It’s extension season!

Spring Training provides a perfect environment to explore long-term pacts, so it isn’t surprising to see some of the conversations result in new contracts.

What is surprising is what these deals look like — and how early they’re happening — because they’re quite different than the extensions of the past.

Teams are handing out larger contracts to pre-arbitration players than ever before. Only 15 pre-arbitration contracts have guaranteed at least $80m, and six of those deals have been signed in the last calendar year. And similarly, eight of the 19 extensions for players with fewer than 75 days of Major League service have come in the past two years.

Subscribe now

The challenge with signing deals earlier is straightforward: there’s less information to work with. Minor League performance can tell us a lot about what a player will be in the future, but it’s no substitute for playing in the big leagues. There we get a better understanding of how skills will translate, how the league will respond, and how the player will adjust and adapt.

At the same time, the cost of these deals suggests that teams are pricing players near their best case outcomes.

Teams are making same type of bets they’ve always made — but they’re making them earlier, with less certainty, and less room for error.

How did we get here?

The practice of giving multi-year agreements to players prior to arbitration dates back to Jon Hart and the Cleveland Guardians in 1993 and 1994. From there, teams began copying and riffing on the strategy, leading to a new category of contract that hadn’t previously existed.

Around 2008, the strategy gained more steam. That year alone, five pre-arb players signed extensions guaranteeing $20 million or more, and two contracts set new records in their respective markets: Ryan Braun’s 8-year, $45 million deal with Milwaukee and Evan Longoria’s 6-year, $17.5 million deal with the Rays. Braun’s deal was the largest for a player with fewer than three years of service, while Longoria’s deal was even more ground-breaking, signed after only seven days in the Major Leagues. No player had ever signed an extension with under a year of service, let alone fewer than 10 days.

Over the past 15 years, teams have signed nearly 70 pre-arbitration extensions with position players.1 When we look at those deals, the initial thing that jumps out is how team-friendly they started out. Over the first five years, there were massively valuable deals for players like Paul Goldschmidt, Jose Altuve, Anthony Rizzo, and Salvador Perez. Deals that traded low guarantees to buy out free agent years and retained flexibility in the form multiple club options. They weren’t all hits — and there was some fortuitous timing in signing players like Goldschmidt and Altuve immediately before they showed up on MVP ballots — but even the worst deals were often closer to neutral than anything.

That’s partially because the dollar figures were so low. From 2012-2016, the median guarantee for extensions was just $28 million. It’s difficult to make bad investments at that level.

The other reason these deals favored teams is they were going to relatively experienced players within the pre-arbitration population. The average service time for extensions signed over the same five-year stretch was more than a year and a half.

Signing players with established track records to low guarantees is a simple, and rather obvious recipe for good contracts (from the team’s point of view).

But there’s a reason they were able to do this. They were signing players who had made very little money to date. We can see that when we look at the amateur signing bonuses for extended players.2

The average signing bonus from 2012-2016 was a little over $450,000; the median was just $90,000 and over half of the extensions were players who signed for less than $100,000. Even adjusting for inflation or league-wide spending, that’s radically different from the last five years where the average signing bonus is more than $2 million and 65% of extensions are with players who signed for at least $1 million.

Earlier extensions often targeted players who hadn’t earned much, making the tradeoff between long-term upside and immediate security straightforward. Rising signing bonuses and structures like the pre-arbitration bonus pool have changed things by putting money in player’s pockets much faster. Players have far more leverage in negotiations.

But MLB is a copycat league, and as teams, players, and agents watched these extensions play out, they saw how valuable they could be. Players still wanted financial stability, only now they demanded higher dollars and earlier commitments to give away upside. Teams saw enough surplus value to be had that they were willing to move on both fronts.3

They continued to push the envelope. The best deals were valuable enough to make up for the losers. Christian Yelich, Jose Ramirez, Ketel Marte, Ronald Acuña Jr. — these were all examples of clubs gaining valuable control of star level players through pre-arb extensions.

By the beginning of 2021, the market had little resemblance to the extensions from ten years prior. Fernando Tatis Jr. and Wander Franco’s deals set new markers for length and guarantee, with Tatis’s deal essentially a full-blown free agent contract signed before he began arbitration. Those were followed by more complex extensions for Julio Rodriguez and Bobby Witt Jr., which layered in player opt-outs to preserve upside if they performed well or if the free agent market moved.

Players have demanded more, and teams have largely acquiesced, continuing to increase the dollar amounts and move the commitments earlier — bringing us to the present.

Are the deals being signed today good bets?

The real question is whether the deals being signed today are actually smart investments.

Generally speaking, I’d say no.

The biggest reason is that the baseline contracts for Major League players are incredibly hard to beat. Teams effectively control players for six or seven years at well-below market rates. Players make roughly the league minimum those first few years before entering a performance-based arbitration system that’s seen little growth over time. And if that isn’t enough, the salaries aren’t guaranteed — teams have the right to walk away each year.4

The tradeoff for guaranteeing early years is supposed to be access to free agent seasons at a discount. But recent deals for Konnor Griffin and Colt Emerson imply free agent AAVs of $33 million and $28 million respectively. Those salaries would rank among the top 25-30 players in the league today, and they’re guaranteed six years out with no Major League track record.

Anyone who plays fantasy baseball knows that predicting player performance is tricky. Doing it years in advance, without baseline major league data, adds another layer of difficulty.

The actual performance of these contracts has not been very good, either. It’s a small sample, but after a strong start with deals for Longoria, Moore, and Perez, the extensions for players with less than half a year of service have been underwhelming.

A number of deals have failed, and the early returns on many others aren’t promising.

Even the best looking deals are complicated to judge. Take Corbin Carroll. He has played about as well as possible since signing his extension, but he’s still years away from free agency. Every bit of production he’s provided Arizona would have happened without the extension (at a lower cost).

And if the deals do succeed, there’s still a common pitfall teams can’t seem to avoid. Players who outperform their initial extensions are often rewarded with worse follow-up contracts, cutting into the surplus value that was created. It’s human nature. Teams want to treat their best players fairly, and it’s hard to walk away from the table after a win. Unfortunately, that human nature makes it harder for teams to capture the upside they’re betting on in the first place.

I understand why teams are signing these deals. I do.

I just don’t think they’re valuing their optionality or the information they’re sacrificing nearly enough.

Things can change quickly. Players’ ability levels, the competitive environment, the needs of the club, and the rules of the game may all look very different seven or eight years from now. Unless a team is getting a significant discount on future control, there’s value in waiting and making a more-informed decision later — even if it ends up costing more.

Of course, that assumes the opportunity exists in the future. Extension negotiations aren’t guaranteed to remain on the table. If teams believe they only have one window to act, the decision is more challenging.

For small market clubs, these deals are understandable. They don’t have access to high-end free agents, especially not on shorter term deals. And to compete, they have to take targeted risks and find ways to be comparatively different.5

It’s less clear that the same logic applies to larger market teams. They can afford to gather information and pay for more certain outcomes. The risk-reward calculus doesn’t make as much sense for them.6

The prevailing thought seems to be that early-career extensions are always good decisions for the team. When that meant minimal commitments to relatively established players, that was fair to say. But that’s not the reality anymore.

It’s early.

Major League Baseball’s regular season, which began last week, won’t wrap up until the end of September. Roughly 95% of the season remains.

Despite the allure and prevalence of #HotTake culture, teams and executives generally ignore what happens in the first week.

The assumption is that it’s too early for anything to matter.

But is that actually true? Wins don’t count any less in March and April, and early contests offer the first real competitive data in months — data that could shift a team’s outlook.

To answer that, I looked at how much the first few games affect teams’ playoff odds.1

Subscribe now

The answer, on average, is not that much.

Over the last ten years, teams see their playoff odds change by 4.4% after the first week. If we extend the window to two weeks, the change moves to 6.2%.

Those aren’t the type of numbers that demand action. But remember, those are only averages.

For a meaningful number of teams, the landscape looks considerably different after only a few games. Nearly 12% of clubs see their odds move by more than 10% after a week. After two weeks, that jumps to more than one in five, with some odds moving as much as 20-30% over those timeframes.2

Most teams — the average teams — can stand to be patient. For teams that fall closer to the extremes, it’s important to examine early-season results. Has the team been especially lucky or unlucky? Did we improperly assess the talent of the team? How much of the movement is driven by the rest of the league’s play?

And by far the most important question: what are we going to do?

