(The full, nine-inning Playbook was originally published in spring 2020. It has been updated for 2025 where applicable.)
By now, you might be fancying yourself a fantasy baseball pro.
You’ve read all five Playbook innings, and perhaps have begun to craft your own cheat sheet for 2025. You’re feeling confident, fully trained for the proverbial marathon that’s ahead. But though the force is with you, young Skywalker, a Jedi yet you are not.
It’s not enough to know the basics of this grand game. No, we won’t stop until we’ve made you a perennial championship contender. After all, it might be fun to play fantasy sports, but isn’t winning ultimately the most fun?
So let’s take these important next steps with nine strategies to embrace — angles that will make you a more competitive player. Though they’re strategies that any experienced player might know, they’re also topics with which anyone could use a refresher course.
1. Wins, batting average and ERA are poor barometers of talent
Rotisserie baseball was spawned from the bubble gum card era, a time when television graphics included just “AVG-HR-RBI” for hitters and “W-L-ERA” for pitchers, and in a season when it was still possible for Steve Stone to win a Cy Young award, despite an ERA seven-tenths of a run higher than and a WAR (Wins Above Replacement) between 2-3 less than that of Mike Norris (depending on your source). Baseball analytics have come a long way and, though the majority of us are more educated players, the game hasn’t kept up quite as well with the times.
That’s not to say that wins, batting average and ERA have no place in fantasy baseball. Consider them a form of accounting for past outcomes, which isn’t an entirely unfair measure of success for our purposes, but rather one that accepts that baseball is a game of occasionally unlucky bounces.
From a future-analysis standpoint, however, the value of these categories stands at zero (or very close to it). The following examples exemplify the folly of chasing wins, batting average or ERA:
Wins: Logan Gilbert pitched the majors’ most innings, averaged the most innings per start (6.3), tied for the fourth-best quality start rate (66.7%) and had the 14th-best ERA among 58 qualifiers. His Statcast expected ERA (3.14) was also fourth best among that same group. Unfortunately, due to pitching for a Seattle Mariners team that averaged the ninth-fewest runs per game (4.17), Gilbert won only nine times and finished with a losing record (9-12). Conversely, Brayan Bello won five more games (14) than Gilbert, despite Bello sporting an ERA nearly a run and a quarter higher (4.49) as well as the majors’ 10th-worst Statcast xERA (4.56).
Batting average: Tyler Fitzgerald hit .280 across 314 at-bats as a rookie, 15 points higher than his career mark in the minors (.265) and six points higher than his mark across three years in college at Louisville (.274). It was backed by a .380 BABIP, the highest among hitters who came to the plate at least as often as he did (341 times). Statcast said that, based on Fitzgerald’s quality of contact, launch angles and batted-ball distribution, he should have hit only .227. That 53-point differential between numbers is the widest in that direction in 2024. On the opposite side, Juan Soto hit a healthy .288, but Statcast estimated that he should have hit .316, the second-highest such swing in the majors.
ERA: Ronel Blanco had the majors’ fourth-best qualified ERA (2.80), nearly three-quarters of a run beneath his career minor league ERA (3.54). But Statcast calculated that, based on the quality of contact as well as how much contact he allowed, he should have had an ERA of 4.00. That near-run-and-a-quarter differential was the widest in that direction among ERA-qualified pitchers. In contrast, Brandon Pfaadt had the majors’ widest ERA/xERA differential in the other direction, with Statcast suggesting he should have had a 3.78 mark to his actual number of 4.71.
Instead of weighting wins or ERA, use FIP (Fielding Independent Pitching score) or SIERA (Skill-Interactive ERA) or Statcast’s xERA. Simpler yet, trust the pitcher’s WHIP over his ERA, or weight his K/BB ratio more heavily.
For hitters, consider a player’s contact rate, line-drive rate or Statcast hard-contact rate rather than put stock in his batting average, at least if your league includes that category. From a hitting-skills evaluation standpoint, wOBA and Statcast metrics such as launch angle and exit velocity are better measures. (Worry not, we’ll dive deeper into those Statcast metrics in an upcoming Playbook.)
