Front Page: Is It Time for a New Bullpen Paradigm?

[Patrick Wozniak]

As the season winds down there has been some talk about Taylor Roger’s ineffectiveness when pitching in back to back games. A quick glance at his numbers will confirm what the eye has observed as Rogers has pitched to a 7.71 ERA on zero-days of rest while pitching to a 2.45 ERA overall on the season, and to their credit, the Twins have shied away from overusing Rogers of late.With the emergence of quality relief arms like Tyler Duffey and Trevor May, who are capable of pitching late in the game, and the addition of late-inning arms in Sergio Romo and Sam Dyson (if he’s ever healthy), the Twins haven’t felt as much need to pitch Rogers multiple days in a row. Add in Zack Littell, who has also been really solid down the stretch, and the Minnesota bullpen appears to be in good shape, especially relative to starting pitching and the dearth of healthy position players.

 

With a bullpen that is overflowing with arms due to September call-ups, I was curious to see how the aforementioned relievers have performed while pitching on back-to-back days. With so many quality relief options and the September additions, it seems less necessary than ever to pitch anyone without a day of rest between outings. Of course, rosters will contract prior to the postseason, but with so many good relievers right now (and problems in the rotation), the Twins would do well to consider the best usage of the bullpen. Let’s take a look at how the top six relievers have performed both on zero-days rest (in the left-side of the box) compared to their overall numbers on the season (on the right-side).

 

Download attachment: BullpenSheet.png

Yikes! Those numbers on zero days of rest are pretty atrocious across the board. The one glaring exception is Trevor May, who contrary to the trend, actually has pitched much better without a day off between appearances. Of course, we’re talking about a very small sample size of just 9 2/3 innings, but compared to his peers it seems that May is the man to go to if you need someone to pitch two days in a row. However, in the course of a full season the Twins would have to be careful not to overuse May if he was to be relied on to go back-to-back days more often than the others.

 

As far as the others go, the numbers while pitching with no rest really stand out compared to their overall numbers, especially considering that the no-rest performances are also included in the overall numbers, causing them to be a bit bloated. Looking at the differences in OPS allowed shows that the pitchers are making hitters look below replacement level overall, while Rogers, Duffey, and Littell are allowing batters to look like MVP-caliber hitters when pitching on back-to-back days. The diminished strikeout-to-walk ratios of Rogers and Duffey point to a lack of control potentially due to the extra fatigue of pitching without a day of rest.

 

Before the trade deadline, Minnesota rode Rogers hard out of necessity, but it is really no longer advisable to do so. With five to six high quality relief arms (depending on Dyson’s health) the Twins don’t really need to use any reliever on back to back days. Rocco Baldelli deserves credit for the overall fluidity of the bullpen roles, but the Twins can afford to be even less strict with the positioning of their best arms. Although Rogers is the preferred reliever to bring in to close out the game, Romo has done so on occasion since joining the team, and both Duffey, and to a lesser extent May, can be trusted in the highest-leverage situations. And if the Twins aren’t married to Rogers in save situations, he can be brought in to late-inning situations with more lefties due up or when facing the heart of an opposition’s order prior to the ninth. While the Twins might not want to bring Littell in to end the game, he has pitched really well and seems to be at the point where he can be trusted in the late innings of a close game. Dyson is a bit of a wildcard as he hasn’t pitched in a while due to injury, but if he comes back strong he also has closer experience and is capable of being a late-inning weapon.

 

Minnesota currently has a plethora of lesser relief options that should be considered before pitching anyone other than May on zero rest days. Relievers like Cody Sashak, Ryne Harper, or possibly even Brusdar Graterol could be preferable to the non-rested options. Depending on how the opener role is used going forward, Randy Dobnak, Lewis Thorpe, and Devin Smeltzer would also be viable options out of the pen. The Twins would then be able to use three of their superior options for the seventh, eighth, and ninth innings, saving the other two or three for the next game whenever possible. They could also let one of the top relievers go more than one inning on occasion, especially if they are not asked to pitch in the next game. And while the “lesser options” may not be as attractive as the top five or six, when compared with the numbers of the “superior” pitchers with no rest, they don’t look so bad, and their effectiveness could be further enhanced by being utilized against the bottom of opposing team’s lineups.

