There are dozens of medium-impact moves that litter the months between the playoffs and spring training, and most of them don’t give us the opportunity to see something new. For example, without much fanfare the Padres signed right-hander Trevor Cahill away from the defending champion Cubs for a relative pittance: only one year and less than $2 million.

A former top starter who washed out and then rediscovered his talent in the bullpen, Cahill may fill either of two roles for San Diego. He could bolster the team’s bullpen by pitching well in relief, as he did in Chicago, or he could roll the dice in the rotation, hoping to recapture his glory years. But what if I told you there was an innovative way that he could sort of do both?

The Padres are in a bad way when it comes to 2017. Not only is their starting lineup pegged to include Yangervis Solarte and Ryan Schimpf near the heart of the order, the rotation looks like something out of a Pacific Coast League team’s media guide. Four BP writers mocked it extensively (and cleverly) last month, implying that guys like Jeremy Sowers could make the team.

Try not to avert your eyes, but here’s what the rotation looks like on RosterResource:

Despite my touting of Jhoulys Chacin, this is an awful rotation, even for a rebuilding team. First, there’s very little youth or upside; each of these pitchers will be over age 29 on Opening Day and few have had extended success in the past. There’s also a lack of reliability or “veteran” experience, with only Chacin and Friedrich surpassing 100 big-league innings in 2016. With little rotation help on the horizon and the offseason winding down, the only potential help coming is the likes of a burned-out veteran like Jered Weaver or Jake Peavy. Things are bad.

When you can’t improve the talent on the roster, you have to look at changing how the talent you have is used. Perhaps Cahill is the linchpin to a bold strategy that could do just that: using a one- or two-inning relief pitcher to begin a game before turning it over to the “starter.” For years, I’ve advocated for a team (or teams) to give this a try, dubbing the relief-pitcher-in-the-beginning-of-the-game role the "opener," a bookend to the end-of-game closer. This new pitching role could be used to get the team off to a better start than a middling starting pitcher might, before turning the game over to the pitcher who will likely throw the most innings.

Reliever usage has evolved over the past few seasons. Andrew Miller—and Erasmo Ramirez, to a lesser extent—have revived the role of the multi-inning fireman and teams like the Rays are more likely to throw “bullpen” games in which no pitcher logs more than a handful of innings. Perhaps now is finally the time for some franchise to give this strategy a try? And the Padres, more than any other team of the past several years, are the perfect team to give it a shot.

The Plan

The Padres would initially set up their “starting rotation” as a traditional five-man rotation consisting of:

The two left-handed starters (Friedrich and Richard) take the second and fourth slots in the rotation, which separates their game days as much as is possible on a five-man schedule. However, instead of entering the game in the first inning on their “starting” days, the Padres make Cahill the first pitcher to appear in the game during the first inning. In my strategy, the ideal opener also should throw from the opposite side of the “starters” that he opens for; put a right-handed opener ahead of left-handed pitchers and vice-versa. Cahill’s a righty, while Friedrich and Richard are southpaws. Super.

Cahill, working as the opener, faces the top of the opposing lineup and pitches through the first inning or two, depending on how things shake out. If he gets banged around in the first inning, perhaps the Friars go to Friedrich in the second inning. If things go smoothly through the first, then maybe the Padres wait until the start of the third inning or Cahill’s batting spot in the lineup before rolling out the longer-term pitcher. At that point, the lefty then enters the game and pitches as if he were the traditional starter. Friedrich or Richard pitches until they are no longer effective or until the team wishes to engage the bullpen—perhaps around the seventh inning or so.

In this scenario, Cahill “starts” two games a week and—provided there are no injury issues—pitches somewhere between 65-130 innings over the course of the season. The lefties throw fewer innings than they might if they were traditional starters, but maintain their rotation slots unless they suffer an injury or untenable ineffectiveness.

The Benefits

So why would a team mix things up and try a strategy like this? I’ve got a half-dozen reasons why it could be a good baseball decision and not just a fun exercise in trying something new.

Put The Better Pitcher on the Mound More Often

To put it bluntly, don’t expect many of these gentlemen to be substantially better than replacement level this season:

That’s not just a list of crushing baseball disappointments, that’s a list of Padres rotation “depth” behind team “ace” Jhoulys Chacin—the only starter currently projected to be anything better than half a win or so by PECOTA. Is there a breakout candidate in the bunch? Probably not, given these pitchers’ ages, repertoires, and pedigrees, though I’ve always had a soft spot for Jenkins.