Returning to Fitzgerald’s example, as further evidence that his numbers are at risk of regression in 2025, Statcast reflected a 65-point wOBA-xwOBA (expected weighted on-base average) differential last season, easily the widest among qualified hitters in the “fluky” direction (.357 wOBA, .292 xwOBA). That casts serious doubt about him coming close to the kind of hot streak he had across a 17-game span in July and August last season during which he hit 11 home runs.
2. Buy low and sell high on the trade market
Trading was covered in the last Playbook installment, but this is a specific, critical angle to understand and exploit. Just as in the stock market, the (perceived) value of baseball players on the trade market varies depending on aspects such as their recent performance, health, role and potentially even the success of their team. To “buy low” means to attempt to trade for a player at a low — and preferably the lowest — point on his valuation curve, while “sell high” means to trade a player at his highest point, when the interest in acquiring his services has reached its peak.
Usually, the way to identify a “buy low” or “sell high” player is to seek those who have underperformed or vastly exceeded expectations, either for the season or in recent weeks. Some of the statistics cited above can help with this: comparing FIP (or SIERA) to ERA, comparing Statcast’s xERA to ERA, comparing Statcast’s hard-contact rate to home runs or comparing line-drive rate to batting average, just to name four. Essentially, you’re engaging in similar analysis to what you should do during draft-prep season, except using in-season data to extract hidden value (or identify overvalued players). You could compare the current year’s numbers to last year — or the past three years — if you wish, though I’d recommend examining skills-driven departments with that.
To extract successful examples from 2024, Pablo Lopez, the No. 7 starting pitcher and No. 28 player selected overall (on average) in the preseason, carried a 5.63 ERA through his 15th start of the season on June 18, the second-worst qualified number in the majors at the time. He was averaging only 8.5 fantasy points per game, well beneath the 13.8 he averaged in 2023 and the 12.3 he averaged from 2021 to 2023. Fantasy managers who noticed at the time that he had a significantly lower FIP — at 3.84, it was more than a run and one-third beneath his ERA — could have reaped the rewards of a Lopez trade. From June 23 (the date of his next start) forward, he scored the eighth-most fantasy points among pitchers, and averaged 14.6 per outing.
On the “sell high” side, Ranger Suarez, who had a stellar finish to 2022 and was a top-60 scoring starting pitcher in 2023, was the majors’ leading scorer in fantasy points through his first 15 starts of last season (through June 19). At the time, he was averaging nearly 10 more fantasy points per start than Lopez, not to mention had seen his roster rate in ESPN leagues soar past Lopez’s (97.8%-91.5%). It’s not outrageous to think that such a straight-up swap might’ve been entertained in many leagues. Suarez, however, posted a 6.67 ERA over his next five starts, missed a month due to lower-back soreness, and scored only 12 fantasy points across his final 12 outings.
Usually, fantasy managers who attempt the “buy low, sell high” strategy make a big mistake. They often attempt such a deal too early in the season, before their competitors’ opinions of players begin to significantly shift, or they’re too unrealistic in gauging the market for such candidates. Such miscalculations can turn off a prospective trade partner, often to the point that there’s no future hope of successfully executing the strategy.
In the above Lopez-Suarez example, June was deep enough into the year that the former’s fantasy managers might have been frustrated with Lopez’s early struggles, or concerned that they were the result of waning skills. There comes a point in every season when a fantasy team languishing in the middle of the standings (or worse) needs to make dramatic moves, making that an ideal time for Suarez’s manager to strike with a trade offer.
3. Stream starting pitchers
“Streaming,” or rostering a player for one day (or week, depending on your league’s lineup-locking format), only to release him the next for that day’s similar replacement, is an increasingly popular strategy in fantasy baseball, especially shallow mixed leagues and those that afford you the maximum opportunities to change a lineup. The idea is that in a league that weighs cumulative statistics — such as a points-based league where every player’s performance is boiled down to a single number, or a rotisserie league light on ratio categories such as batting average, ERA or WHIP — you want to maximize your number of player opportunities to accumulate such stats. This means trying to get an active game out of every one of your active lineup spots, every day, and in ESPN standard leagues, you get the benefit of changing your lineups each day.
