04 November 2014

Constructing the Orioles 2015 Bullpen

It's no secret that the bullpen was a strength of the 2014 Orioles team. The bullpen finished 6th in MLB in ERA and 5th in fWAR (although just 15th in FIP). Zach Britton was a revelation, going from failed starter to being one of the better closers in baseball in 2014. Darren O'Day put together another great campaign, making picking up his $4.25 million option a no-brainer. Andrew Miller was brought in at the deadline to bolster the back end of the bullpen and was dominant during his time in Baltimore.

Orioles fans saw first-hand just how important bullpens can become in the postseason. In the ALDS, a strong effort by the club's bullpen, combined with an atrocious Tigers' bullpen, helped to lift the team to the ALCS. The Orioles' bullpen wasn't awful in the ALCS, but it sputtered at times, while the back-end of the Royals bullpen was strong.

Looking at the success of the Orioles' bullpen this season, and the importance of bullpens in the postseason, might lead one to believe that bolstering the bullpen with the best relievers available on the free-agent market should be a top priority for the Orioles this off-season. We've written here before why relievers shouldn't be overvalued. Orioles' fans need only look so far as Britton, Jim Johnson, Tommy Hunter, Brian Matusz and so on to know that great seasons from a reliever can come from unexpected and cheap options. So, offering a reliever top dollar (Hey, Andrew Miller!) is generally not the most efficient use of resources and is certainly risky.

The Orioles are pretty brilliant with how they approach their bullpen strategy. They get rid of relievers such as Jim Johnson once they become overpriced, and they are also really good at accumulating players with options, such as T.J. McFarland, Brad Brach etc. in order to increase the odds of finding lightening in a bottle during any particular season.

With that in mind, I'm going to take a look at the internal options for the Orioles' bullpen this year as well as a few external options that they should consider.

Zach Britton-Britton, coming off a dominant season as the Orioles closer, is entering his first year of salary arbitration. While he will certainly make more money in 2015, Britton's arbitration number will still be reasonable. If he continues to put together solid seasons as the closer, he could eventually get the Jim Johnson treatment and be shipped out of town, but for 2015 Britton is a lock to enter the season as the closer.

Darren O'Day-As I said above, the Orioles smartly picked up O'Day's option for 2015. $4.25 million isn't dirt cheap for a reliever, but it's certainly reasonable given O'Day's level of production and, more importantly, it's only for one year.

Photo by Keith Allison

Tommy Hunter-Hunter is entering his third year of salary arbitration after earning $3 million in 2014. He will get a raise. If the Orioles are looking to reallocate resources, he could be moved. But I think the club should keep Hunter around given the fact that the financial commitment is only for one year. After a rocky stretch as the team's closer to start 2014, Hunter settled in to become an integral part of the pen, leading all Orioles' relievers with a second-half fWAR of 0.6.

Brian Matusz-Just like Hunter, Matusz is entering his third year of arbitration. He earned $2.4 million in 2014 and will also be getting a raise. I think the Orioles should cut ties with Matusz, whether that be through non-tendering him or trading him. Don't get me wrong, Matusz had a great second half for the club, but the Orioles will get better value by letting a cheaper option handle his innings and reallocating his salary elsewhere.

Ryan Webb-Webb seemed to fall out of favor in Baltimore, but I hope he's back and given a chance to contribute next season. He's under contract for a reasonable $2.75 million. He also was pretty good when he was given the chance to pitch in Baltimore this season.

T.J. McFarland-McFarland is one of the players who could receive an expanded role should the team move Matusz. He'll be cheap and he has two option years reaming, giving the team the roster flexibility it covets. That being said, I'd expect McFarland to spend the majority of the season in Baltimore.

Brad Brach-Brach isn't eligible for arbitration until next season, so he'll be back on the cheap. He is out of options, but Showalter clearly trusts the guy, so it seems as if a spot in the bullpen is his to lose in 2015.

The Orioles have other options, although not all currently on the 40-man, who will compete for time in the bullpen next season such as Evan Meek, Tim Berry, Mike Wright, and Suk-min Yoon. Dylan Bundy could also be an option at some point in the season, given the Orioles' crowded rotation and the fact that he's had trouble staying healthy (not a long-term move to the bullpen, but he'll be considered if the team is in a pinch and he's rested). Buck Showalter recently hinted at the fact that some of the Orioles' young arms in the system could be used as relievers in Baltimore before being given a shot at the rotation. Expect the Orioles to add a few guys with option years to minor league deals or off of waivers, as they recently did with Patrick McCoy.

