th among MLB starters
- 3rd time through lineup – .790 OPS, 16 K, 5 BB – 102nd among MLB starters
Plus, Liriano has been inefficient as a starter so far, exiting during or directly after the 6th inning in the majority of his starts.
4. A.J. Burnett is another pitcher that decreases in relative effectiveness as the game goes on. Look at his numbers since the start of 2010 among 232 Major League starters. This is probably the best list of bullet points I have, as it has the largest sample size.
- 1st time through lineup – .682 OPS (93rd), 254 K, 88 BB
- 2nd time through lineup – .723 OPS – (103rd), 202 K, 80 BB
- 3rd time through lineup – .852 OPS – (182nd), 118 K, 74 BB
Beyond that, Burnett has experience pitching every fourth day a few times for the Blue Jays and Yankees and did so effectively.
“When I did it in Toronto, I loved it. Even in the playoffs, I loved it,” Burnett told me last year. “Some guys don’t like it. Some guys want the days of rest and there’s days I wouldn’t want to do it. Just depends.”
It is also a good time to re-start Burnett’s schedule. Coming off an injury, he is not back in the habit and throwing schedule of pitching every fifth day, so it would be easier to get him on a plan of starting one day, resting the next, throwing the next, resting the next and starting the next.
5. Jeanmar Gomez has not been given the workload of a swingman replacement starter rather than a full-time rotation pitcher, never throwing more than 79 pitches in any of his 12 outings. He has produced good results under unique circumstances, posting a 3.07 ERA. However, he is due to regress because his BABIP is low, he strikes out too few hitters and walks too many to be considered a true control pitcher. His SIERA is 4.50, denoting a below-average pitcher and perhaps the least effective of all seven guys, but Gomez is someone who can still be effective if used correctly like in combination with Burnett.
6. With seven starters for eight spots, one pitcher has to go without a dance partner, and I would choose Wandy Rodriguez.
- 1st time through lineup – .683 OPS (94th), 212 K, 81 BB
- 2nd time through lineup – .758 OPS (144th), 173 K, 71 BB
- 3rd time through lineup – .685 OPS (32nd), 133 K, 46 BB
Even with left forearm tightness causing Rodriguez to miss a few weeks, he has been incredibly durable: started 30+ games and pitched 190+ innings each of the last four seasons to earn a workhorse reputation. As with Burnett, the Pirates can reset Rodriguez’s schedule right now prior to his return to the rotation.
7. The only pitcher that does not fit very well into this plan is Jeff Locke. His results this season have been a 2.01 ERA, best in the National League. He has also been one of the NL’s more efficient starters at 15.5 pitches per inning. Why mess with a good thing? And indeed, it would probably be best to not change anything around Locke.
Two counterpoints: He has been making the bullpen labor a bit. Locke pitched seven innings his last two times out, but did so only twice in his first 13 starts. Pirates manager Clint Hurdle has also kept him below 104 pitches in every start, only hitting double-digits three times.
In addition, Locke will continue to pitch well but not likely at NL-best levels. I’ll shorten my normal speech into a trio of stats: a .706 OPS-against with the bases empty vs. a .329 OPS-against with runners in scoring position, a 4.40 SIERA, and a league-high 85.6% strand rate. If you believe Locke’s results will go south, take the victory you have had with Locke in a five-man rotation and see if you can continue to get decent results in a four-man rotation.
Why the Pirates Should Go for a Big Change
The same idea as Locke goes for the Pirates as a whole. One might wonder why the Pirates would mess with the tremendous results they have received in a normal five-man rotation. Since Pittsburgh’s 3.24 staff ERA and 3.37 starters ERA are both hovering among the best in the NL, why should the team do something so radical during the middle of the season?
Number one, the past is the past. The Pirates have prevented runs well, but other stats about the pitching staff are less optimistic about the future: 3.87 xFIP (8th in the 15-team NL), 20.1% strikeout percentage (7th), 9.0% walk percentage (15th) and 3.97 SIERA (8th). The biggest reason the Pirates have been successful is defense (best in NL at turning balls in play into outs) and lucky sequencing (No. 11 OPS-against with bases empty, No. 1 OPS-against with runners in scoring position). Don’t confuse past success as a bellwether of future success. We have better numbers to predict the future, and those numbers see an average staff instead of an NL-best staff.
What the Pirates need is depth to stave off the regression monsters for another half-season, and they have already used the second-most pitchers in baseball this season, behind only the record-setting Toronto Blue Jays. They can’t afford to let any of their Major League-caliber pitchers go to waste.
This plan prevents any pitcher from having to go onto waivers, as odd men out Tony Watson and Bryan Morris have options to return to the Minor Leagues and Morris is in the middle of an option year. Gomez and Vin Mazzaro do not have options, but both have performed very well to earn a spot on a tight roster. Tim wrote that James McDonald could work well out of the bullpen needing only a fastball-curveball combo, and I agree, as McDonald’s numbers have been good his first time facing a lineup: .234 BA, .711 OPS, 2.2 K/BB ratio. The plan does not hinge on McDonald, though, and could work just as well using a pitcher that has started and relieved before like Morris.
Three Other Benefits
- The three primary relief pitchers for the Pirates become Mark Melancon, Jason Grilli and Justin Wilson, the three best relievers the Pirates have had this season. Each would be capable of closing out the 9th inning when needed after getting four innings from each of the “starters.” More appearances from those three pitchers and fewer from effective-but-not-dominant relievers like Watson, Reid and Morris would prevent runs.
- When the pitch count target becomes 80 pitches instead of 100, starters can still go six innings or more. Locke, Liriano, Burnett and Rodriguez have all gone six frames on fewer than 85 pitches at some point this year. When starts like that happen, great! A Four-Starter Piggyback encourages flexibility. If a starter doesn’t pitch one day or only goes two innings, he could be ready again on the next day he throws. The goal is to use all 12 pitchers effectively and keep them healthy.
- If someone important gets hurt and the Pirates need to trade for a starter, they can find better value than other teams. Astros pitcher Bud Norris posted a 3.37 xFIP last year in the first three innings and 4.56 xFIP after that. Tigers pitcher Rick Porcello has held opposing hitters to a fantastic .646 OPS his first time through a lineup over his career but a poor .830 OPS after that. When you are looking for value where other teams are not, you can acquire talent that will give you good bang for your trade buck.
As I said at the start, the “problem” of having too many starting pitchers usually works itself out via injuries or poor play. But there should be a Plan B in place like a Four-Starter Piggyback. It’s radical, at least in today’s risk-averse game. If the Pirates believe it can work, though, and all the pitchers get on board with the belief that the team can win with it, there could be many benefits. Such a system could add two extra wins over a whole season, according to Baseball Prospectus’ Russell A. Carleton, or one win over a half-season. Where the Pirates sit in the playoff race, every win is important.
There is no bible that says every team must carry five starting pitchers and seven relievers, or that such a format is the best way to use a pitching staff. It simply became the staff of choice over the last few decades. Even if the Pirates don’t use a four-man rotation or a piggyback system, the front office should examine closely the strengths and injury risks of each of its starting pitchers to find the optimal plan.
Lightning Round: What Should The Pirates Do With Their Rotation? Open up to new ideas. Trust your professional pitchers to accept different roles. Be careful with the health of these important players.