 

The Twins have yet to clinch the division, but they are well on their way to doing so and with the number of relievers that they have available there is really no reason to run out any relievers on back-to-back days. The postseason is a completely different animal, and with the current lack of starting pitching the bullpen will be paramount if the Twins hope to advance. The good news is that there are plenty of days off between games in the postseason, which should prevent the top bullpen options from too much overuse. Obviously, Minnesota wants its best pitchers throwing in the postseason, but they only seem to be at their best when they are properly rested. It will certainly help to have such a high number of quality late-inning arms, and hopefully an extra lefty or two.

 

Beyond this season, Minnesota may be able to take advantage of limiting the use of relievers on back-to-back days next season as well. Sergio Romo is the only impending free agent of the top six arms, so the Twins should have plenty of good options for high-leverage situations. They also have plenty of young arms with options left, so they will be likely to keep the Rochester-Minneapolis shuttle going strong, making it easier to have fresh arms available. Beyond that, the new rules for 2020 could also lead to less reliever overuse. With pitchers having to face a minimum of three batters (unless the inning ends first) there should be fewer situational pitcher changes, although the Twins don’t do a whole lot of that due to the lack of LOOGYs in their pen. Rosters will also expand to 26 players which will make it all the easier to carry an extra arm.

 

Utilizing the bullpen is essential not only for the remainder of the regular season, but for the postseason if the Twins are to go anywhere at all. After much fan dissatisfaction with the bullpen in the first half of the year, the Twins now have one of the best pens in all of baseball and are poised to be strong in 2020 as well. It’s always nice to have your best pitcher in the game, especially when the game is on the line, but it appears that the best are only the best when they have proper rest (all you coaches out there feel free to use that handy rhyme with the youngsters). It’s already a huge advantage to have such a large quantity of quality arms, and if the Twins are able to fully utilize their relievers with rest between outings, the bullpen will that much more of a weapon going forward.

 

Click here to view the article

[Shaitan]

It's interesting. But it's also an incredibly small sample size.

[mikelink45]

The entire league is wrestling with relief pitcher overuse.  When Starters only go 5 innings per game - that is 810 innings in a season (162 games) and 1458 innings if they all go just 9 means that 648 innings must go to relief pitchers.  Since most go only 1 inning, that 648 appearances would be spread over 8 (on average bullpen arms).  That means 81 appearances per arm (every other game).  Not sustainable.  

Stat heads its time to figure out how to apply your data in a way that fits with the human arm and shoulder.

 

If the starters went 7 innings we have half the bullpen stress.  Of course if you have what we have left in the rotation we have to double the usage - and next year in September there will be limited call ups. 

 

[Patrick Wozniak]

 

The entire league is wrestling with relief pitcher overuse.  When Starters only go 5 innings per game - that is 810 innings in a season (162 games) and 1458 innings if they all go just 9 means that 648 innings must go to relief pitchers.  Since most go only 1 inning, that 648 appearances would be spread over 8 (on average bullpen arms).  That means 81 appearances per arm (every other game).  Not sustainable.  

Stat heads its time to figure out how to apply your data in a way that fits with the human arm and shoulder.

 

If the starters went 7 innings we have half the bullpen stress.  Of course if you have what we have left in the rotation we have to double the usage - and next year in September there will be limited call ups. 

Yeah, teams definitely need to be creative to avoid overusing relievers. There are still plenty of opportunities to use long relievers to eat some innings though and the Twins have a lot of young relievers they can shuttle back and forth from AAA.

[Shaitan]

 

Stat heads its time to figure out how to apply your data in a way that fits with the human arm and shoulder.

I've never gotten the impression the stat heads are concerned with the human element of the game.

 

[Mike Sixel]

Bullpens, in meaningful games that aren't blowouts, are as effective as ever.... That's what the stats show across the league.

 

Reliever stats overall look bad because there are so many blowouts and teams not even trying to field good players.

Edited September 14, 2019 by Mike Sixel

[yarnivek1972]

Roster sizes haven’t kept up with the way pitchers are used. It’s pretty much that simple. This is the first permanent roster increase since World War 2. Owner collusion did reduce rosters to 24 in the late 1980s. Even going back to 1991, the Twins got 1000 IP from their starters, 448 IP from their relievers.

 

Last year, 833 IP from the starters, 610 IP from the relievers.

[DocBauer]

One thing I wonder about is Rogers' stats being skewed somewhat due to 4+ plus batters faced, and sometimes 2IP earlier in the season.

 

I also wonder, as bad as some of the stats listed above appear to be, how do they compare to the rest of MLB?