Because these shifts are meaningful. To put them in context: major trade deadline acquisitions improve playoff odds by 10-15%, and cost teams tens to hundreds of millions in surplus value.

For teams moving toward 50% odds — teams that were expected to make the playoffs off to a slow start or teams that are now surprise contenders — there’s opportunity to take action. To call on reinforcements from the Minor Leagues, make a tactical adjustment, or acquire outside help through a trade or signing.

Take last year’s Atlanta Braves. Coming off a seven-year run of playoff appearances, they were poised to contend for a title once again. Their playoff odds were 93.4% before their first game, but they quickly dropped to 69.4% after an 0-7 start.3

Other than shuffling the lineup a bit, the Braves decided to let things ride without intervention. After the seventh loss, manager Brian Snitker told the media, “It’s hard. And there’s nothing you can do but to fight your way out of it.”

That’s not true, though. Something can be done. Teams always have options, and yet, too often they do nothing.

Teams default to inaction because it’s less risky. Waiting draws far fewer eyes than promoting a prospect or making a trade. No one wants to overreact to a slow start and make the wrong move. But doing nothing, regardless of reason, is still a choice. And we rarely talk about the consequences of underreacting.

The 2025 Atlanta Braves failed to make the playoffs. It was the first time since 2017 they stayed home. It’s not clear what, if anything, they could have done, but the odds had shifted significantly after a mere seven days.

After this year’s first week, five teams have moved more than 10%.4 Two in particular (Boston and Houston) are in altogether different situations than they were prior to Opening Day.

These shifts should force real questions and dialogue within organizations. But most teams will ignore the results. They’ll stick to their priors and choose not to act. For some, that’s a decision they’ll regret in October.

Amateur and international baseball scouts write between 75 and 200 scouting reports a year. For pro scouts, the numbers jump to 400 to 700. That is a lot of writing.

Teams rely on the grades from those reports because they’re comparable across players and easy to incorporate into models. The overall grade is supposed to map cleanly to the written summaries — as the old scouting adage goes “words equal numbers, and numbers equal words.”

But while important in their own right, grades take full evaluations and mental models and compress them into a single number.

I’ve always had a theory that teams are missing valuable information hidden in the text. Evaluators don’t choose words at random. They take great care to find the right language to capture players.1

The question is whether that information adds anything beyond the grades. Does it tell us something meaningful, or is it just noise? So I decided to examine scouting summaries to test whether teams could make better decisions if they used that information.

Data

I used FanGraphs professional prospect rankings from 2017 through 2020.2

Among publicly available sources, FanGraphs most closely mirrors traditional team scouting reports. Their reports provide an overall grade — their Future Value — along with a written summary that relies primarily on in-person evaluations rather than aggregated opinions.

They are also consistent year over year. Eric Longenhagen led the process across all four years, writing two lists himself and collaborating with Kiley McDaniel on the others. That consistency helps reduce variation in grading scale and writing style.

The dataset includes year, organization, rank, player, age, highest level, position,3 estimated year of arrival, FV, report text, and report length, along with whether the player reached the major leagues, their debut year, and fWAR in their first three seasons.4

In total, the dataset contains 3,448 reports, with some players appearing multiple times.5

Approach

If you’re not interested in the technical details, feel free to skip ahead to the findings.

With the dataset in place, I tested different features of the report text, using Claude to help structure and iterate on the analysis.

• TF-IDF — Identifies words that are distinct to individual reports, focusing on terms that may carry evaluative meaning. Structural terms that teams could access elsewhere like level, age, and signing context were excluded.

• Sentiment scoring — Measures the overall tone of each report using a custom baseball-specific vocabulary to score sentences and determine how positive or negative the language is.

• Hedging analysis — Compares declarative language (e.g., will, has, shows) with conditional language (e.g., could, might, if) to estimate the level of confidence in a report.

• Sentence structure — Examines where positive or negative language appears (e.g., caveats at the end) and how much tone varies from sentence-to-sentence.

• Makeup language — Tracks the presence and density of character or makeup-related terms. (This didn’t turn up anything interesting.)

• Disagreement between grade and text — Measures how much more (or less) positive a report reads than we’d expect given the player’s FV.

In all models, FV, year, and log report length were used as controls. These controls ensure the results aren’t driven by differences in player quality, changes in the prospect pool over time, or the fact that better players tend to warrant longer writeups.

Subscribe now

With that out of the way, let’s dive into some results.

Does Text Add Signal?

The obvious place to start is a simple question: does layering text on top of FV improve our predictions?

The answer is yes.

FV alone explains a reasonable chunk of outcomes, but when we layer in text features, the R-squared6 improves, both for MLB Debuts and for 3-year WAR.

In both cases, our correlation essentially doubles. And that signal isn’t coming from everything in the report. When we split the text into evaluative language and more factual details like level, age, and signing context, the improvement is almost entirely from the evaluative words.

Tone

Knowing there’s some signal, we can move to sentiment. Do overly positive or negative reports mean something for players that are graded similarly?

Sort of. Tone matters, but not in the way we would assume.

If we’re predicting 3-year WAR for players who reach the majors, the results are essentially the same across players.

Looking at the likelihood of reaching the majors, the relationship flips. Players with the most positive reports reach the majors 60% of the time, while those in the bottom quartile do so nearly 65% of the time. The effect isn’t large, but it’s consistent and counter to what we’d expect.

Digging in a little further, the sentiment seems to have something to do with how the reports discuss athleticism and ceiling. The positive words that make reaching the majors less likely include frame, feel, athletic, physical, and any variations on the word projectable. Words that increase those chances include bench, role, and reliever.

But tone isn’t only about specific words — it’s also how consistent the language is.

When reports remained consistent throughout — neither switching between enthusiasm and concern nor saving reservations for the end of the report — players tend to produce more. Among players who reached the majors, the ones with consistent reports averaged 17% more WAR over their first three seasons than those who did not.

Conviction

Tone isn’t the only thing that shows up in scouting reports.

When I started scouting, my reports were full of hedging and conditional language. I didn’t have the experience to state things confidently.

Even as I gained experience, I felt more certain about some players than others. I spent more time with them, saw their full range of abilities, or simply had a stronger intuition.

In our sample, conviction has a small but measurable effect on outcomes. Unlike sentiment, the effects show up in production for players who reach the majors, and in the direction we’d expect. The most declarative reports result in players that produce 3 or more WAR 15.5% of the time, compared to 12.7% for the most hedged reports.

“He has plus bat speed and will hit .280” is not the same thing as “He has plus bat speed and could hit .280.” The grade and language may be similar, but the confidence is not.

Distribution

Until now, we’ve looked exclusively at averages. That’s not really how teams think about players, though.

Teams think in ranges — what a player could become, and how likely each outcome is.

Take players with a FV of 45. If we combine the analysis in the last two sections, reports that were confident and not overly effusive averaged 1.4 WAR against 1.0 WAR for hedged and enthusiastic reports. That’s a 40% difference in the mean outcome for players with identical grades.

The language in the report allows us to see the distribution around those outcomes. Among those same 45 FVs, reports that used ceiling and projection language — language that correlates with a lower rate of reaching the majors — produce 3 or more WAR at almost double the rate (17.2% vs 9.9%) as reports that don’t. Those players have more upside and downside, which intuitively makes sense.

Teams will sometimes try to capture this explicitly, asking scouts to provide a range of outcomes. But some of that information is already embedded in the language itself.

The same FV can describe very different players. Different risks, different paths, different shapes of outcomes. A 45 FV is a lossy format. The text helps us recover what was lost.

When Text and Grades Disagree

One final place where text adds information is when it disagrees with the grade.

When the grade and language are misaligned, outcomes are worse. Players reach the majors at a lower rate, produce less on average, and are less likely to become stars.

That holds true across all misalignments, but the direction matters, too. Reports that read more positively than the grade suggests produce slightly better outcomes than when the opposite is true.

There are a number of reasons the disagreement might occur. Risk, uncertainty, or inconsistency are all plausible. Regardless of the cause, the disagreement itself carries information.

Caveats

There are a few limitations to this analysis.

The data comes from FanGraphs, which, while the closest analogue to team scouting reports, is still a public-facing product with different audiences and incentives from teams. Teams also generate multiple independent reports on players, while this dataset relies on a single report per player.

The methods are relatively simple, and there is room to refine how language is measured and modeled. The results may also be overly sensitive to the specific sample or overfit to the data.