So, here's where the bullpen stands now, as I see it.

1) Britton
2) O'Day
3) Hunter
4) Brach
5) McFarland
6) Webb
7) ???

Webb might not be back given his 2014 usage, 2015 salary, and the fact that he's out of options, much like all of those relievers except for McFarland. The Orioles may prefer to use his spot on a cheaper reliever with options that gives them roster flexibility or on a LHP, particular if Matusz is not brought back. But, regardless, the Orioles have a spot or two in the bullpen that they should look to fill this off-season, probably with a LHP given that McFarland is the only LHP of the group above. Here are a few players I think they should consider.

2014 Salary
Zach Duke
Tom Gorzelanny
$2.95 mil
Neal Cotts
$2.2 mil
Joe Thatcher
$2.38 mil

I've ordered the pitchers by K/9, as the Orioles should target strikeout relievers to help boost a bullpen that is all but certain to lose Andrew Miller after finishing 24th in the league in K/9 in 2014.

Duke is a very appealing option after a stellar 2014 campaign. His cost shouldn't be anything too outrageous, as he wasn't even the Brewers primary setup option last season. This was his first full season dedicated solely to being a reliever and he thrived. While Steamers projects him to come back down to earth a bit in terms of K/9 next season, he would be welcomed addition to the Orioles bullpen, likely at a very reasonable salary.

Gorzelanny pitched well for the Brewers once he recovered from his 2013 shoulder surgery. With questions about his durability and limited high-leverage use in 2014, Gorzelanny could be a good value in the Orioles 2015 bullpen.

Cotts is another LHP option that strikes more batters out than the average pitcher in the Orioles' bullpen. Of the 4 pitchers I've listed, I think Cotts may draw the most attention around the league. However, a reliever who has just 4 career saves and is coming off a good but not dominant season shouldn't command too high of a salary.

Thatcher is another left-handed reliever on the market who can likely be had for a very reasonable value. Thatcher's 2014 BABIP of .340 is well above his career mark of .316, suggesting a beneficial change in luck may be headed his way in 2015.

The Orioles 2015 bullpen should include Britton, O'Day, Hunter, Brach, and McFarland. I'd like to see Ryan Webb in the bullpen as well but wouldn't be terribly upset if they shed his salary and allowed multiple players with options (such as McCoy), on minor league deals, and Spring Training invites to compete for a spot. That still leaves one spot open for the Orioles to go out and sign one of these 4 LHP that would provide a boost to a bullpen that needs to strike more people out next season.


Matt Kremnitzer said...

Nice write-up. The O's showed interest in Neal Cotts at the trade deadline last year (http://rangersblog.dallasnews.com/2014/07/report-rangers-orioles-discuss-potential-trade-that-could-send-neal-cotts-to-baltimore.html/). So it wouldn't be surprising if they were still interested in bringing him aboard. I'm not sure it would be much of a fit, though, especially if he replaces Matusz. Cotts is actually better against right-handed batters, and he's not exactly dominant against them to begin with.

Pat Holden said...

Yeah, I should have been more specific in saying that this LHP need not necessarily be a matchup pitcher depending on what they do with the 7th bullpen spot if Webb doesn't return.

And Cotts is a bit better against RHP, especially in 2014 and 2008. But in 2013 he was pretty effective against both. His BABIP vs. LHP is also higher against lefties, despite a better LD%, so there could be something else going on there as well.

Thanks for visiting out site, Matt.

Philip said...

I'm delighted to see Neal Cotts' name on this list. I live in Texas, and the Rangers were my team before I started following the Orioles, and I've liked Cotts for a long time.
What about Brett Anderson? He's been injured a lot but is an extreme ground ball pitcher and wouldn't be very expensive, and he'd probably accept a BP role, given his injury history.
Btw, I'd really love to get rid of Hunter. He's just scary, but he does have a modicum of trade value.
RH relievers can be had a lot cheaper than Hunter's expected cost, so I'd love to move him in favor of almost any other 60 inning RH reliever.

Pat Holden said...

I think Hunter could be moved. But the commitment will only be for 1 year and he was pretty darn good last season after he was moved out of the closer role.

Anderson could be a candidate if he was looking for bullpen work but I'd rather see them go for guys with better K/9.

John Morgan said...

Strikeout pitchers cost money, hence the reason Miller will most likely not be back. Duquette has obviously decided to go for pitchers that are undervalued because they have lower strikeout rates and higher FIP's, with the idea that the defense will complement their skills. The strategy has worked nicely thus far, so there really is no reason to change.

Matt Perez said...