Substituting an inning or two from any of these pitchers for an inning from a solid bullpen arm like Cahill is likely a small upgrade. Cahill’s time with the Cubs saw him emerge as a moderately successful reliever with a great ground-ball rate and solid strikeout numbers. Though he once lived the life of a high-quality starter, I’m not convinced he can return to an extended starting role at this point in his career. The right-hander likely needs to maintain the velocity he’s flashed out of the ‘pen if he’s going to be valuable in an inning-by-inning context and I’m not sure he can do that throwing five innings per game.

The best candidates for an opener position are pitchers who used to be starters, are effective enough to be setup men, and won’t completely fall apart in two-inning stints. Given his performance and background, Cahill seems to tick all of those boxes. Unless the Padres are truly desperate for a large quantity of bad innings, then I’d advocate keeping him in short relief where he’s more effective. This way, the Padres get more bang for their buck. At the same time, using him as an opener allows the team to possibly get more of those short outings and allow him to face more batters without losing too much velocity. Instead of using him when he’s not particularly valuable—say, in a 5-0 blowout—Cahill is used when the game is tied, every time. Sure, there’s not always a bunch of leverage in the first inning, but at least he’s pitching more often.

Avoid the Times Through the Order Penalty (TTOP)

You’ve seen the trends and the data—the more times that a starter turns over the lineup the less effective he becomes. The first time through the order results in slightly better performance for the pitcher compared to their usual overall performance, then decreases a bit the second time and even more the third time through. For that reason, limiting a starter’s innings in favor of the bullpen tends to benefit the team and you see that in the increased bullpen use throughout the league. Leveraging the bullpen for an extra inning or two instead of using the starter a third time through the order likely makes a team’s run prevention better, unless the starter already has a great true talent level to begin with. (As we’ve established, the Padres don’t have any of those guys.)

Not only does using Cahill early limit starter innings, it also shifts when that starter turns over the lineup. Cahill does the hard work of taking on the top-of-the-order hitters. By starting the inferior “starter” off against the bottom half of the order—whom the lefty is more likely to face in the second or third inning—that second pitcher benefits from facing the worst hitters first. That gives them a little leeway for when the lineup has turned over twice; if the starter is going to be less effective the third time through the order, let those extra batters faced be the worst hitters. Perhaps the Padres will be more inclined to give Richard or Friedrich a few extra hitters the third time through the order, since the hitters they’ll be facing are already the least effective batsmen? If not, well, this strategy makes it even easier to shorten the starter’s workload, as managers are already used to going to the ‘pen in the sixth and seventh inning.

Destroy Enemy Platoons and Force Opposing Managers to Make Mistakes

One of the biggest benefits to using an opener is that it allows the Padres to put opposing teams on their heels, especially if the team tends to structure their daily lineups to obtain the platoon advantage. By using the right-handed Cahill early, then switching to the left-handed Richard or Friedrich, the opposing team must choose to do one of the following:

  1. Structure their lineup against the right-handed "opener" and then give up the platoon advantage when the lefty comes in or go to their bench early.
  2. Structure their lineup against the left-handed "starter" and punt the platoon advantage early against Cahill.
  3. Or split the difference, running a mixed lineup and only changing it partially when the "starter" comes in (or not at all).

Using Cahill first before diving into either Richard or Friedrich allows the Padres to get something of an asymmetric advantage early in a game. If they choose to stack the order against Cahill, great! Richard/Friedrich should have a slightly easier time or the opposing team could dig into the bench in the third inning. If they choose to stack the order against Richard/Friedrich, great! The team would’ve done that anyway, but now there’s a slight advantage in the first couple of innings, perhaps; as we’ll see a little later, that may make a difference.

This will affect teams that are platoon-heavy much more than teams that rely on roughly the same seven or eight hitters every day. I’d expect that the Giants, Dodgers, and Rockies are all going to lean on some sort of platoon for two or more regular positions, while the Diamondbacks probably won’t. At least against the teams in the NL West, this could give the Padres a slight advantage before the game even starts, as managers might have to get more creative early on.

A corollary here is an adage that I agree with, but don’t have a lot of data to back up: making the opposing manager act (pull a player, switch a reliever, call for a bunt, etc.) opens the door for that manager to make the wrong decision and give your team an advantage.