Finally, there’s an open question of causality. Most of the effects are small, and the relationships could run in either direction.

Conclusion

My hypothesis was that the language scouts use to describe players provides information above and beyond their grade.

This analysis doesn’t prove that definitively. But it does show something. The text adds signal, and more importantly, it gives us a better sense of confidence, upside, and the range of possible outcomes.

Those are the exact things teams are trying to understand when they make decisions.

And yet, most of that is compressed and stripped away. Discarding that information — for any reason — is the wrong choice.

I once wrote that teams could create a meaningful advantage by hiring and retaining more scouts. Over the last decade, many organizations have moved in the opposite direction, using technology to reduce their scouting operations.

At the risk of being stubborn, I think I had the right idea with the wrong conclusion.

Scouting, at its core, is about collecting and generating information. Technology can replace that information, or allow us to use it more effectively. And the advantage is in that choice.

Baseball is back! Spring Training has kicked off and we’re less than a week away from exhibition games. And yet, even with camps well underway and plenty of storylines to cover, the narratives that continue to dominate coverage are payroll inequality and competitive balance.

Major League Baseball famously operates without a salary cap (or floor). Depending on intent, it’s not hard to suggest that the lack of a cap has a major effect on competitive balance — or that it doesn’t matter much at all. Consider that all of the following things are true:

• 8 of the past 10 World Series Champions carried a top ten payroll.

• No champion was lower than the 17th highest payroll.

• Only 3 of the past 10 World Series winners finished with a top 5 payroll.

• The team with the league’s highest payroll has missed the playoffs in 2 of the last 7 years.

The popular move — at least among the owners — is to adopt a salary cap. That wouldn’t eliminate the spending discrepancy, but it would shrink it, and move baseball to a system that more closely matches the other major sports.

Would a salary cap help level the playing field? That’s hard to say, and even if it would, players aren’t interested in adopting one.

That doesn’t mean there are no options to address competitive balance. Spending is one lever — but it’s hardly the only one that impacts winning. If the league is willing to push past the obvious and get creative, there are other ways to increase parity.

Shorten the Season

One angle that doesn’t seem to get talked about much is the number of games in the season. Major League Baseball has played 162 games in both leagues since 1962. Maybe it’s time to rethink that number.

Theoretically, shortening the season has two benefits to competitive balance. The first is that a shorter season introduces more variance. Baseball already is more unpredictable1 than other sports, with a fair amount of luck involved in who advances to the playoffs with the current schedule. Reducing from 162 games to 150 or so would introduce more chaos and take away some of the advantage of spending at the top of the scale.

We saw an extreme version of this in the 60-game season in 2020. It’s probably not a total coincidence that was the only season in the past ten years that a team with a bottom five payroll played in the World Series2.

The second benefit is that a shorter season could also reduce payroll disparities over time. With fewer games, the league wouldn’t necessarily have to shorten the length of the season — it could simply insert additional off-days. Those off-days would likely fall on weekdays and other off-peak attendance dates, increasing attendance per game3 and shrinking the gap in local revenue totals per club — which could, in turn, narrow spending differences.

Introduce Relegation

In other parts of the world, one of the biggest incentives for teams to put out a competitive product is the threat of relegation. The lowest-performing clubs in most European soccer leagues don’t get to try again next year; they are demoted to a lower division while new teams take their place.

Implementing a system in the United States would be enormously complex, but big picture, it could mitigate tanking strategies, inspire urgency around winning, and introduce additional drama and excitement for fans.

The commissioner has long discussed expansion. Maybe instead of growing to 32 teams at the top level, the league should shrink to 28 and introduce a secondary tier to enable promotion and relegation.

The popularity of a product like the Savannah Bananas — while a vastly different experience — suggests there is space for compelling baseball outside of the traditional MLB structure. And who wouldn’t be captivated by simultaneous races to avoid relegation and earn promotion?

Allow Advances on Revenue Sharing

The main way that MLB currently addresses imbalance among clubs is through revenue sharing. The basic idea is that all 30 clubs contribute a portion of their local revenue to a central fund, and the proceeds are distributed equally — often resulting in large-market clubs providing payments to small-market clubs.

These payments happen on an annual basis. But what if the league introduced flexibility that allowed small-market clubs to use those payments more strategically to push chips in during a particular competitive window via an advance on revenue sharing?

Take the Pittsburgh Pirates. Per FanGraphs, they currently project to be the 12th best team in Major League Baseball in 2026 with a 42% chance to make the playoffs. They also project to have the 22nd highest payroll in the league. Rather than taking their expected annual revenue sharing check this past year, they might have decided to “borrow” from future payments to give themselves a better chance to land a top tier free agent like Kyle Schwarber or Framber Valdez.

The trade deadline also provides an opportunity for small-market clubs to infuse some future capital into their club. This past year, the Milwaukee Brewers had a relatively quiet trade deadline despite being one of the best teams in the National League. Had they been able to inject some future revenue sharing payments into their budget, maybe they would have been bigger players for expensive bats like Josh Naylor, Ryan O’Hearn, or Eugenio Suárez.

With this change, revenue sharing would do more than determine how much support small-market clubs receive. It would allow them to choose when to invest, aligning spending with their best opportunities to win.

Increase Player Limits for Small-Market Clubs

Most of the suggestions to this point would be fairly visible for fans. A more stealth idea is to increase the number of reserve roster spots from 40 to 42 for small-market clubs.

Changing the number of active roster spots — the 26-man roster — would be extremely noticeable and introduce the possibility of unintended consequences that aren’t worth dealing with. But the 40-man roster? Most fans probably don’t really know that it exists. And in-season, because of the 60-day injured list, teams often carry more than 40 players under reserve already.

These extra roster spots probably aren’t worth millions of dollars or tens of wins, but they would give teams additional time for development and in-season depth — particularly important for clubs without the financial flexibility to upgrade their team through the open market. They would also allow for more speculative waiver claims and less concern about pending Rule-5 status in trades.

This past World Series featured heroic performances from Will Klein and Miguel Rojas — two players much closer to player number 40 than player number 1 on the roster. Extending the reserve roster for a portion of clubs could mean the difference in advancing to the playoffs or hoisting a championship trophy.

Implement a Draft for Senior Staff Roles

No one would suggest that staff members affect wins more than players, but the right leaders absolutely make a difference. They assemble rosters, dictate philosophies, build infrastructure, establish culture, and make decisions that affect the outcomes of games and seasons.

So if we care about competitive balance, why do we only regulate player labor?

Enter a draft for senior staff members.

It would be just what it sounds like: a draft designed to distribute key leadership talent and the associated organizational competencies. Teams would have to announce their intention to move on from existing staff members and enter the draft via blind submission by October 1 in a given year, creating a hiring pool. Selection priority could then be determined through similar mechanisms to the amateur talent acquisition system — rewarding clubs with smaller markets or sustained on-field struggles.

In practice, the system would likely be limited to a small number of senior roles (such as Manager, President of Baseball Operations, or General Manager) where compensation and influence are significant enough to justify participation.

Defining eligibility may prove challenging4 given that titles and reporting structures vary widely from club to club. Teams would also likely resist anything that moves away from full autonomy on hiring. There would need to be strong incentives and conviction to outweigh the concerns.

Still, if leadership and decision-making infrastructure contribute meaningfully to long-term success, distributing that talent systematically is another lever the league has to influence competitive balance.

The conversations around competitive balance are getting louder, and they all seem to center around the same topic: payroll.

There aren’t nearly as many pleas for more variance, more optionality, or better leadership, even though those things all have an impact on winning as well.

The ideas above certainly aren’t the only possible ways to address imbalances — but that’s sort of the point. The current rhetoric focuses almost exclusively on one aspect of competitive balance, when in reality there are many other avenues to explore. Once we push past spending as the only option, new possibilities become much easier to imagine.

With Spring Training just a few short weeks away, we are firmly in the midst of prospect ranking season. Seemingly every day, a new top-100 or team list is published at one of the major sites. For many fans, prospect coverage is the highlight of the offseason. It offers not only entertainment, but also hope — hope for a better future.

As I’ve seen new content released, I started to wonder: do these lists offer more than hope and entertainment? What do farm system rankings really tell us about an organization’s future?

After all, the goal isn’t (or at least shouldn’t be) to have the number one farm system. It’s to win games and championships at the big league level.

