Anybody who has watched Sean Doolittle pitch this season knows that when it comes to building a successful major league bullpen, sometimes a team just has to get lucky. Doolittle, the hard-throwing 25-year-old set-up man for the Oakland A's has yet to complete his second full season as a pitcher. Not as a reliever. As a pitcher, period. A supplemental round pick as a first baseman and right fielder in 2007, Doolittle converted to the bullpen in the minors in 2011. This year, the rookie appeared in 44 games for Oakland, posting a 3.04 ERA with rates of 11.4 K/9, 2.1 BB/9 and 0.6 HR/9.

Cases like that of Doolittle and Cardinals closer Jason Motte, who began his career as a catcher, underscore the inexact science that is bullpen construction. So what you are about to read is less of a blueprint and more of a case study in the multiple methods a team can use to build a unit of relievers capable of pitching for a contender.

We aren't going to breakdown the bullpen the same way we have broken down the other positions on the roster this week, because a bullpen really is a conglomerate of four different positions: the closer, the set-up man, the specialist, and the long man.

Along the way, we'll attempt to make some rough generalizations about the nature of each position. Really, though, there is no one way to build a bullpen. There is, however, a wrong way, as the Phillies showed in 2012, when they lost 27 games in relief, the fifth most in the National League.

I. Economic Stratification

Logic suggests that the wrong way to build a bullpen is to pour a disproportionate amount of resources into one reliever. Assuming that a general manager is budgeted a specific amount for payroll, he must then decide how to allocate that payroll across the various positions on the roster, which means he must then decide what percentage of that payroll he will allocate to the bullpen, which means he must then decide what percentage of that allocation to spend on each reliever. Basic business accounting. If all of this is true, then every dollar spent on one reliever means another dollar that cannot be spent on another reliever. The economists call it opportunity cost. So when the Phillies decided to spend $11 million on Jonathan Papelbon this offseason, that meant $11 million that they were not able to spend on the rest of the bullpen.

A quick look at the rest of the National League suggests that teams should err on the side of distributing that bullpen spending equally rather than tying up a large percentage of it in one player.

Below are the six relievers in the National League whose salaries accounted for more than 50 percent of the money that their team spent on the bullpen. I calculated bullpen spending by adding up the six highest salaries paid to relievers on Opening Day. Teams carry six or seven relievers. If they carry a seventh, he is usually a long man who does not factor into the situations that end up defining a bullpen. Which is why I chose to calculate the top six salaries. In parentheses are the bullpen's final NL rank in ERA followed by the final regular season record of the player's team.

Costliest Relievers (as percentage of bullpen spending)

1. Jonathan Papelbon, 67 percent (10th, 81-81)

2. Brandon Lyon, 67 percent (12th, 55-107)

3. Heath Bell, 60 percent (11th, 69-93)

4. Matt Guerrier, 58 percent (4th, 86-76)

5. Huston Street, 56 percent (5th, 76-86)

6. Carlos Marmol, 55 percent (13th, 61-101)

The one thing that all six of these players have in common is that their teams did not qualify for the postseason. In fact, of the 12 relievers whose share of the bullpen payroll was higher than 40 percent, only two pitched for teams that qualified for the postseason (Ryan Madson, who was injured for the Reds, and Eric O'Flaherty, who pitched for the Braves).

The three best bullpens in the National League, at least according to ERA, all featured at least four players who earned at least 10 percent of the bullpen spending.

Here was the ratio of spending for those three teams, along with the Phillies

Highest Paid Reliever

Reds: 41 percent

Braves: 45 percent

Nationals: 34 percent

Phillies: 67 percent

2 Highest Paid Relievers

Reds: 63 percent

Braves: 61 percent

Nationals: 60 percent

Phillies: 84 percent

3 Highest Paid Relievers

Reds: 78 percent

Braves: 72 percent

Nationals: 78 percent

Phillies: 91 percent

Look at it another way: The median salary of the Phillies' six highest-paid relievers was $810,000. The median salary of the Nationals' six highest-paid relievers was $1.325 million, despite the fact that Washington spent half as much as the Phillies did on those six relievers.

Occupy Broad Street, anybody?

Everything else about the economics of bullpen construction varies greatly. The Reds spent 24 percent of their Opening Day payroll on their top six relievers and finished with the best bullpen in the majors. But the Braves only spent six percent of their Opening Day payroll on the bullpen and finished second. the Nationals spent 10 percent, the same as the Phillies.

Bullpen spending as percentage of Opening Day Payroll

1. Reds, 24 percent

2. Padres, 24 percent

3. Giants, 17 percent

4. Pirates, 17 percent

5. Mets, 16 percent

6. Brewers, 14 percent

7. Diamondbacks, 14 percent

8. Cubs, 12 percent

9. Astros, 12 percent

10. Rockies, 12 percent

11. Marlins, 11 percent

12. Phillies, 10 percent

13. Nationals, 10 percent

14. Dodgers, 8 percent

15. Cardinals, 6 percent

16. Braves, 6 percent

Long story short, the Phillies need to spend money on a couple of veteran arms this offseason, but spending money is no guarantee that things will get better.