Nowhere does streaming benefit a fantasy manager more than on the pitching side. Pitching statistics tend to be much more volatile than hitting statistics, and starting pitchers work significantly less often than hitters — generally once every five days, so keeping the same starting pitcher in your lineup for an extended period means getting generally one start (and maybe two) from him each week. Streaming starters in a daily league provides the opportunity to squeeze a start out of every pitching lineup spot every day, maximizing your chances at getting fantasy points or, in a roto league, wins and strikeouts. (In the latter, however, bear in mind that this strategy can come at the expense of your ERA and WHIP because most pitchers readily available on a league’s free agent list are less talented than those already rostered.)
Again, the format of your league comes into play, as does whether your league limits the number of transactions or starts you’re allowed in a given week, but the closer your league is to fully points-based, daily transactions and no limits on moves or starts, the more the strategy of streaming starters benefits you. Only 18% of all starts last season resulted in a negative point total in ESPN standard points leagues, which was less than the rate of starts that were worth 19-plus points (19%), giving you good odds of a strong return on the strategy (albeit with a hint of risk).
In a weekly league, incidentally, streaming starters is also a valid strategy, but there it’s often referred to as loading up on “two-start” pitchers in a week, picking those set to start early enough in the week that they’d squeeze in a second turn before Sunday’s games conclude.
As an additional piece of advice regarding ESPN standard leagues: blow past the weekly starts cap, if your league has one. This means that if your league limits you to 14 starts in a week (an average of two per day), then on the day that you expect to reach your maximum for that week, you should stream everywhere you can. Our cap rules take effect only at the beginning of a day, but they don’t lock you out on the day you reach or exceed said cap, meaning that a clever manager could enter a Sunday with 13 starts already in the tank, then stream six starters on Sunday for a total of 19. (Incidentally, one reason to argue this should be allowed is that, in the event of a team exceeding the cap, it would be impossible to tell which pitcher was responsible for the final start under said cap — would it be the one whose game started first, whose game became an official game first, or the one whose game finished first?)
4. Volume is king, especially in a points league
Tying to the previous point about streaming, you want to squeeze as many opportunities to generate statistics out of your players as possible. Besides manipulating fantasy lineups, there are other ways to do this. Drafting or acquiring hitters from more productive offenses, hitters who hit earlier in the lineup, hitters whose teams have more favorable daily or weekly matchups or pitchers who can claim the same on that side. Returning to the previous topic about wins, too, in those leagues you can also accumulate pitchers who work for the most successful teams.
Seeking players from productive offenses is self-explanatory: The more runs a team scores, the more runs and RBIs it will spread up and down the lineup. For example, of the 12 hitters to score at least 100 runs last season, eight played for teams that ranked among the top nine in runs per game, and five of the top six in that category played for one of the top four offenses in runs per game. On the pitching side, eight of the 11 pitchers to win at least 15 games last season pitched for teams that averaged at least 4.50 runs per game.
It’s the lineup advantage that’s often overlooked in fantasy, but it’s a relevant one. Coupling this somewhat with the previous point, the more times teams score, the more times they cycle through their lineup. Therefore, the higher a hitter bats in the lineup, the more opportunities he’ll get to hit in a given game, and during a season, that can amount to some noticeable volume advantages. The chart below breaks down the average number of plate appearances by each of the nine lineup spots for the 2024 season, with the totals by the majors’ best and worst from each spot.
The difference in plate appearances between each of the nine lineup spots is roughly 18 for each successive slot we move down. Though 18 PAs might not seem like much over a 162-game schedule, it represents an opportunity advantage. The 123-PA difference between Nos. 1 and 8 hitters, meanwhile, is massive, which is why it was such a big deal when the St. Louis Cardinals decided in June to move Masyn Winn, until then a seven-eight-nine hitter, to the leadoff spot, where he started in 98 of the team’s final 107 games. Winn averaged only 3.58 plate appearances per start before the move, compared with 4.58 after it, and he averaged 0.38 more fantasy points per game from June 1 than before that date, despite his OPS being 88 points lower after being moved up.