To examine the relationship between farm system strength and future Major League success, I gathered Baseball America’s1 organizational rankings from 2004-20212 along with each club’s win totals3 . Once I had those, I bucketed the rankings into four tiers: teams ranked 1-5, 6-10, 11-20, and 21-30. Looking at nothing more than these simple tiers, we see a clear pattern.

Teams in the top tier averaged roughly 87 wins per season over both the three- and five-year windows following their ranking. Teams in the bottom tier averaged closer to 79 wins per season.

The relationship holds when we look at the change in wins. Taking into account the prior year win totals, Tier 1 organizations improved by roughly four to five wins per season over the same three- and five-year windows. Teams in the bottom tier moved in the opposed direction, declining by two to three wins per year.

To test this more formally, I ran regressions linking farm system rank to future wins. I did so both with and without including the prior season wins to control for the quality of the big league team.

The results hold. Even after accounting for a team’s starting point, moving from a bottom-tier system to a top-tier system is associated with about 30 additional wins over the next five seasons — about six wins per year. That’s enough to potentially move a team from mediocrity to consistent contention.

I have to admit, I’m a bit surprised by the strength of the findings.

For starters, BA is not actually trying to predict which teams are going to be the best over the next 3-5 years. They’re ranking farm systems based solely on the talent of players yet to exhaust their rookie eligibility, with no consideration for the quality of the ML roster, the state of the payroll, the club’s market size, or their competition.

Additionally, predicting prospects is hard. We’re not particularly good at forecasting how established veterans will perform next season, let alone projecting young players several years into the future. Many players will over- or under-perform expectations in ways that can dramatically affect win totals.

And yet, we see a strong relationship between where a team ranks and their short- to medium-term outcomes. The results reinforce how important the health of a farm system is to an organization’s future.

That’s part of what makes prospect season so compelling. The lists aren’t merely offseason filler — they’re signals of where a franchise may be heading. High rankings don’t guarantee parades, but it does seem that the associated hope is well placed.

This morning Eno Sarris and Will Sammon published an interesting piece on The Athletic titled Tommy Pham makes his case for MLB’s next big metric: He calls it PhamGraphs. In it, player Tommy Pham makes the case that advanced baseball statistics are missing important context.

At the heart of one matter for Pham is the chasm between playing for a winning or losing team. Pham, who has been on both sides of that equation, says that playing for a powerhouse means a chance to feast on lesser pitchers when your team is leading comfortably. Alternatively, playing for a non-contender sometimes means facing high-leverage relievers more often as superior opponents close out another victory.

“It was just so many games like that where I’m like, damn, rather than facing the mop-up guy, where you know you can tack on another hit or walk, whatever it is, you’re losing so many at-bats,” Pham said.

Pham has thoughts on defense, too. He wants advanced numbers like outs above average to incorporate elements such as wind and sun.

I promise I won’t make all of my posts reactions to Sarris’s articles, but as I read and reflected I had a series of reactions that felt worth sharing.

Pham’s chief concern seems to be that even the best widely available statistics aren’t taking enough information into account. He notes that wRC+ — an all encompassing hitting stat that credits hitters for each outcome while controlling for park effects and the league run scoring environment — doesn’t consider the quality of the pitchers the batter faced.

Subscribe now

My immediate take was that Pham is right, but not exactly in the way he’s framing the problem. The specific examples he cites are real, but most teams already account for these factors in their internal models, even in the public never sees the work.

At first glance, it seemed that this was simply a misguided understanding of the difference between what public-facing metrics — even the best ones — do and what teams are using to make decisions.

Almost all of the statistics available on sites like FanGraphs, Baseball-Reference, and Baseball Savant are descriptive. They exist largely to explain what happened. Even more theoretical stats like xwOBA represent a version of what happened; in this case what the player’s wOBA would have been with average game conditions. Some statistics are going to be more indicative of future outcomes than others, but by and large, the purpose of public metrics is to describe and summarize the past. And that makes sense! If I’m a fan, I want to know what has happened. Debates about awards and the Hall of Fame should be based on results that actually occurred.

But while fans, journalists, and commentators spend time dissecting and debating the past, teams really only concern themselves with the future. A player’s past performance has already been banked. The games have been played and the wins and losses have been recorded.

That’s why internal team models and tools attempt to do just what Pham is asking for: incorporate as much information and context as possible — quality of competition, usage, environment, health, aging — to estimate future performance. They’re focused on providing predictive information. Past performance is only important in that it helps inform bets about the future. If Pham wants a rough sense of how teams are likely to value him in free agency, he doesn’t need entirely new statistics. Public projection systems like Steamer or ZiPS already layer in much of the contextual data that has been shown to improve forecasts.

But the more I thought about Pham’s contention, the more I realized that my initial reaction was too dismissive — and that he is unfortunately closer to the truth than I’ve described.

Because the reality is that while teams should only use forward looking statistics, every organization I’ve ever worked for spent a meaningful amount of time discussing descriptive statistics when evaluating potential acquisitions. Some of that comes from misunderstanding. Some of it is hubris and the belief that experience or intuition should consistently override models. Some of it is the ubiquity of the numbers Pham references. And some of it is just human nature and our biases.

Even the most devout Superforcasting-fan might hesitate before committing real money to a hitter who just hit .200. It’s hard to reconcile when a model suggests a player is going to be good when the most visible evidence points in the opposite direction.

The irony is that discomfort is exactly where the opportunity lies. Players that both project and play well are going to be expensive in free agency or via trade. The edge typically exists when there is a disconnect — when the numbers that describe the past conflict with the projections.

Organizations that can truly learn to separate those two things — what a player has done and what we think they will do — give themselves a chance to find value where others won’t. That doesn’t mean abandoning judgement, context, or human elements. It means anchoring evaluations in objective, future-oriented tools and adjusting from there.

Tommy Pham is correct to say that context often gets lost in the way performance is discussed. Major League teams would be wise to keep that in mind as they go about building their rosters this offseason.

A few weeks ago, the Tampa Bay Rays did what the Tampa Bay Rays do. They traded away veteran players for younger, cheaper talent, sending Shane Baz to Baltimore for four players and longtime second baseman Brandon Lowe to Pittsburgh in a three-team deal that brought back two more prospects.

To most observers, these moves exemplify the unavoidable reality of being a small-market club. Teams like the Rays are constantly reacting to the financial constraints imposed by baseball’s economic system.

Major League Baseball operates with neither a salary cap nor a salary floor, creating larger spending gaps than the NFL, NBA, or NHL. In 2025, the Los Angeles Dodgers spent roughly $270 million more on Major League players than the Chicago White Sox.

That extreme gap fuels persistent claims that big spenders are ruining the sport. And to be clear, spending is a real advantage. Over the last decade, nine of the ten World Series champions ranked in the top ten in payroll, and six winners — including the past three — finished in the top five.

It is harder to win as a small-market club. There’s not really an argument to the contrary.

And yet, over that same ten-year span, the Rays have compiled the eighth-most regular-season wins in baseball. Ahead of them sit the Cleveland Guardians and Milwaukee Brewers — organizations operating under many of the same financial constraints.

So what if these teams aren’t merely succeeding in spite of their limitations, but in some cases, thriving because of them?

When we look closer, a single hidden advantage emerges from being a small-market club: focus. Small-market teams don’t have the option to pursue every path, hedge every decision, or compete in every market. They’re forced to narrow their ambitions, choose fewer priorities, and concentrate their resources with unusual clarity.

Let’s take a look at what that focus looks like in practice.

Subscribe now

Choosing Fewer Markets

The most obvious place that spending constraints show up is in free agency. Despite interest — and even competitive offers — headline free agents overwhelmingly sign with teams in the largest markets.

By total dollars, the ten largest free-agent contracts in baseball history have been signed exclusively by the Mets, Dodgers, Yankees, Phillies, Padres, and Rangers. Juan Soto’s $765 million deal carries an average annual value of $51 million. That’s not far off from the largest total commitment the Brewers ($80 million for Lorenzo Cain) or Guardians ($60 million for Edwin Encarnación) have ever made to a free agent, and it’s $11 million more than the Rays’ largest contract ($40 million for Zach Eflin).

For small-market clubs, this reality effectively removes free agency as a primary team-building avenue. And while that limitation is often framed as a disadvantage, it quietly forces a kind of strategic focus that better-resourced teams rarely adopt.

That focus shows up in two distinct ways.

The obvious one is that Major League free agency generally represents a poor return on investment. Teams are typically paying for past performance, buying decline years at premium prices, and accepting limited flexibility when deals go bad. Even good processes can fall privy to tacking on additional years and dollars a la the winner’s curse.