II. Potential offseason acquisitions

The good news is that the Phillies could be one shrewd offseason acquisition and a couple of good breaks away from having a dominant bullpen. While 2012 was difficult to watch, the turmoil did force the club to call upon a number of young relievers who could contribute in the future. When all was said and done, the Phillies' bullpen led the National League in strikeouts-per-nine (10.05) and finished fifth in strikeouts-per-walk (2.62). Their biggest problem was home runs. They allowed an average of 1.01 per nine innings. They also allowed 32 percent of inherited runners to score, the third highest total in the NL.

Despite Papelbon's $12.5 million salary over the next three years, the Phillies should have plenty of flexibility to renovate their bullpen. No other relievers are signed to guaranteed deals, and only Antonio Bastardo is eligible for arbitration. Bastardo is the only player who, if healthy, is guaranteed a spot in next year's bullpen.

The biggest hole in this year's 'pen was the absence of a veteran strikeout arm who could handle the eighth inning in the event of injuries or regression by Bastardo and Mike Stutes, who were the only things standing in the way of a complete bullpen implosion in 2011. Jose Contreras was on the wrong side of 40 years old and was coming off of major elbow surgery. Chad Qualls had struggled the previous couple of seasons and was not a strikeout pitcher.

The market is less bountiful than it was last season. You have to assume that pitchers like Jonathan Broxton, Jose Valverde and, if healthy, Joakim Soria and Ryan Madson, will look first for situations where they have the opportunity to close. The most accomplished set-up man is probably Mike Adams, but there are some red flags there, specifically the fact that he was diagnosed with Thoracic Outlet Syndrome, which is the same condition that Antonio Bastardo had at one point, and the Phillies spent the first half of the season shooing off questions that the reliever's struggles might be due to a recurrence. Adams' average fastball velocity was down nearly two miles an hour this season compared to where it was in 2010. His strikeout rate also dipped. He will be 34 years old. That should be enough to convince the Phillies not to spend significant dollars on him. Rather, the best model might be the one employed the Rays every offseason. Instead of locking up a lot of money in one potential set-up man or closer, they target several players who have good arms but are not at the top of the market. This year, it was Joel Peralta and Fernando Rodney, who combined to earn about $3 million. Instead of Adams, the Phillies might be better of looking at teammate Koji Uehara. At 38 years old, he probably shouldn't be the marquee acquisition, but he posted a 1.75 ERA with sparkling strikeout and walk numbers in 37 appearances for the Rangers this season. He also made $4 million. On a one-year deal at a reasonable number he could be a low-risk, high-reward candidate.

The guy the Phillies could end up targeting is 33-year-old Brandon Lyon, who they had some interest in three years ago before he signed a three-year deal with the Astros. Lyon was traded to the Blue Jays this season and he pitched very well. He's one of the few guys available who you might feel reasonably safe spending $4 or $5 million on a two-year deal. After that, it is going to come down to good pro scouting and good luck. You can make an argument that all this bullpen needs is one trustworthy veteran who can throw strikes, avoid home runs, and miss a bat or two in runners on base. As we mentioned before, that probably means acquiring more than one pitcher with the potential to do so, because you have to account for the volatile nature of the position from year-to-year. I once heard Andrew Friedman say that the organization's plan was to sign several pitchers with the hope that one ends up stepping up. Last year, that guy was Kyle Farnsworth. A couple of years ago it was Joaquin Benoit. The Phillies have actually had some good luck with this in the past, specifically in 2008, when Chad Durbin and Brad Lidge both proved to be pivotal acquisitions. J.C. Romero, acquired the year before off waivers, and Rudy Seanez, acquired in the first week of the regular season, also provided important contributions at various points. So did Scott Eyre, who was acquired in August.

Last year, the Reds finished ninth in the NL with a 3.55 ERA. In the winter, they acquired Sean Marshall from the Cubs and signed Madson as a free agent. Madson, of course, went down with an injury. But Marshall stabilized the eighth inning while Aroldis Chapman blossomed into one of the most dominant closers in the game. The rest was all about the contributions of pre-existing younger players like Jose Arredondo, J.J. Hoover and Sam LeCure. They also added Jonathan Broxton in the middle of the season via trade. In 2009, Jack Taschner did not pan out, but Chan Ho Park did. Same goes for Danys Baez and Jose Contreras in 2010.

Those guys profile similarly to the kind of relievers available on this year's market. You wouldn't want to go into a season with Luis Ayala as your set-up man, but as an experienced veteran who can throw strikes and miss a bat every now and then? Same goes for Peralta and Uehara. Lefty Jeremy Affeldt might require more of a commitment, and he has thrown a lot of bullets in his career. But he is out there. Again, price is everything.