If the Os keep all six starters then there may not be room for a LOOGY. Either that or McFarland may be sent to Norfolk.

Matt Kremnitzer said...

I wouldn't be surprised if they traded away a starter. Norris and Chen are free agents after this year anyway.

Matt Perez said...

Wouldn't be surprised either. I'm just not sure what they'd receive in return. And lots of starters will be on the market.

Michael said...

I'm concerned about O'Day's decline over the last month or so of the season. He gave up numerous costly home runs (McCann, Ortiz, Cabrera, Gordon).

Given that, I would be in the mix for Miller. Buck's use of him in October speaks volumes as to his value.

Pat Holden said...

John, I don't think any of these guys will cost the team much.

Matt-I think they'll deal a starter. Have you seen Norris' arbitration estimate? Yikes.

Matt Perez said...

The MLBTR arbitration estimates have better PR than accuracy. But I think he may be right about Norris.

Lots of teams have starters to deal. Just ask Oakland and the Reds. And then there's the FA market. Should be interesting.

Pat Holden said...

I thought, at least the first couple of years with their model, they were pretty successful with their estimates, but maybe that's changed or I'm not remembering that correctly.

Matt Kremnitzer said...

Their estimates are fine to use as a guideline. They are certainly better than anything I could estimate.

Jim said...

I'm not sure I would call this more PR than accuracy. It seems their model does pretty well.

2013 MLBTR estimates vs reality (arbitration and agreed terms):
Jim Johnson: 10.8 MM vs 10 MM
Chris Davis: 10 MM vs 10.35 MM
Matt Wieters 7.9 MM vs 7.7 MM
Bud Norris: 5 MM vs 5.3 MM
Tommy Hunter: 3.1 MM vs 3 MM
Brian Matusz: 2.1 MM vs 2.4 MM
Nolan Reimold: 1.2 MM vs 1.025 MM
Troy Patton: 1.2 MM vs 1.275 MM
Steve Pearce: 1.1 MM vs 0.7 MM

Matt Perez said...

This is going to be complicated. I'm not sure how best to explain this.

The 2014 Orioles were brought up so let's discuss Brian Matusz. He made 1.6M in 2013. The model predicted 2.1M for 2014 while he received 2.4M.

On the one hand it's off by .3M or 12.5% (2.4/2.1). One looking at this may say it's close enough. On the other hand, one may say it's off by nearly 20% ((2.4/1.6)-(2.1/1.6)). Is it accurate?

Fortunately, it doesn't matter whether it's accurate or not. The range of plausible values is such that it's impossible to be off by a significant amount. He almost definitely wouldn't make more than $3.2M. He wouldn't make less than $1.6M. His range of plausible value is at most within $1.6M and realistically between 800k. Anyone with a reasonable understanding of math, baseball and reading patterns would be able to come within a few hundred thousand of the right answer.

For most players whatever numbers the model spits out will be legitimate because pretty much every value in a plausible range will be seen as close enough.

I'm not sure any of that will be understood. So let me put it like this. When it comes to a player like Norris the model has a 45% chance of being within 5% in either direction. If Norris is in that 45% then he'd earn somewhere between 8.265 and 9.1M. It's a guideline but not exactly gospel.

Jim said...

I think you might be a bit too fixated on percent. One would expect players on the lower end to fight more strongly for an extra hundred grand than those at the top end. Additionally, one would expect players earlier in the arbitration process to fight more strongly for a higher total because that impacts future arbitration years.

That means we should expect greater variation on the lower end of the projection than on the upper end of the projection. In assessing populations and overall payroll, this seems pretty sufficient to me.

It is not gospel, but people certainly have staked gospel on far less evidenced things.

Pat Holden said...

Fair points, Matt.

Back to Norris...anything within the range of the projection makes me think we should move him, given our other options at SP.

Matt Perez said...

The reason why I fixate on percent is because that's one of the criteria they use to declare their model successful. The other is whether they're right within a million dollars for a given player. They neglect to note that half the sample makes less than 2M.

If you ignore percent (which I agree you should) then their model is accurate because they say it is. You're right that there is slightly higher variation for lower paid players than higher paid. But not as much as you'd think because the plausible range is much smaller for lower paid players.

I agree that you need to look at the upper end of the projection. Has anyone written an article or anything about those results?

It's not necessarily inaccurate (especially as I'm not sure how you'd define accuracy in this case). It just includes a large amount of lower paid players which bias the results and make it look better than it is.

Jim said...

I have not seen an article.

Go forth!

Would be interesting to see how accurate the model is with the lower end included or excluded.