Get Off to a Solid Start

I’ve always kind of thought that when a team gets out to an early lead they might be more likely to play better in the future. Yes, this is a soft factors thing, but mindset and confidence certainly play a part in athletic performance, even if we don’t know how to quantify it. But I did take a cursory stab at it. Thanks to Rob McQuown and the BP stats team, here’s the data for runs allowed by teams with an early lead over the past 10 seasons:

RA9 in all situations RA9 with a lead
After 1st Inning 4.40 4.36
After 2nd Inning 4.46 4.38

As you can see, there’s a very slight benefit in terms of runs allowed when a team is winning after the first inning or two, compared to all games. With the exception of 2010 and 2014, there’s been a slight benefit to coming out to a lead in each individual season as well. Of course, the possibility exists that this relates to the fact that teams are more likely to use better relievers (such as their closer) when they’re in the lead, or that they’re in the lead in the first place because of a better starting pitcher. Still, there’s the possibility that there is some psychic benefit to coming out to an early lead that makes the pitchers and defense play better.

Mitigate the First-Inning Effect (Maybe?)

Would you believe that there’s one inning in which teams are more likely to score a run than any other? Several years ago, Jacob Peterson of Beyond the Box Score published a short piece tracking which inning offered up the most runs allowed. It was the first inning, and not by too small of a margin. I re-did some of his work (using the entire MLB instead of splitting it between leagues), and found that the same thing holds true over the last 15 seasons. There’s still a better chance of runs being scored in the first inning than in any other.

I think this requires a great deal of examination as to the cause—my best guess is that maybe hitters are fresher and more physically prepared in the first inning, plus there’s the guarantee of facing the top-of-the-order bats—but again, if the first inning is likely to result in more runs being scored wouldn’t a team want to put a more talented pitcher out there to mitigate the damage? Instead of pitching an iffy starter like Richard or Friedrich, Cahill gives the team a slight advantage during the inning in which the opposition is most likely to strike.

Play to Your Strengths

I’m not sure that any team over the past several years has had the personnel in place to make the opener gambit work as well as the Padres do. Step 1: Have a rotation with few established veteran starters you can rely on. The Padres definitely accomplished that mission. Step 2: Be comfortable punting a reliever slot later in the game. This is a little tougher, as we all know how teams have been prizing extra bullpen arms. With that in mind, a team using an opener may want to add an extra pitcher to the bullpen to compensate for Cahill’s loss, which is tough to do given the 25-man roster limit.

Enter Christian Bethancourt.

Bethancourt is ostensibly the backup catcher to Austin Hedges heading into 2017, but the Padres have also been trying him out as a sometimes pitcher thanks to his ridiculous arm. He pitched twice in 2016, throwing 1.2 innings, and his career ERA is 0.00. This offseason he’s been working out as a pitcher in the Panamanian Winter League. Now, I’m not certain that he’s a replacement for Cahill in the seventh or eighth inning, but he does give the Padres an emergency relief arm that no other team—save the Royals and secret relief ace Drew Butera, maybe—has on the positional side of the ball. If there’s a garbage-time pitching situation, Bethancourt might be able to handle it.

The Padres have a viable bullpen without Cahill, with good-but-not-outstanding arms like Ryan Buchter, Brad Hand, and Brandon Maurer. While this isn’t exactly the Kevin Towers build-a-bullpen-out-of-wishes-and-dreams era, the Friars can certainly survive without Cahill in the mix. And there’s another factor at play here: Cahill has stated a desire to return to the rotation. If the Padres agree with me that a regular five-or-more-inning role isn’t best for the team, perhaps this opener role would be a bit more palatable than a return to the sixth or seventh inning. At the end of his Padres run next offseason, if the gambit is a success, Cahill could say to prospective suitors: “Hey, I can do this for five innings every game instead of two.”

The Potential Downsides

I’ve put together a laundry list of benefits to this strategy, and yet there’s quite a bit of hedging and supposition as well. Unfortunately, there’s simply a lack of hard data to tell us how much overall value San Diego might get out of trying this experiment. Beyond just that, I’m certainly not blind to the fact that there are downsides to the opener plan. Most of these cons aren’t specific to the Padres, but rather are concerns that any team might have when trying a new strategy.

Saving Your Best Pitchers for Higher Leverage

When I first posited this idea of the opener a few years ago, Russell A. Carleton made the most compelling argument against it.

The way that the modern pitching staff is structured is to play for the endgame, and really it has been since the bullpen became a thing. You put the starter out there in a rotating because someone has to pitch bulk innings and you wait for the game to unfold to see what sort of game you have. Then, with the benefit of that information, you decide whether to use the good reliever or the not-so-good one. This sounds like cart before horse. You don’t know in the first inning whether this will be a 13-0 laugher or a 4-3 nail biter, so whether you pitch your best reliever or your fourth best as the opener, why declare in advance of when you need to? Since bullpen use has a carry-over effect to the next game, you want to be careful of whom you have pitch and when.