 

Arguments have raged for years about pitching, both SP and BP. Statistics can be played many ways, both pro and con. Maybe the Twins pen is horrible with no days rest. Maybe its average or well above average compared to the rest of the league. And that's what is missing here. Just like, the Twins rotation is nothing special and maybe even poor. Except, they have been mostly healthy and pitched amongst the ,leaders in GS, IP, etc.

 

Even when we KNEW the pen needed arms for the long haul, the numbers indicated they were still OK. And then kids came up and did well and a couple trades were made, and the numbers indicate the pen is still good. Or am I just missing something?

 

 

[Huskertwin]

 

One thing I wonder about is Rogers' stats being skewed somewhat due to 4+ plus batters faced, and sometimes 2IP earlier in the season. 

Good point. I was also wondering about Rogers stats when he only pitched one inning compared to 1+ innings. I know earlier in the season he seemed our only option at times, but with the couple of trades and the September call-ups, would he be more effective just pitching one inning a game on back-to-back days than these overall stats show?

[terrydactyls]

In my opinion, the best way to build a strong bullpen is to have starters that can pitch seven innings. Currently, the Twins only have one (Berrios) who might go seven. The front office needs to find at least two starters this off season through free agency or trade.

[notoriousgod71]

 

In my opinion, the best way to build a strong bullpen is to have starters that can pitch seven innings. Currently, the Twins only have one (Berrios) who might go seven. The front office needs to find at least two starters this off season through free agency or trade.

 

Even when they are pitching lights out, the starters are not allowed to pitch into the seventh on most occasions. We had the stretch about a month ago where Gibson, Berrios, and Pineda all dominated with pitch counts in the low 80s and left after six or seven innings.

 

But, dammit, we needed to get Matt Magill some extra work before we released him.

[tarheeltwinsfan]

Nice article. Thanks.

[Richie the Rally Goat]

In my opinion, the best way to build a strong bullpen is to have starters that can pitch seven innings. Currently, the Twins only have one (Berrios) who might go seven. The front office needs to find at least two starters this off season through free agency or trade.

The best starting rotation in baseball, the Cleveland Indians, get 5.76 innings per start, to the Twins 6th most 5.55.

 

Acquiring two starters in one offseason who are mid rotation or better in one offseason is a very difficult task. This particular ask of 22% more starting pitcher innings than the best rotation in baseball isn’t feasible.

[wabene]

This is revealing and good news to me as I was under the impression our bullpen ace had a flaw. Now it appears every pitcher is worse on a back to back. This is an area where a team could gain an advantage with some bullpen creativity. With Alcala coming up there is quite the arm talent in our pen. Today is going to be a great test too see how the staff deploys these weapons.

[Riverbrian]

If Taylor Rogers is our best pitcher. (I Believe he is... Starters included).

 

You will need more innings out of him. Come playoff time... the walls are down... the 162 game structure of rotations, closer, set up, long relief... no longer apply.

 

Meaning... if they need Rogers to throw... He will have to throw, especially if he is our best pitcher. Saving him for save situations and thereby limiting the innings that he can throw is an incredible waste of talent. 

 

The data that states that Rogers struggles with no rest is a small sample size but if it is a direct cause and effect. They got to throw him on multiple days to get him used it or better at it. 

 

Because we will need him to do such things in the playoffs. 

 

Many people look at data and say... :Uff Da we got to turn this down... it isn't working".

 

Sometimes that same data is really saying. We got to turn this up. 

 

 

[yarnivek1972]

In my opinion, the best way to build a strong bullpen is to have starters that can pitch seven innings. Currently, the Twins only have one (Berrios) who might go seven. The front office needs to find at least two starters this off season through free agency or trade.

You can hold to that opinion as long as you choose, but it simply isn’t realistic.

 

There were thirteen pitchers with more than 200 IP in 2018 in all of MLB. That’s probably pretty close to what there will be this year as well. At 33 starts that’s barely 6 per start. Even Scherzer who led MLB with 220 IP in 2018 averaged 6 2/3 IP per start.

[Richie the Rally Goat]

Today was a good template for the Division Series. Smeltzer/Littell is a darn good combo to start the game

[sdangus]

Didn't Rogers have a couple of back to back games that were total meltdowns. A couple of bad outings could really skew the numbers on such small sample sizes. He might have been perfectly fine on the other 12 outings. I seem to remember a good number of sucessful back to back outings he made.

 

[ashbury]

Didn't Rogers have a couple of back to back games that were total meltdowns. A couple of bad outings could really skew the numbers on such small sample sizes. He might have been perfectly fine on the other 12 outings. I seem to remember a good number of sucessful back to back outings he made.

