Bullpen Strategy: Holds, Middle Relievers and Ratios

By 4Baseball Staff · November 5, 2026 · 11 min read

When relievers help a fantasy baseball team and when they hurt it: how saves, holds, ERA and WHIP interact, worked examples of ratio math, and a practical way to build a bullpen in roto and points leagues.

Relievers are the strangest position in fantasy baseball. A closer can win a category by himself, a middle reliever can sink your ERA and WHIP in a single evening, and every manager learns that the closer drafted in March often does not hold the job in August. Some fantasy players spend their top picks on saves. Others refuse to spend a dollar on them and still win. Both groups are sometimes right, because the value of a reliever depends on your league's scoring far more than on the pitcher himself.

This guide is about that dependence. It explains what each relief statistic is, shows with arithmetic how a reliever changes your pitching ratios, uses the historical record to show how scarce saves have become, and then turns the math into a set of strategies you can apply to a rotisserie league, a head-to-head league or a points league. It assumes you know the basics of a draft; if you want the foundation first, read how to win a snake draft and how fantasy rankings are built.

A note about scope. Nothing here predicts any player's performance, and no player is recommended. The goal is to make you a better judge of a reliever's value in your own format, so the advice is about structure, not names.

The statistics that define a reliever

Start with the definitions, because leagues use them loosely.

Save. A pitcher earns a save when he finishes a game his team wins, is not the winning pitcher, and meets one of three conditions: he enters with a lead of no more than three runs and pitches at least one inning; he enters with the tying run on base, at bat or on deck; or he pitches at least three effective innings. In practice, most saves come from the ninth inning with a one-, two- or three-run lead.

Hold. A hold is an unofficial statistic: a reliever who enters in a save situation, records at least one out, leaves with the lead intact, and does not finish the game. It was invented to credit the setup men, and it is tracked by fantasy sites and statistics services but not by the official scorer.

Win. Relievers can win games, but the credit is random. A reliever who gives up the lead and sees his team retake it can be charged with a blown save and still collect the win.

ERA and WHIP. Earned run average is nine times earned runs divided by innings pitched. WHIP is walks plus hits divided by innings pitched. Both are rate statistics, which means they do not accumulate. They average.

Strikeouts. A counting statistic, but one where relievers are efficient: they pitch few innings and often strike out many batters per inning.

Most leagues use some combination of these. The rest of this guide treats three common setups: saves-only leagues, saves-plus-holds leagues and points leagues.

How a reliever changes your ratios

The rate statistics are where relievers cost and help the most, so it is worth doing the arithmetic.

Suppose your pitchers have thrown 1,200 innings with a 3.80 ERA. That is 506.7 earned runs (3.80 times 1,200 divided by 9). Now add a reliever who throws 60 innings with a 2.00 ERA, which is 13.3 earned runs. Your new total is 520 earned runs in 1,260 innings, and your ERA becomes 3.71. A very good reliever for a full season has improved your team ERA by about 0.09.

The same reliever with a 1.00 WHIP pulls a 1.25 team WHIP (1,500 baserunners in 1,200 innings) down to 1.24 (1,560 in 1,260). The gain is a hundredth of a point. The point of the example is that ratio help from a reliever is real but small, because 60 innings is only a few percent of a staff's workload.

Now the downside. Take a team 50 innings into the season with a 3.80 ERA, which is 21.1 earned runs. One reliever allows four earned runs in one inning of work. Your team now has 25.1 earned runs in 51 innings, and your ERA is 4.43. One bad night added 0.63 to the ratio.

Do the same arithmetic with 1,000 innings behind you and the same four earned runs raise the ERA by only about 0.03. The lesson is that ratio damage depends on when it happens. A bad outing in April costs a lot and fades by July, and a bad outing in September costs a little and cannot be undone. That has a practical consequence: a risky reliever does more harm to your ratios early in the season than late, because a small sample moves further.