The second effect is less discussed, but far more important.

Because small-market teams can’t realistically build through free agency, they can largely ignore the market altogether. They don’t have to spend months every winter evaluating, courting, and debating expensive veterans they’re unlikely to sign. That’s a meaningful organizational advantage. Every team operates with finite resources — not just dollars, but human attention and decision-making bandwidth.

Consider this year’s free-agent market. How will Dylan Cease age as a two-pitch pitcher? Which version of Cody Bellinger is going to show up? How will Framber Valdez fit in a new clubhouse and new environment? Large-market teams have no choice but to wrestle with those questions. Small-market teams don’t only avoid answering them. They don’t have to ask them at all.

Instead, those resources get redirected to areas with better returns like scouting, player development, research, and process improvement. Because these teams can’t buy talent as easily on the open market, they’re forced to turn to other avenues to improve. That helps explain why the Rays trade more frequently than any other team in the league. Trading has to become a primary way to reallocate talent and shift windows to remain competitive.

Even if a large-market club wanted to ignore free agency entirely, it likely couldn’t. The expectation to participate — from fans, media, ownership, and even players — makes disengagement difficult. Small-market clubs don’t have that problem.

Freedom From Optics

Although a bit of an oversimplification, small-market clubs generally exist because fewer people are paying attention. They’re from smaller cities, they have fewer fans, they have shorter histories — whatever the reason, the audience is smaller. That changes decision-making in important ways.

Small-market clubs can focus on what they’re doing, free from pressures that their large-market competitors struggle to ignore. They don’t need every move to be immediately legible or publicly defensible. That’s another real advantage that puts a wider range of ideas on the table.

If the Rays want to trade a star player, they don’t have to consider pushback from their fanbase in nearly the same way large-market teams do. When they have information that suggests a player is improperly valued, they can act without hesitation.

Small-market clubs have no choice but to believe in their projections when it comes to player acquisitions. They have to find players that can provide value without name recognition or full sticker price. When the Rays traded Chris Archer in 2018, it wasn’t immediately obvious that Tyler Glasnow would become a frontline starting pitcher. Glasnow had a 4.34 ERA in low-leverage long-relief that year and a 7.69 ERA over 13 starts the previous year. But they believed in their information to take a shot on a pitcher with elite stuff and strong strikeout rates in a different role. Many clubs likely had access to similar information. Few were in a position to trust it.

(As an aside, I used to think there was little reason for large-market clubs to operate this way. It seemed silly to pay full price for players when cheaper alternatives existed, assuming teams were willing to accept a bit more risk. With more reflection, I’ve come to see that this can be a rational — and often prudent — choice. As Rory Sutherland says, “People do not choose Brand A over Brand B because they think Brand A is better, but because they are more certain that it is good.” If you can afford it, certainty provides more than just a luxury.)

Over time, the Rays have built a reputation for acquiring players with flawed profiles and coming out ahead far more often than not. But not every bet works. An underrated benefit of operating with less scrutiny is that they don’t overreact when something goes wrong. No one is fired for a single misstep and they don’t abandon their processes simply because of a bad outcome. Small-market teams understand they won’t hit on everything — and that their only choice is to stay consistent and keep taking informed swings.

Calculated Risks

If small-market clubs try to do the same things as their competitors, they will almost certainly lose. The goal isn’t to simply be competitive; it’s to be comparatively better in specific areas. They must be different if they want to win. That reality forces them to think differently and take targeted risks.

Because small-market teams have already narrowed their priorities, they can afford to push harder in those areas. Risk becomes concentrated and grounded. The upside, if they’re right, is meaningful.

Some teams will take established principles to their extremes in search of an edge. The Guardians’ hitters had the platoon advantage in just over 84% of their plate appearances this year. The Brewers limited their starters to facing batters for a third or fourth time to a league-low 573 plate appearances. Others have experimented with openers, aggressive defensive positioning, or more fluid roles.

These aren’t attempts to be clever. They’re examples of small-market clubs implementing creative strategies to level the playing field.

The lone team in the past decade that won the World Series without a top ten payroll was the 2017 Houston Astros. They were not a small-market club, but they leveraged the same innovative spirit to make up for a lower payroll. They were early adopters of Edgertronic cameras, they reimagined scouting, and they thought differently about who they hired and why. They took risks in areas where they believed it would matter most.

Small-market clubs know they have to be different. They can’t play it safe. But when they narrow their focus and block out external pressures, it makes their bets feel less risky. And that puts them a little closer to closing the gap created by spending.

Spending less on Major League payroll is a disadvantage. There’s nothing romantic about it, and there’s no bonus for doing more with less.

But constraints have a way of clarifying priorities. Teams that accept what they can’t do are forced to narrow the field, concentrate their effort, and make deliberate choices about where they compete.

Small-market successes like the Rays didn’t choose this model. They were pushed into it. What separates them is that they’ve built an organization designed around focus. And in a league defined by imbalance, that focus is often the difference between staying competitive and falling behind.

I don’t think it’s an accident that most professional sports have head coaches while baseball has managers.

Unlike other sports, managers aren’t designing schemes, calling plays, or overseeing a specific unit. In most cases, they aren’t running drills or providing daily instruction. The job is less about coaching and more about, well, managing. It’s about managing personalities, information, relationships, and expectations over the course of a long, grueling season.

That doesn’t make the role easier or harder than a head coach’s job. But it does make it different, and harder to evaluate and hire for using traditional screens.

Subscribe now

This series stemmed from a few curiosities about managers in Major League Baseball. I wanted to know who teams hire and if any obvious traits or experiences actually predict success. What we’ve learned is that teams hire from a fairly narrow profile, visible qualities don’t tell us much, and teams cycle through managers at an extremely high rate.

While interesting, these findings don’t offer anything definitive. Hiring a manager is always a high-stakes decision that comes with uncertainty and pressure. The data can tell us what hasn’t been especially informative, but it doesn’t suggest what teams should do differently.

From here, we can shift from results to reflection. Let’s take everything we’ve learned and walk through how we might structure a reimagined managerial hiring process.

Experience

Experience is an understandable place to start when hiring a manager. Unfortunately, as we’ve seen, resumes aren’t terribly informative. People develop very different skills and capabilities even when they’ve held similar roles for similar lengths of time.

I’m more interested in evidence of leadership than in any particular career path. That could come from managing in the past, leading a department in a front office, or operating in a completely different environment altogether. Baseball is treated as uniquely insular, but the core demands of the managerial role aren’t unique to the sport. In that sense, leadership may be more transferable than we tend to assume. If we believe that, it opens up the range of candidates worth considering.

Skills, Traits, and Ways of Thinking

If experience isn’t particularly predictive — and the job of a baseball manager is fundamentally different from that of a head coach — then hiring decisions need to shift away from background and toward the skills, traits, and ways of thinking that actually drive success in the role.

That sounds straightforward in theory. In practice, it’s more complicated. Traits like communication, curiosity, adaptability, and relationship-building are harder to define and harder to evaluate than years of experience or past titles. Two people can watch the same interview and come away with very different impressions.

I saw this firsthand during a managerial hiring process I was part of. We identified ten characteristics we wanted to evaluate and independently ranked candidates across each dimension. Our lists were kept separate until interviews were complete, specifically to reduce groupthink. When we compared rankings, I was surprised to see significant differences with a colleague I usually aligned with. When we dug in, the discrepancy wasn’t about how we assessed the candidates’ abilities. It was entirely about how important we believed each trait was.

That experience clarified something essential: identifying desirable traits isn’t enough. Hiring requires figuring out how those traits are weighted relative to one another. No candidate is going to be exceptional across every dimension. It’s important that teams are able to compare people with different strengths.

That isn’t to say that teams should design a rigid scoring system and blindly follow it. But it does mean their process should be structured enough to surface these differences early and intentionally, rather than discovering them accidentally at the end.

I would prioritize traits that are hard to teach and costly to miss on such as relationship-building, a willingness to communicate directly, curiosity, and adaptability. Other qualities like critical thinking, in-game decision-making, and comfort with the media matter, but they’re more teachable. A manager who is open and adaptable can make better bullpen decisions over time. It’s far more difficult to make someone trustworthy.

Interview format

Once we establish the leadership traits we want to prioritize, it’s critical to design an interview process that allows those traits to actually surface. That’s easier said than done, as many of the qualities teams care most about are going to be contextual, relational, and highly subjective.