If you go to his game log on b-r.com

 https://www.baseball-reference.com/players/gl.fcgi?id=rogerta01&t=p&year=2019

and click twice on the DR (days rest) column, this sorts his games with 0 days rest at the top. You can give him a pass for one or two worst outliers, and it's still pretty visible that he's been a different pitcher than his normally spotless self

 

Or, you can sort on the Earned Runs column, and see that in the 11 games where he gave up any, 4 were with rest and 7 were without rest, which is basically the opposite of that ratio when he didn't give up a run.

 

This way of looking at it reduces the impact of one bad game, because you're not looking at averages, but just counting "good game" and "bad game".

 

Last year was a little less extreme, but of the same general pattern.

 

I wouldn't be surprised if this were typical for relievers, and as always with relievers you are right to be wary of Small Sample Size, but I don't think this ought to be ignored.

[Twodogs]

I think this shows also that Rogers really should be an 8th inning guy or that guy that comes in and matches up with the other teams best lefty in a critical situation during the game. Could be 6th or 7th inning. But because he is the closer now he has to wait until the 9th. Real good closers are able to come in day after day typically, that's why they are given the 9th all of the time. So it would be nice if the Twins can develop a closer and move Rogers back to 8th inning or whenever he is needed type of guy. Also that would probably save Rogers a little as he wouldn't have to get 5 outs ever, he could come in and get 1 - 3 outs and be done and probably save some wear and tear on him.

[railmarshalljon]

I like where this is going but I want to take it a step further. Half of our bullpen is guys that were starters once. Why not have them go more than one inning and make them unavailable the next day? Look for 5 innings out of your starter (less if he starts to faulter) then hand the ball to a guy like Littell for the next 2. At this point you've gotten to the 8th inning with 2 pitchers. Rinse repeat. Imagine trying to hit Odorizzi for 5 innings then getting cut down by Graterol for the next 2. 

[Linus]

I like where this is going but I want to take it a step further. Half of our bullpen is guys that were starters once. Why not have them go more than one inning and make them unavailable the next day? Look for 5 innings out of your starter (less if he starts to faulter) then hand the ball to a guy like Littell for the next 2. At this point you've gotten to the 8th inning with 2 pitchers. Rinse repeat. Imagine trying to hit Odorizzi for 5 innings then getting cut down by Graterol for the next 2.

I agree with this. If you also consider that just warming up represents a considerable amount of the pitches thrown it makes even more sense to get more mileage out of them once they are warmed up. Meanwhile the other reliever that would have used doesn’t do a thing that day

[Tomj14]

 

You can hold to that opinion as long as you choose, but it simply isn’t realistic.

There were thirteen pitchers with more than 200 IP in 2018 in all of MLB. That’s probably pretty close to what there will be this year as well. At 33 starts that’s barely 6 per start. Even Scherzer who led MLB with 220 IP in 2018 averaged 6 2/3 IP per start.

I agree there really aren't any pitchers that can average 7 innings a start only being allowed 90-100 pitches. To pitch 7 innings and have 105 pitches you need to average 15 pitches an inning. When pitchers like Ordo and Gibson (and a ton of others in the league) get to 0 -2 and it takes 3 or 4 more pitches to get the guy out, that isn't happening.

On top of these pitchers (IMO) feel like anytime after the 5th if they get into even a little jam they will be taken out, so they nit pick and run up their pitch count.

In the past if a pitcher got somebody 0 - 2 they didn't have to waste close pitches because the batter was in a protect mode, now they batter doesn't care what the count is they are freely swing for the fences.

 

[Tomj14]

 

I agree with this. If you also consider that just warming up represents a considerable amount of the pitches thrown it makes even more sense to get more mileage out of them once they are warmed up. Meanwhile the other reliever that would have used doesn’t do a thing that day

On TV the other day (I think Morris) said he talked to Romo and Romo said he told the other relief pitchers not to waste all their bullets in the bullpen, they get 8 warm up pitches. (seems like good advice)

 

[ashbury]

On TV the other day (I think Morris) said he talked to Romo and Romo said he told the other relief pitchers not to waste all their bullets in the bullpen, they get 8 warm up pitches. (seems like good advice)

No offense to Sergio but he doesn't have any bullets, just Nerf darts, so what does he know?

[railmarshalljon]

 

No offense to Sergio but he doesn't have any bullets, just Nerf darts, so what does he know?

 

Talked to Romo. He said "No offense to ashbury but he doesn't even have nerf darts, just poorly folded paper airplanes, so what does he know?