The argument that Carleton makes here is sound: you’re making the decision to use Cahill with less information than if you were to make that decision later in the game. I’m not sure that there’s much of a way around this—my only claim is that the value of that prior information isn’t quite worth the value of those extra innings on Cahill’s arm along with the other benefits of the opener. How much is using the opener preventing the game from becoming one of those 13-0 laughers? Until we see this role in action, it’s challenging to try to compare the costs and benefits of using an opener, but I do think this line of reasoning is worth considering.

“Shorten” the Bullpen

I’d peg Cahill as the third- or fourth-best pitcher in the Padres' bullpen. Any manager might balk (baseball pun!) at the thought of being without one of the team’s nicest late-game setup options for the entire season just because the front office listened to some idiot’s weird idea. Late-game relievers are something that every team wants more of and it’s possible that this idea would cause teams to think that they’re “losing” a reliever for end-game purposes. Yes, the innings are still being eaten up by that pitcher—and potentially more of them!—but what happens in those circumstances where Buchter or Maurer go down with an injury? What happens when you need a counter for Bryce Harper with runners on in the eighth? In any case, the late-game choices would be fewer with Cahill throwing early.

Change is Hard

Admit it, it would be weird to look up the seasonal leaders for “games started” and see Trevor Cahill looking like Old Hoss Radbourn at first glance. And if it’s weird for you, imagine how strange it is for Friedrich or Richard, who’re unlikely to get the credit (or pitcher wins) as starters that they might’ve picked up otherwise. Getting buy-in for a move like this isn’t exactly a given, especially considering the history of slow uptake of new ideas in sports.


I think the time’s ripe for the Padres to give this strategy a go; with A.J. Preller at the helm, you could maybe even see the notoriously unpredictable GM trying something wild like this. There’s a collection of tiny potential advantages here, as going with an opener before the starter could allow San Diego to get more good innings out of Cahill without making him a full “starter,” blow up opposing platoons, limit the questionable innings of two indifferent pitchers, and maybe get the team out to an early lead—if the bats cooperate.

Yes, there are certainly some potential buy-in issues and the endgame may get slightly dicier, but I’m not sure any franchise is better equipped to try my theory out or has a bigger cushion on which to land. After all, the Padres going for a gambit like this likely won’t make the difference between missing or making the playoffs this year. It’s certainly a big break from the traditions of baseball, so if anyone’s going to give it a shot, it’s most likely a team and rotation with nothing to lose.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
I love this. Traditional roles are so ingrained in baseball that big changes like this are rarely tried, but I'd love to see teams try different approaches. I'd like to see teams use a piggyback system when they break young pitchers into the league. They are always looking for ways to minimize innings anyway and if they use a piggyback system, they would limit innings and also avoid over-exposure of young pitchers to multiple times through a lineup. For example, the Dodgers could have piggybacked Urias and DeLeon last year or the Cardinals could have piggybacked Reyes and Weaver (they still could, actually).
The other nice thing about using a young good arm for this strategy is they get more comfortable opening games vs. coming out of the bullpen. I would have to assume that it's a bit of a difference mentally.
This seems like a good experiment to me too. But I'm pretty sure we have to work through all the other modern bullpen experiments first - eg the obvious - closer, backup-closer, 8th inning lefty, 8th inning righty, 7th inning lefty, 7th inning righty, 6th inning lefty, 6th inning righty, guy who pitches in the blowout losses, guy who pitches in the blowout wins, guy who only comes to pitch in extra innings, guy who doesn't actually pitch much but who warms up pretty much every game in order to deceive the other team. Add the 5 starters and we're already up to 17 pitcher slots. And finding all those lefties is tough enough. Adding an opener to that means we have to eliminate one of the position players. It's just too high risk.
As long as you are experimenting go with some 2 inning relievers. The 6-8 lefty/rightys should be 4 pitchers, two of each. In fact the closer role should be dumped and just throw in your best available reliever into the 8-9 mix. 'Blowout guy', 'extra inning guy' and 'warm-up guy' should be all the same guy and in fact the 5th starter on a skipped rotation can be this guy. So now we are down to 6 RP and 5 SP. Relievers presumably are interchangable on a bad team. So get the AAA train on a roll and you only need 11 pitchers tops.
The other problem is that this might be moderately successful, meaning that it makes the Padres better but not playoff good. The Padre's are clearly tanking this year to improve their draft position. So, finishing near .500 is a worse case scenario for them.
I love this too, and I'd take it one step further. I'd have my three best bench bats, as long as they are better than the Padres' worst three hitters, bat 1,2,3 in every away game and then pull them in the bottom of the first for the starter/defensive replacement in the bottom of the first. Why not put pressure on the other team and maybe give your team an early lead? Then your normal leadoff hitters will hit and you won't see your worst batters till maybe the third inning.