Why saves are scarce, and getting scarcer

Saves feel like a fixed pool, and in total they are. The league recorded 977 saves in 1985, 1,006 in 1995, 1,254 in 2005, 1,292 in 2015 and 1,201 in 2025, according to the historical record. What changes is how they are distributed. The database shows the number of pitchers with at least 30 saves in a season:

The change from 2015 to 2025 is the headline. The total number of saves barely moved, yet the number of pitchers who reached 30 fell from 21 to 8. The saves are being spread across more bullpens, more committees and more pitchers who hold the job for part of a year. The 2025 leader had 42 saves, and the next two had 40 and 38.

For a fantasy manager, this has two effects. Elite saves totals have become scarcer, which makes the few pitchers who still accumulate them more valuable in a saves-only league. And the average closer is less reliable, which means that a team that waits on saves has a larger pool of mediocre options and a more volatile one.

It also helps to see how long the very best careers last. The historical record shows Mariano Rivera at 652 saves, Trevor Hoffman at 601, Lee Smith at 478, Kenley Jansen at 476 and Craig Kimbrel at 440. Those careers were built by pitchers who held a job for a decade or more. The pool of pitchers who do that is small.

Evaluating a closer

A good closer in fantasy is not the one with the best earned run average. It is the one most likely to be handed the lead in the ninth inning. Four things matter most.

1. Role security. Who else in the bullpen has the stuff to take the job? A closer with an obvious backup is one bad week from losing it.
2. The team's chances of leading. Saves require a lead. A closer on a team that wins 95 games gets more chances than one on a team that wins 65.
3. The manager. Some managers name a closer and stick with him. Others use the best available reliever against the highest-leverage batters. The manager's habits are the best predictor of how rigid the role is.
4. The pitcher's skill. Strikeouts, walks and ground-ball tendencies matter, because a closer who is bad enough will eventually be replaced even by a patient manager.

Notice that skill is last. Fantasy managers tend to overrate it and underrate role. A mediocre pitcher with a secure job will beat a brilliant one in a committee.

Middle relievers and the value of holds

If your league counts holds, or saves plus holds as a single category, the middle reliever becomes a real asset. Holds are more plentiful than saves, since every team uses several setup pitchers, and they are less tied to a single job. An effective setup man may accumulate as many holds as a closer does saves, and he costs far less at the draft.

Three points guide a holds-based strategy.

The best middle relievers are not always the ones with the title. Setup men are chosen by the manager's judgment of leverage. A pitcher who gets the seventh and eighth innings of close games for a winning team will collect holds, and one who pitches the same innings for a weak team will not get the chances. Holds are a product of both role and team success.

Strikeouts are the stable part of the line. A reliever who throws 60 innings with a high strikeout rate gives you a usable contribution in a counting category regardless of how many holds he gets. When you evaluate setup men, look at strikeouts per nine innings, which is nine times strikeouts divided by innings, along with walks per nine, rather than saves or holds from last season.

Holds are noisier than saves. Because the statistic depends on a save situation in the sixth, seventh or eighth inning and a manager's choice to use a particular arm, a reliever's hold total can vary a lot from one year to the next without any change in his pitching. Treat last year's holds as a weak signal.

In a saves-plus-holds format, treat closers and top setup men as roughly interchangeable for the category and choose on price, skill and role. In a saves-only format, middle relievers have little value except as a source of strikeouts and ratio help, or as a future closer if you can spot one early.

Four ways to build a bullpen

There is no universal best approach, but four have emerged. Each suits a different league.

Spend early on one or two elite closers. This works in saves-only leagues where closers are scarce. You pay a premium in draft position for a pitcher who will give you a dependable share of the category. The danger is that closers get hurt or lose the job, and the premium is wasted.

Take closers in the middle rounds and manage the position all year. This is the common approach. You draft two or three pitchers with plausible roles, watch usage, and replace failures from the free-agent pool. It rewards attention and punishes inattention.