The risk in most interview settings isn’t the questions themselves, but formats that are safe and conventional — and limit what candidates can actually reveal.

With the caveat that I don’t know the rigor or structure of every managerial hiring process league wide — I’ve generally seen baseball operations interns put through more taxing and comprehensive screening than managers. Inertia and pride make it hard to design a process that feels appropriate for well-known and highly accomplished candidates.

The same process I mentioned above utilized a two-hour written portion the candidates would complete in-person the morning of their interview. I thought it was reasonable to supplement the rest of our process with these questions, but when news got out, it was unthinkable to most that we would ask so much of the candidates.

That highlights the importance of ignoring outside noise and pressure to design a process that more closely resembles the job itself. That might include scenario-based discussions that introduce ambiguity, exercises that require candidates to piece together information in real time, or conversations that test how they’d navigate disagreement. It could also include more outside the box things like utilizing group interviews to see how people interact outside a one-on-one setting.

I would think intentionally not only about structure, but also about who is involved in the interview process. This is a major organizational decision to hire a leader who will touch and oversee a lot of people. The decision on who to hire will come down to a small handful of people atop the organization, but that decision should be influenced by a wide range of perspectives. I would minimally consider including all of the following:

• Front Office Staff - GM, AGMs, Director of Baseball Operations, Director of R&D, Pro Scouting Director, Advance Scouting staff

• Performance Staff - Head ATC, Head S&C Coach, Director of Performance Science, Director of Mental Health & Performance

• Player Development Staff - Farm Director, ML Coaches, MiLB Coordinators

• Clubhouse and Travel Staff - Clubhouse Manager and Director of Team Travel

• Players - 2-5 Major League players, preferably veterans with some influence

Players are the group that I’ve rarely seen included and it’s sort of indefensible to me. A manager needs to be able to establish trust, have difficult conversations, and inspire confidence with people in the clubhouse — why wouldn’t we want to get their feel for those things during the hiring process? And don’t we think involving them — even if we don’t go with their choice — would make immediate buy-in far more likely?

The process wouldn’t be fast or comfortable. And that’s the point. It would utilize a number of tried and tested interview staples — like rubrics, independent scoring, and skill assessments — and include a wide range of people. Our process won’t eliminate risk, but it should increase the likelihood that we’re evaluating things that will predict success.

Supporting the Hire

If we consider experience appropriately, identify and value the traits that matter, and design a thoughtful interview process, we should make a good hire, right? Maybe. Hopefully. But even then, a great deal of uncertainty remains.

That’s where organizations can make a difference. Hiring a manager isn’t just about selection; it’s also about what happens once the decision is made.

What I tend to see is that once a hire is made, organizations offer surprisingly little in the way of structured support. Outside of getting deeply involved in in-game decision-making, many teams largely leave managers to figure out the rest of the job on their own.

It reminds me of Minnesota Vikings head coach Kevin O’Connell’s sentiment on quarterback development. O’Connell said, “I believe that organizations fail young quarterbacks before young quarterbacks fail organizations.” When I look at the turnover in managerial hiring, it’s hard not to see a similar dynamic at play.

Support starts with clarity and autonomy. Managers benefit from explicit expectations around decision-making authority, communication norms, and how disagreements will be handled. General managers and managers are responsible for different things. Those boundaries should be clearly defined and then respected.

Beyond that, support should focus on the parts of the job that are the hardest to develop. Rather than being critical and overly-involved with in-game decisions, organizations can add more value by coaching managers on building relationships, navigating difficult conversations, and dealing with stress. They can hire coaches that complements the manager’s strengths and addresses their weaknesses. They can expose the manager to front office processes since we saw that experience shows some signal as far as who will be successful.

In an ideal world, our new hire would have a meaningful runway to settle in, learn on the job, and find their footing. But we know that decision won’t always be in our control. Finding ways to support our new manager will make it all the more likely that we’re happy with our hire rather than designing a new process for a replacement.

This series began with simple questions: who gets hired to manage Major League teams, and what predicts success once they do? They didn’t lead to perfect answers because there is more than one blueprint for a successful manager.

What we did instead was challenge some conventional assumptions and move toward a different way of thinking about the role. A way that places less emphasis on background and more on how we can evaluate, hire, and support managers over time. Hiring a manager isn’t a single, discrete event. It’s the beginning of a longer commitment that plays a meaningful role in shaping whether the decision ultimately succeeds or fails.

Earlier this week, The Athletic’s Eno Sarris reported that Major League Baseball is making a move to regulate in-game technology available for use across all minor-league parks.

“We are taking this step to ensure that all 30 Clubs are working with the same set of information as they make Baseball Operations decisions,” an MLB spokesperson wrote in a statement confirming the change. “Each Club can use the information however they see fit, but it is important that we provide a level playing field on access to information.”

MLB had been discussing a rule change for the past few years. Now they’ll approve in-game data and technology vendors and standardize access for all 30 clubs.

On its face, the goal is straightforward: reduce disparities in access to information. Sarris suggests that the main reason for the change is to curb the advantage wealthy teams have.

Deep-pocketed organizations have gained an advantage in recent years by flexing their financial muscle in the minor leagues. By spending freely on cutting edge tools like high-speed cameras and motion capture technology to gather data in-game, those teams widened the information gap over rivals with fewer resources.

That explanations is plausible, but I don’t think that’s what is going on here. The issue doesn’t seem to be who is spending, but simply that spending exists and continues to rise.

To start, I don’t think market size or payroll correlate all that well with investments in data and technology. Some of the lowest revenue clubs in the sport — like the Rays, Brewers, and Guardians — invest as heavily in these areas as anyone else. While large market clubs could create an overwhelming advantage, it’s not happening right now. And paradoxically, blanket limits will probably hit the most innovative small-market teams the hardest because they rely on creativity and implementation to compete.

Spending limits themselves don’t have to eliminate advantages. Constraints often spur creativity, and the best organizations will respond by finding edges in new places. More importantly, the changes seem to only affect in-game data and technology. There is a lot teams can — and already do — in non-game environments to learn about their players and develop them. There is no impact on scouting, staffing, or (I’d bet) on operations in the Dominican Republic. That leaves plenty of opportunities for teams to invest and get ahead if they don’t let the rule changes limit them.

Subscribe now

The change also isn’t likely to reduce spending for all clubs. It will cut spending on the whole, but some teams will see costs rise if they weren’t investing in the new standards. When the article says “MLB will pay for and manage the technology” it’s a bit misleading. Money spent by the league in practice means that all 30 clubs are sharing the bill equally; not that individual clubs aren’t spending.

It’s hard not to view this change through the lens of upcoming labor negotiations. MLB has made it clear that they would like a salary cap. To even broach the topic, the league needs to be able to guarantee players a revenue split at or near 50%. Curbing spending on things like data and technology may signal to players that they’re serious about investing a larger percentage of their spending directly in player salaries.

Still, the biggest concern around the rule change is potential consequences on innovation. If the league limits spending and standardizes vendors, it’s effectively discouraging competition in the marketplace. New vendors and data sources emerge every year, pushing existing partners to improve their products or find new solutions. Advances might stall if teams are unable to experiment or work with new partners.

If this move is part of a broader effort to reshape the league’s economic structure ahead of labor negotiations, it might make sense in the in the short term. But no revenue split works if the underlying product suffers and the overall pie doesn’t continue to grow. I hope the league knows what it’s doing.

When the Los Angeles Angels introduced Kurt Suzuki as their new manager in late October, he spoke about how much it means to him to get an opportunity to manage. He expressed gratitude and excitement to lead a team he once played for. The comments could have come from any other introductory press conference — until General Manager Perry Minasian revealed that the organization had committed to Suzuki for just one year.

“It’s a one-year deal, so he’s tied in with me,” Minasian said. “But for us, in sports, in general, everybody’s on a one-year deal. That’s just the way professional sports is.”

It was an unusually blunt acknowledgement of something that has become increasingly true: managers don’t have long to find success before teams start looking for their replacement.

In Part I of our series, we looked at who teams hire. In Part II, we tested which of those public-facing traits, if any, predict success. Now we turn to what happens after the hire.

How long do managers last? And what influences the decision to retain or move on?

Tenure isn’t oft-talked about, but it’s meaningful. It captures the point where expectations, performance, and patience meet. And it shows — in at least a broad sense — how long teams are willing to give a hire before they’ve seen enough.