Hitters similarly slated for, or stuck in, bottom-third-in-the-order roles are at a significant opportunity disadvantage. That’s increasingly true when the competitive levels of the offenses are unequal — note the 134-PA difference between the best team’s No. 3 and worst team’s No. 7 hitter, an even wider margin than the aforementioned one between an average team’s Nos. 1 and 8 hitters.
Daily or weekly matchups also influence opportunities. Hitters set for a week of games at nothing but hitter-friendly ballparks are likely to see their teams score more runs, meaning more trips to the plate for the offense and more runs/RBIs up and down the lineup. These are just as important to weigh — if not more so — in your lineup-setting as the players’ roles.
5. Spring training stats don’t really matter
I get the lure of these silly numbers. Assuming that it starts on time, spring training baseball represents the first moments of competitive, recordable game action in four months, and as stats-obsessed baseball fans, we crave new statistics. By March 1, we’re ready to dive into these new numbers, often to the point we get carried away with players’ spring performances and make unnecessary, and almost always unadvisable, adjustments to our cheat sheets.
Here are the problems with spring statistics: They’re drawn off a minuscule, roughly one-month or 30-day sample, and one that, unlike during the regular season, features prominent players playing only fractions of the games or often not many of them at all (especially in the early weeks). They’re also played in states where weather conditions are quite different from what the same teams will see during the regular season, as Cactus League games in Arizona are played at 1,000-plus-foot elevations, often in humidity, pumping up the offensive numbers, while Grapefruit League games in Florida are played at or near sea level, in often larger ballparks that favor pitchers. And, perhaps most importantly, games in both states are played against far more variable levels of competition than what we’d see during the regular season, as expanded rosters mean certain players could capitalize from facing nothing but inexperienced, Class A competition for a good number of their at-bats or innings.
Remember when Jordan Hicks struck out 28 spring batters, second most in the majors, with a 2.65 ERA and a .167 BAA? You should, considering it happened last year.
Nowhere is the absurdity of spring statistics more apparent than in saves. Over the past six full spring trainings (2018-19 and ’21-23) — we’ll consider 2022’s spring training “full” for argument’s sake, despite its lockout-influenced abbreviation — 18 pitchers had a three-save spring: Jonathan Aro, Ryan Brasier, Cody Carroll, Dietrich Enns, Caleb Freeman, Justin Hancock, Eric Hanhold, Nolan Hoffman, Gavin Hollowell, Andrew Kittredge, Joe La Sorsa, Jose Leclerc, Dominic Leone, Lucas Long, Riley O’Brien, Evan Sisk, James Teague and Hunter Wood. These pitchers saved four big league games during the regular seasons that followed, all four of them recorded by Leclerc in 2023. The reason is that big league teams tend to lift their veteran players from spring contests early, usually by the sixth inning, meaning that it’s those same Class A-caliber players who are often left to pitch the eighth and ninth, not to mention that teams prefer to get their real closers work against real big league hitters earlier in the game if they can. You can expect to see Emmanuel Clase probably pitching the fifth, not the ninth, for the Cleveland Guardians during spring training.
If there is a spring-stats angle worth exploiting, it’s less-proven and/or younger types who have something to prove or a job to claim. Wyatt Langford‘s .365/.423/.714 hitting line, six home runs and a league-leading 20 RBIs were instrumental in earning him a place in the Texas Rangers‘ Opening Day lineup. Luis Gil, the eventual AL Rookie of the Year, earned the New York Yankees‘ fifth starter role out of spring training in large part to his 2.87 ERA and 23 strikeouts during Grapefruit League play.