Punt saves. Some managers ignore the category and put their picks into hitters and starting pitchers. In a 12-team rotisserie league, you give up the points in the saves category, but you gain depth elsewhere. The approach works when the other categories offer cheaper points than saves do, and it works poorly in leagues where saves count for extra or where you lack the depth to make up the difference.

Stream and rotate. Instead of drafting a stable closer, pick up a reliever who has a hot hand, a new role or a favorable matchup each week. This approach fits leagues with weekly transactions and low waiver costs, and it requires monitoring bullpen news daily.

The right choice depends on three features of your league: the number of teams, the transaction rules and whether saves, holds or both are scored. A 15-team league, like those described in expert and deep league play, has fewer useful relievers on the free-agent list, so drafting them matters more. A 10-team league with a deep free-agent pool can stream relievers or punt more safely.

Roster slots, innings and the rules that shape value

Your roster rules change relief value more than most people expect. A league with a dedicated relief-pitcher slot rewards you for carrying relievers, since the slot cannot be used by anyone else. A league where all pitchers share flexible slots makes every reliever compete with every starter for a spot. The guide to lineup slots and position eligibility explains how those slots are defined.

Innings limits matter in a different way. If your league caps total innings or starts, a reliever uses very few innings and contributes ratios and strikeouts almost for free. A reliever with a 2.00 ERA in 60 innings does not eat into your cap the way a starter with a 3.50 ERA in 180 does. In capped leagues, elite relievers are efficient, and the standard advice to avoid paying for relievers becomes weaker.

Minimum-inning rules work in the other direction. Some leagues require a minimum number of innings to qualify for the ratio categories. A team that has punted pitching can fall short of the minimum, and in that case relievers help only by supplying innings. Read the constitution of your league before you decide how to spend.

Points leagues are a different game

In points leagues, the calculation is simpler. Each pitcher earns points for innings, strikeouts, saves, holds and wins, and loses points for earned runs and hits. A reliever's value is the sum of those.

Because relievers throw fewer innings, they accumulate fewer points than starters unless the league awards a lot for saves. Many points leagues give a save a value similar to a win. Check how many points a save is worth compared with a strikeout, and compare the weekly total of a closer with the weekly total of a starter who pitches twice. In many formats the starter scores more, so relievers become a depth play rather than a core one.

Mistakes to avoid

• Paying for last year's saves. Saves are partly a product of role and team success, and both change. Buy the role going forward.
• Ignoring bullpen news. Role changes are announced in postgame comments and reported by beat writers. A reliever who loses a job on Tuesday should be dropped by Wednesday.
• Hoarding closers. A roster with five closers has no room for hitters. Fill roster spots by their expected return.
• Forgetting ratio exposure. A weak reliever can hurt more than a good one helps, especially early in the season. Use your bench and your reserve slots to avoid starting a pitcher you do not trust in a week where you are protecting a ratio lead.
• Forgetting to check your league settings. The guidance for a saves-only roto league can be wrong in a holds league.

Common questions

Are relievers worth drafting early? Only in formats where the category is scarce and weighted heavily, such as a saves-only league. In most formats, the middle rounds are a better place to buy.

How much can a reliever help my ERA? Less than most managers think. In the example above, a full season of elite relief lowered a team ERA by about 0.09, while one bad April inning raised it by more than half a run.

Should I carry a setup man in a saves-only league? Only as a future closer or as a source of strikeouts and ratios, and he costs a roster spot.

What is the difference between a closer and a setup man? The closer pitches the ninth inning in save situations. The setup man pitches the seventh or eighth, often against the toughest part of the opposing order.

How do I know when a closer is about to lose the job? Watch for a manager's reluctance to use him in close games, repeated blown saves, and a rising walk rate. Role changes usually have warning signs, but not always a long one.

Where to go next

To see who leads the league in saves and strikeouts, use the site's leaders pages, and for the broader approach to building a team, see the fantasy baseball tools and the fantasy baseball topic page. If you need a refresher on how to judge a pitcher's usage in a deep league, read the article on expert and deep league formats, and for hitters in platoons, evaluating platoon and two-way players.