When we look at the data, we see that the average tenure for managers hired this century is just 4.1 years, and the median is even lower at 3 years. Those numbers actually overestimate typical tenure because we credit managers who were fired mid-season with a full season. What we’re really saying is that the average hire lasts about three Opening Days.

Averages can be skewed by outliers, though. Another way to think about tenure is to focus on the distribution. Nearly 40% of managers never reach a third season. More than 70% never make it to a fifth. Only eight managers have reached ten or more seasons this century.

Now, you might assume that active managers will push the numbers up. Maybe. But while there are some longer tenured managers like Kevin Cash and Aaron Boone, almost two thirds — 19 of 30 — of active managers have been in their current positions for two seasons or fewer entering 2026.

These number add up to a lot of hiring. The average club has hired 6.1 managers this century. Kurt Suzuki is the sixth different Angels’ manager hired since 2000 — and that’s with the first 19 seasons covered by Mike Scioscia! The Marlins “lead” the pack, having hired ten different skippers:

One more thing about tenure worth flagging: when we tested tenure with our success metric, we found a faint but real signal. Managers who last longer tend to outperform expectation. Each additional year a manager lasts is associated with roughly two extra wins relative to preseason predictions.

That’s not to say that tenure causes success. It doesn’t. It means that the industry does a reasonable job identifying who is doing a good job and sticking with them. Tenure is usually a reflection of success rather than a predictor of it.

Taken together, we see a job that has little security and even less predictability. Why is that?

Subscribe now

I would bet that no owner or GM hires a manager with the expectation that they’ll be going through the same exercise again in just a few years. They’re aiming to add someone who can lead the franchise through a sustained period of excellence. A number of forces contribute to the high turnover rates we see across the league.

The standards are high and unforgiving

Baseball is a zero-sum game. To win more games, someone else has to lose more. Judged solely by wins and losses — or by a metric that compares actual results to preseason expectations — it’s not possible for every managerial hire to look successful. The structure of the sport ensures turnover.

And sometimes, even winning isn’t enough. Only one manager can win the World Series each year; the other 29 fall short. That’s why, despite one of the highest regular-season winning percentages in history (.621), two World Series titles (to that point), and 27 wins over expectation during his tenure, fans and media outlets were calling for Dave Roberts to be fired this fall — just months before he guided the Dodgers to another championship.

Front office alignment matters

We’re talking about tenure for managers, but managerial stability is directly linked to front office stability. When a new general manager or president arrives, they often want to hire their own manager — someone they trust, communicate well with, and share alignment on team building.

Even when the front office doesn’t change, that relationship still matters. Managers can find themselves out of a job if they don’t maintain strong working relationships with baseball operations leadership. “Philosophical differences” has become a common refrain during managerial transitions.

The clubhouse effect

One of a manager’s core responsibilities is to inspire players and create an environment where they can be their best. If a manager loses credibility with players or the staff, it’s hard to recover. Culture isn’t an easy thing to measure, but it’s often one of the strongest drivers behind midseason firings.

Right manager, wrong time

The sentiment exists that the manager who is right for a rebuilding team won’t necessarily be the right one when the club is ready to compete. Rebuilding teams prioritize teaching, development, and working with young players. Contending teams want someone who can make in-game decisions under pressure and hold veteran players accountable. Whether that approach is right or wrong, it contributes to short leashes as organizations shift between competitive windows.

The hire doesn’t work out

Occam’s razor says the simplest explanation is usually the best one. Sometimes teams just get it wrong. No matter how thorough the interview process, there’s no substitute for how someone will perform once they’re in the role. Maybe the communication style didn’t translate, the staff didn’t gel, or the manager struggled with the demands of the job. In those situations, teams understandably prefer to move on quickly.

The job is complex and hard to evaluate

Finally, the role of the manager has evolved and expanded over the years. It’s not simply a matter of filling out a lineup card and managing a bullpen. Managers must process and integrate insights from research & development, sports science, and performance departments. They’re tasked with overseeing coaching staffs that are larger than ever. They are asked to talk to the media more frequently about a wider range of topics. It’s a difficult job!

At the same time, it isn’t easy for teams to evaluate managers. We discussed in Part II some practical challenges utilizing wins and losses to judge managers. It’s not necessarily easier to evaluate how well someone is establishing a culture, motivating players and staff, or communicating clearly. And even if teams feel they can evaluate the different elements well, there’s still the matter of what to do when the manager is doing well in some areas but struggling in others.

All of these factors combine to make managerial tenure one of the most unpredictable outcomes in the sport — driven as much by context and alignment as by anything the manager does on the field.

Part I of our series opened with the argument that manager is “arguably the most important non-playing position in a Major League organization.” The turnover we’ve seen this century doesn’t always make it feel like teams treat the role that way.

Looking at the tenure data, a clear reality emerges: managers have very little time to find their footing before teams consider making a change. Tenure is less a referendum on a manager’s ability and more a reflection of everything around them — expectations, alignment with the front office, clubhouse dynamics, timing, luck, and the growing complexity of the job.

That makes tenure hard to interpret from the outside and even harder to predict. And it reinforces a broader theme of this series: the visible traits that shape hiring decisions don’t tell us much about who will actually succeed once the job begins.

In the final part of this series, we’ll step back from the data and look at what all of this means for the hiring process. If teams want to make better decisions, what should they prioritize? And what qualities really matter when choosing someone to lead a clubhouse?

If you had to hire a new manager for your favorite Major League Baseball team, what traits would you target?

In Part I, we looked at who Major League teams have hired when they have managerial openings. We found that they favor a similar candidate – a former professional player in their late-40’s who managed in the minor leagues or coached in the big leagues. There has been some variation, but by and large teams tend to stay well within the box when they hire.

Today we’re going to dig deeper into whether those hiring practices actually produce good results.

To answer that, we first need to define what success looks like. Unfortunately, there’s no clean or perfect way to do that when it comes to managers. Looking only at wins and losses suggests that a manager can’t be good with a bad roster (or bad with a good one), and I don’t think anyone would argue that Manager of the Year voting is a reliable measure of quality.

Instead, we’re opting for a simple, objective proxy: the difference between preseason Vegas win totals1 and actual wins. Using the gap between projected and actual wins offers a clean way to isolate how much a team over- or under-performed relative to expectations. We assign that difference to the manager.

For every manager hired after 20002, we looked at the difference between the projected wins and the actual wins during their tenure. Excluding the ten managers hired this offseason, we’re left with 172 managerial hires. Of those 172, 68 (40%) outperformed expectations while 101 (59%) underperformed. Here are the top five managerial hires based on our metric:

This measure is far from perfect – injuries, front office decisions, and projection errors all affect outcomes – but it’s a straightforward way to estimate the influence a manager had on team success. And the leaderboard above passes the sniff test. Expand to the top ten and you’ll find reigning National League MOY Pat Murphy and World Series winning manager Dave Roberts as well.

Now that we have a simple measure defined, we can move on to the real question: do certain backgrounds or experiences actually make a manager more likely to exceed expectations?

Analysis

I tested the traits and experiences teams clearly value in a simple regression with our success metric to see which, if any, actually correlate with better outcomes. The full list of traits included:

• Former MLB Player

• First Time MLB Manager

• Internal vs External hire

• Age in first year

• ML Coaching experience

• MiLB Managing experience

• Front office experience

• Years of work experience

• Managerial tenure

The regression produced an R-squared of 0.18, indicating that our traits only explain about 18% of the variability in manager success. Given what we’re including, that isn’t necessarily surprising. That’s about what we should expect for something as context-dependent as managing.

When we break down individual traits, the picture gets clearer. The things that teams seemingly value don’t appear to make a meaningful difference to a manager’s success.

• Despite being the overwhelming preference (75% of hires), former MLB players don’t perform differently than non-players. In fact, they slightly underperform those that didn’t play at the highest level.

• First-time managers perform basically the same as veteran managers.

• External hires have been slightly more successful than internal promotions, although the effect is small.

• While teams strongly favor managers in their 40’s or 50’s with real work experience, the data suggest that age and total years of experience have effectively no impact.

• Neither prior Major League coaching experience nor Minor League managing experience predicts success.

In fact, there were only two things that show any faint signal.

The first is previous front office experience. In our data, “front office” covers a wide range of backgrounds – including advance scouts, special assistants, farm directors, and assistant general managers – and often represents a very small portion of prior work experience. But managers who spent at least some time in a front office role outperformed expectations more than those who didn’t.