Another statistical factor to consider is whether a player’s strikeout or walk rate has noticeably shifted from previous seasons, such as when Jack Flaherty, who added velocity to all three of his most prominent pitches (four-seam fastball, slider, knuckle curve) during spring training, struck out 26 of the 70 spring batters he faced while walking only four, looking much more like the pitcher who finished 17th overall in fantasy points during his career year of 2019. He went on to total his most single-season wins (13), and most starts (28), innings pitched (162), strikeouts (194) and fantasy points (405) since that 2019 season.
For a final note on spring stats, if you’re insistent on placing any stock in them, a wise move is to peruse Baseball Reference’s “strength of competition” number, which in recent seasons the site has provided as an additional column beside their spring statistics. If a player’s level of competition faced falls in a Class A-level tier described by their metric, his stat line is much less relevant than one who faced a great deal of Triple-A or MLB talent.
6. Go bargain-shopping for saves
Speaking of those saves, though I’ll stop considerably short of the blanket “don’t pay for saves” declaration, there’s still a lot of merit to the strategy. Saves are typically the easiest of the 10 traditional roto statistics to find on the free agent list, or at worst, at a discount price on the trade market.
To that point, 31% of the majors’ total saves last season came from pitchers who were not drafted in ESPN leagues (specifically both outside the top 300 in ADP as well as selected in fewer than 5% of drafts), including Kirby Yates, the No. 3 scoring relief pitcher, Trevor Megill, who saved 21 games, and 15-save performers James McArthur, Chad Green and Michael Kopech. Note that the percentage of saves secured by the fantasy baseball “draftable” pool has increased significantly in back-to-back seasons, but we’re still talking about nearly one-third of the saves market in 2024 having been widely available via free agency.
Again, though, I hesitate to use the word “DON’T” when it comes to investing in saves because a lackadaisical approach to the category is another mistake. That the percentage of total saves amassed by the draftable pool has changed in back-to-back years makes it foolish to set an ironclad rule regarding the category..
Especially the deeper the player pool your league uses — think AL- and NL-only — the more likely it will be that managers will roster players who might sniff a save chance, meaning the free agent list won’t be nearly as populated with prospective save-getters. Worse yet, trade partners are much less likely to trade a pitcher once he’s handed his team’s closer role, especially with the recent, growing tendency of major league teams shifting to closer-by-committee strategies.
7. Resist recency bias
Fantasy managers, and not just baseball but in all sports, tend to find chasing yesterday’s statistics irresistible. A hitter slugs three home runs in one game, and he becomes the hottest commodity by the next morning. The same goes for the pitcher who just threw a no-hitter. But even for the more experienced players, who aren’t fooled by a one-night outburst, some get fooled by lengthier stretches, albeit over still-small samples of time, of player success. If you see the phrase “small sample size” bandied about on these pages, this is what we’re cautioning against.
Recency bias can reveal itself with the out-of-nowhere wonder, such as the aforementioned Tyler Fitzgerald, who saw his ESPN roster rate pass the 50% threshold for the first time the day after his 13-homers-in-26 games hot streak in mid-August, in time for him to bat .230/.281/.311 with one home run over his final 36 games, or Kyle Tucker, whose 79-game absence due to a fracture in his shin last season was more of a fluky circumstance than something likely to linger. Tucker’s metrics were otherwise excellent, very much in line with his previous seasons, and there’s little reason to expect he’s any less of a first-round candidate than he was a year ago.
Another area where recency bias traps even the best of us is early in the regular season, when the freshness of new statistics lures us and causes us to believe outcomes that haven’t crystallized. Connor Joe, through 48 days of the 2024 season, might have convinced his fantasy managers that he was destined for a breakthrough season — and his reaching a 40% roster rate in ESPN leagues around then hinted he did for a good number of managers — thanks to his .293/.370/.520 hitting rates. Those same managers might have begun to consider swapping him in for Paul Goldschmidt, whose .206/.291/.300 rates and 32.6% strikeout rate at the time made many wonder if he was done as a fantasy contributor.
Be patient, especially early in the year, because baseball tends to even out over longer stretches.