In Durbin, Park and Contreras the Phillies had a lot of luck signing a mediocre starter with the thought that he would eventually land in the bullpen. It wouldn't be a shock to see them try that route again, although the money in starting pitching is such that it would probably be a difficult task convincing somebody like Francisco Liriano or Scott Feldman to sign with the potential of ending up in the bullpen. The candidate would have to view the bullpen as a legitimate way to increase his future earning potential, and he would have to have a decent fastball with a good out-pitch. Eyeballing the field, I don't see any obvious candidates.

Again, the free agent crop is not overwhelming, so the Phillies might consider adding at least one piece via trade. This is all stuff that we'll delve further into as we draw nearer to the start of the offseason.

III. Birds in hand

Now, I know what you are thinking: Dave Murphy, I realize you've been covering this team for five years, and you're one hell of a guy, but you're telling me that the only thing this bullpen needs is a couple of low-cost veteran arms? What bullpen were you watching?

But if you really look at the current personnel and the strides they made this year, the potential is there. The Phillies just need a safety net for them.

Keep in mind, one of the commonalities among all the successful bullpens that we mentioned earlier is that they featured two or three young, homegrown relievers who, at some point, transformed from a talented but frustrating youngster to a legitimate power arm. It is impossible to predict when, or if, that transition will occur. It's a combination of getting used to the physical and mental grind of pitching in unpredictable situations, and trusting your stuff enough to pound the zone and not walk batters. If you watched the second half closely, you saw some of these young guys show signs of making that transition. The physical talent is there.

-Justin De Fratus: I'm starting off with him because I've considered him the top near-major-league-ready reliever in the system for the past couple of seasons. The elbow soreness that sidelined him for spring training and the first half of the season was a huge injury. If he is healthy, he might get the call-up instead of Mike Schwimer or Jake Diekman, and who knows what happens at that point. I know this: he has moxie to go with his mid-90's fastball. The question is consistent command and the continued development of his slider. He ended up pitching 10 2/3 innings, walking five against eight strikeouts. But he didn't allow a home runs, posted a very good ground ball rate, and stranded all four of his inherited runners while going 5-for-5 in hold situations.

-Jake Diekman: We've all seen the talent. It comes down to him not walking guys. Everything else you want to see is there, a good ground ball rate, 19-of-25 stranded runners, and strikeout stuff. The guy only allowed one home run in 27.1 innings. He still walks too many lefties to count on him as a lefty specialist (one out of every seven plate appearances resulted in a walk against them). But if you enter the season with him as the sixth reliever in your bullpen, you get a chance to work him into the flow while at the same time having a guy who, if he is forced to pitch in a pivotal spot (whether it is the middle innings or extra innings), can at least strike batters out (32 in 25 1/3 innings over his last 30 appearances of the season). One of the Phillies' problems this year was that, by May, that sixth man in the bullpen was often a guy who clearly did not have the stuff to give you a fighting chance (see interleague play).

-Jeremy Horst: Also known as Wilson Valdez's most important contribution to the team, Horst was one of the huge bonuses of the 2012 season, because he did look like a guy who could be used as a No. 1 lefty specialist. In 52 plate appearances against lefties, he struck out 17, walked four, and allowed just one extra base hits. He wasn't bad against righties either. In 31 1/3 innings overall, he struck out 40, walked 14, and allowed one home run while stranding 10-of-13 inherited runners. Attempted base-stealers were just 1-of-4 against him.

-Phillippe Aumont: He is still something of an unknown, but he clearly has a weapon that will make him a big league reliever in some capacity. That weapon, of course, is his nasty sinker. Like Diekman and De Fratus, he still must prove that he will not kill himself with walks in an high-pressure role (he walked 9 in 14 1/3 innings this year).

Along with Mike Stutes, who missed most of the season with a shoulder injury, the Phillies have five young relievers who have the stuff to make you think that they could at least hold their own if forced into pressure situations. Obviously, Stutes health is a big question mark. But keep in mind we are operating on the assumption that these young relievers will not be in a situation where they have to be set-up men. In an ideal world, the Phillies enter the year with Papelbon, Bastardo and two veterans as the top options for late-inning duty. The other spots in the bullpen are then doled out to three of the aforementioned five players, and Rich Dubee sees what develops. Inevitably, the young guys will be forced into situations where they will be tested, and as they earn or lose trust, their roles change accordingly.

Guys like Josh Lindblom, Michael Schwimer and B.J. Rosenberg all gave up way too many fly balls. And, by extension, home runs (although Schwimer's home run rate was not awful). They will all get plenty of opportunities in spring training, but I think you have to do a lot less projecting to get De Fratus, Horst, Diekman and Aumont into the 2013 bullpen.