I wouldn’t overstate the effect or declare FO experience a necessity in hires. Still, it’s at least directionally interesting – especially given the lack of signal with everything else tested – and it lines up with the way the game is evolving. I’d frame it this way: exposure to the front office in some capacity tends to lead to slightly better performance.

The second thing that shows signal is managerial tenure – for every additional year of tenure, managers outperform expectations by an additional two wins. That doesn’t mean that teams should hire someone and leave them in the role forever – although there might be an argument to stick with promising candidates a little longer – but rather that the industry does a reasonable job of deciding who to retain and who to let go once they’ve been hired. Tenure isn’t a predictor as much as a reflection of performance.

Conclusion

If we zoom out, the only consistent theme is that these observable traits don’t tell us much about which managers will outperform expectations. And honestly, it’s not especially surprising that they don’t.

The takeaway here isn’t that managers don’t matter. They do — but while manager performance is real, it’s extremely hard to quantify. Our success measure is simple by design, and it may not capture the nuance, context, and day-to-day dynamics that shape outcomes.

There also might not be enough variation in the candidate pool to produce measurable differences. Teams hire from a relatively narrow pipeline, so it’s not shocking that the pipeline produces similar results.

But the bigger point, in my view, is that relying on past experience is a poor way to predict future success. HBR and Forbes have been publishing research on that very idea for years. The resume almost certainly doesn’t matter as much as the soft skills that allow a manager to establish a culture, communicate effectively, and put players in a position to be successful.

In Part I of this series, we identified the visible traits that define who teams hire. In Part II, we tested whether those traits correlate with success. In Part III, we’ll look at what happens after the hire — how long teams stick with managers, how quickly they move on, and what that churn says about how the industry evaluates the job.

The manager is arguably the most important non-playing position in a Major League organization. It’s certainly the most visible, and when you combine the influence on players, in-game decisions, and culture, it’s hard to overstate how critical it is to have the right person in that role.

When the 2026 season begins, 10 teams will have a different manager than the one they started last season with. That’s literally one third of the league! The final opening was filled earlier today when the Colorado Rockies announced they would permanently install Warren Schaeffer as their manager.

With so many openings being filled, Ken Rosenthal recently wrote an article questioning the hiring strategy this offseason:

All industries should welcome new ways of thinking. Baseball has done a good job of that over the past two decades, benefiting from advances in analytics and technology. Yet there is still risk in trying to be too creative with hires. It’s doubtful all of them will work out as intended.

What exactly is it some teams want, other than to do something different? Beats me. Their supposed boldness goes only so far.

His article got me thinking about managerial hires. Who do teams hire? What makes a successful hire? How long does a manager have to make things work?

This will be part I of a multi-part series on managerial hires. Today, I’m going to try to answer a simple question: Who do teams actually hire?

Before we dive in, some quick things about the data:

• I looked at every managerial hire that Major League teams made from 2000 through this offseason. Managers hired prior to 2000 that managed during this century (such as Bobby Cox with the Braves or Joe Torre with the Yankees) were not included. The year 2000 is an arbitrary cutoff, but it’s a reasonable dividing line for what’s broadly considered the “modern” era of Major League front offices and their hiring practices. That gives us 182 hires composed of 132 different individuals.

• I excluded interim managers. The purpose here is to evaluate who teams actively chose to hire, not which names happened to be available when a mid-season change occurred. Interim stints do not count toward “prior managerial experience” either.

With that out of the way, let’s profile our managerial hires.

Professional Playing Background

If you want to manage in the Major Leagues, there’s one nearly universal prerequisite: you must be a former professional baseball player.

Almost 75% of hires played in the Major Leagues. And while that requirement has seemed to fade a bit in recent years (56% of hires since 2020), there have only been five hires this century who didn’t at least play minor league baseball: Mike Shildt (twice), Dave Trembley, Carlos Tosca, and the Giants hiring of Tony Vitello this offseason.

The recent hiring trend is more aligned with what we see in the NFL and NBA, where only around one-third of the current head coaches played at the highest level. But compared to those leagues, MLB remains deeply tethered to professional playing as a prerequisite.

Internal or External?

Any hiring process involves a basic choice: promote someone internally or bring in someone from the outside. There are benefits to continuity with known quantities. There is also the allure of bringing in a fresh voice. For Major League teams, it doesn’t seem to be that much of a choice.

In two-thirds of instances, teams have opted to hire someone from outside the organization. That makes sense when you consider the competitive context. Most teams hiring a manager are doing so after a disappointing season. On average, these clubs won only 75 games the year before they hired a new manager.

Internal hires seem to more frequently happen in extenuating circumstances like when:

• A manager retires (e.g., Walt Weiss taking over for Brian Snitker),

• A manager leaves for more money (e.g., Pat Murphy following Craig Counsell), or

• A team feels it has a better option in-house (e.g., David Ross over Joe Maddon).

The data roughly reflects these teams aren’t positioned as poorly. Teams averaged 80 wins the season before making an internal hire and just 72 wins before an external hire.

Age

Managerial opportunities overwhelmingly go to mid-career candidates. The average age of managers in their first season is 51.6. Removing non-first timers, the average falls but only slightly – to 47.7 – indicating that even first-time hires skew toward more seasoned candidates. Here’s the distribution among all hires:

More than three-quarters (77.5%) of hires fall between 40 and 59 in their first season. The hiring of an extremely young or old manager still makes headlines for a reason: it’s rare.

First-Time Managers vs. Returning Managers

One of the simplest ways to examine the hiring landscape is to look at whether teams prefer first-time managers or someone who has previously held the job. Since 2000, 56% of hires have been first-timers, with the rest being managers returning to the post for another go-around. That split has remained stable over time. In other words, “retreads” aren’t outliers - they’re a consistent part of how teams fill these roles.

Previous Work Experience

We know that nearly half of our hires previously managed in the Major Leagues and that almost every one of them played professionally. But what does the rest of their post-playing resume look like?

• 80% coached in the Major Leagues for at least one season.

• 56% managed at least one season in the Minor Leagues.

• 37% had spent at least one year in the front office in some capacity (most often as a Special Assistant).

Nothing here is particularly surprising. Teams value candidates who have been in a big-league clubhouse, have run a team (at some level), and/or have exposure to the front office.

In fact, only 14 hires (7%) lacked any experience as a manager (Major or Minor) or a Major League coach. Interestingly, three of those hires – Tony Vitello, Kurt Suzuki, and Craig Stammen – have occurred this offseason.

How Much Experience

We covered what experience they had, but what about how much experience managers had?

On average, first time managers had 12 years of professional experience (managing, coaching, front office) before they were hired. That number has dipped very slightly in recent years – to just over 11.5 since 2020 – but the drop is modest, and sample size is small. By and large, teams want candidates that have real off-field experience, which makes sense when we consider the typical ages of our hires. If we look at the extremes, we see that:

• 27 first-time managers (26.5%) had five or fewer years of experience. All 27 were former Major League players (save Tony Vitello).

• 17 first-time managers (16.7%) had 20 or more years of professional experience. These hires represent the other end of the spectrum: “baseball lifers” who finally got a shot. Only six of those 17 played in the Major Leagues.

Conclusion

Pulling that information together, the composite profile that teams have hired this century looks something like this:

• Former professional baseball player (most likely reached the Major Leagues)

• Finished playing around 12 years ago

• Mid-to-late 40’s or early-50’s

• Several years of MLB coaching and/or MiLB managerial experience

• Hired from outside the organization

From that lens, I don’t know if this season’s hires are terribly outside the box. Eight of the ten will be between 41 and 56 in 2026, five played in the Major Leagues, and nine played professional baseball in some capacity.

The two most “unique” hires are Blake Butera with the Nationals and Tony Vitello with the Giants. Butera is one of only three managers hired this century to be 35 or younger in his first year - though he also hits many traditional markers: he’s a former minor league player, he managed in the minor leagues, and he also has experience as a farm director in the eight years since he finished playing.

Vitello is unusual in that he brings zero experience – playing or working – with Major League teams. But his extensive college coaching background overlaps meaningfully with both front-office and minor league managerial work, and he’s got far more practice with media and fan attention than an average first-time hire.

For all the talk about innovation, Major League teams seem to consistently target the same type of manager they always have. That could be for good reason. Today’s goal wasn’t to judge - simply to understand who gets hired.

In Part II, we’ll look at something far more complicated: what, if anything, has actually worked when hiring a manager.