8. Resist the rookie hype
Who doesn’t want to be the first person to discover the next big thing? The lure of rookies has taken on greater weight lately, with such recent standouts as Paul Skenes, whose 1.96 ERA last season was the lowest by any rookie pitcher with at least 20 starts during the live ball era, or Corbin Carroll, who became the first rookie to have at least 25 home runs and 50 stolen bases (2023).
Additionally, Jackson Chourio, Michael Harris II, Gunnar Henderson, Jackson Merrill, Julio Rodriguez, Spencer Strider and Bobby Witt Jr. captured many a headline as rookies in recent years, while the graduation of several of the game’s top prospects to the major leagues during each of the past two seasons (Elly De La Cruz and Jasson Dominguez in 2023, James Wood and Colton Cowser in 2024) fueled a perception that rookies are the “name of the game” nowadays.
The problem with rookie-chasing, though, is that for every Skenes or Carroll, there’s a Jackson Holliday, a can’t-miss prospect who, well, missed, or an Evan Carter, Jordan Lawlar or Brooks Lee, prospects who either got hurt, disappointed or didn’t get the call in 2024. Yes, rookies and younger players have had greater odds of success in recent years than at any other time this century, but it’s still important not to overrate each season’s freshman class, especially not at the expense of ignoring a more seasoned, yet still young big leaguer who has yet to reach his peak at the MLB level.
9. Have patience through streaks — if the player’s skill set warrants
To repeat, baseball is an unpredictable game, full of ups and downs that even themselves out over a 162-game schedule. Narrowing the scope, however, there is a subset of baseball players who are even more subject to peaks and valleys than others, and it’s with these that you must be the most patient.
On the hitting side, big sluggers who hit a lot of home runs at the expense of many strikeouts, often referred to as “three true outcomes” players because of the high likelihood that the outcomes of their plate appearances will be either a home run, strikeout or walk, represent the streakiest.
Kyle Schwarber saw 49.3% of his plate appearances in 2024 end in either a home run, strikeout or walk, the second-highest rate among qualified hitters (Aaron Judge led, with 51.4%). Schwarber had two separate stretches of at least 20 games with a sub-.200 batting average and sub-.700 OPS, but also three spans of 20-plus games in which he hit at least .300 with a 1.000 OPS. He has endured maddening hot and cold spells, most notably having outrageous June numbers, as his 61 home runs and .934 OPS in that month are easily his best in any of the season’s individual six months.
Though one could attempt to use a hitter like Schwarber as a buy-low or sell-high candidate based on where he’s at on the performance curve, it’s a poor idea to acquire him at his high points or sour on him at his lowest. Such players are best utilized over lengthier time frames, when their fluctuations have more time to even out, as it’s difficult to predict their next hot or cold streaks.
On the pitching side, truly “streaky” types tend to be those who have an incomplete ingredient in their games. It could be the lack of blazing, raw stuff, perhaps shaky control, or maybe a durability question. Carlos Rodon is an excellent example from 2024, when he had an up-and-down season. He began the season by going 9-2 with a 2.93 ERA over his first 14 starts, only to post a ghastly 9.67 ERA across his next six outings. Rodon also finished the season on a high note, posting a 2.92 ERA and 28.2% strikeout rate over his final nine starts.
In Rodon’s example, though patience remains a worthy strategy, remember that the greater degree of pitching volatility — especially for a pitcher with the number of durability questions he has — supports a strategy of greater turnover. The takeaway is not to distrust the streaky pitcher, but to be more prepared to move on when opportunities present themselves, or make greater effort to find replacements to fill in the gaps between their cold spells.
Always consider the nature of the player and what his skills tell you. Returning to Goldschmidt’s example, keep in mind that, after his aforementioned slow start to 2024, he bounced back with .260/.307/.460 rates and 19 home runs over his final 114 games; his 230 fantasy points in that time were 10th best among first basemen. His overall level of career consistency warranted greater patience than with an average player, and his strong finish was representative of that.
Now, you’ve got the skills to be a competitive, well-educated fantasy baseball manager, so it’s time to shift our focus to prepare you for the upcoming season. In the next edition of the Playbook, we will examine the shifting trends in today’s baseball game. Stay tuned!