If you don't have three great bench bats, how about two, or even one?

The Padres should be the Sonoma Stompers of MLB this year, maybe Ben and Sam can pull a consulting gig out of this, the weather is sure nice at Petco.
Not a fan of this idea, with benches already super short. The idea of going into the eighth and ninth innings with few or no bullets left to fire is a non-starter, no pun intended.
Isn't some of the problem Mr. Carleton identified offset by forcing the other team to account for the different handedness of the opener from the "starter"? Do they start the guy who mashes lefties but should never see a righty, or bat him lower in the order hoping the opener is gone before he bats? Or do they wait until the opener has left the game and bring him in then but lose the now-starting pinch-hitter for a possibly lower-leverage at bat in the second inning? Yes, the team using an opener is using a pitcher in lower-leverage innings, but they're planning to do it. The opposing team also has to account for that in their planning and either punt on an at-bat early in the game or burn on of their players as well.

Of course I also believe teams likely to struggle should usually go with a more high variance strategy. My employment isn't tied to a team looking competitive even if it isn't so it's easy for me to want that.
Every football coach I've ever talked to hates coaching against a gimmick offense because it's not worth the time to get really good at defending it, which is why so many small under-talented schools use gimmick offenses. They do this in basketball too. Maybe there's not enough wiggle room in baseball rules, but I'd love to see the envelope stretched.
I was thinking of football when mentioning variance. At the NFL level gimmicks don't work as well because the difference in skill level between teams is minimal; the read-option craze petered out because NFL defenders can move laterally too fast and hit too hard so the advantage just isn't worth the beating read-option QBs take.

The skill level between the best and worst MLB teams is relatively thin they still have to deal with the same rules about roster use. The use of an "opener" can force the opposing manager into making a move they might not want to make at a time they don't want to make it, or sacrificing an opportunity.

It's not a good idea for the Cubs, Dodgers, etc, but the Padres?

Of course a counter-point is that even the tanking-phase Astros didn't use anything as gimmicky as this so it might be something considered too obviously damaging to winning.
This would break play index!
Who was the last player who was both a pitcher and position player? Brooks Kieschnick?

The problem with Betancourt as a part-time pitcher is that the Padres might then need to carry a 3rd catcher unless one of the existing position players could do it on an emergency basis. And the emergency route doesn't work if you want to lift Hedges for a pinch hitter.
" imagine how strange it is for Friedrich or Richard, who’re unlikely to get the credit (or pitcher wins) as starters that they might’ve picked up otherwise"

Actually, the opposite is true. Cahill would never be eligible for the win even if he was pitching when the Padres got the lead for good while Friedrich or Richard would be eligible the win even if they only pitch 3 or 4 innings (and would likely be awarded the win if they pitched the most innings)
I had the same thought, but then I realized the downside -- Friedrich and Richard might be entering games SD is already down in, and knowing they have little chance of getting a W knowing the "quality" of offense supporting them.

And now I wonder if SD has or can find a lefty "opener" for two of the starter righties, and then you're only letting your best real starter actually start games.
One of the interesting things about this idea is that the Padres would essentially be be putting a relief pitcher on a fixed schedule. Cahill would presumably never have to warm up multiple times in a game nor would he ever warm up and then not come into a game. Every pitcher and pitching coach that I have heard discuss relief pitcher usage has commented that these two activity really wear down a pitcher.

I suspect Cahill could probably be counter on to pitch 2-3 innings per game if he were on a fixed schedule. Over the course of a season he could open 60+ games and probably log 120+ innings. Assuming he remains effective, 120+ innings like this would be far more valuable than 60-65 innings in a traditional relief roll.
Like the invention of the 6 pack for beer: brilliant!
"It's a bold strategy, Cotton, let's see if it pays off".
The more pitchers you use in a game, the faster you find the one guy who really doesn't have it that day.

Then the manager gets to explain TTTO syndrome post-game and why he thought it was good idea to pull a moderately effective pitcher for the guy who gave up the losing 3-run homer.
I don't know if you want this to happen, if you really like your strategy and want it to be adopted. Any strategy will largely succeed or fail depending on the quality of the players at your disposal, and the Padres are BAD. If they adopted this stratgey they would still lose, and the strategy would be widely panned as one of the contributing factors.

Now if the Yankees did it, with Betances as the Opener and Chapman as the Closer, that could turn heads