How Rare Was Baltimore’s 49-Win Climb?

Baltimore improved from 52 wins in 2021 to 83 in 2022 and 101 in 2023. Those were annual gains of 31 and 18 wins, producing a 49-win improvement from the first season to the third. Among the documented comparisons examined here, it is the largest endpoint gain—and an unusually sustained climb—but the available evidence does not establish it as the all-time MLB record. MLB two-season win improvement comparisons require separate definitions for endpoint gains, consecutive increases and schedules of different lengths.

The short answer: Baltimore gained 49 wins, but the all-time claim remains unverified

The calculation is straightforward:

• 2021 to 2022: 83 − 52 = +31 wins
• 2022 to 2023: 101 − 83 = +18 wins
• 2021 to 2023: 101 − 52 = +49 wins

MLB’s review of major divisional-era turnarounds confirms Baltimore’s three season totals and the two annual increases: 52 wins in 2021, 83 in 2022 and 101 in 2023. The Orioles therefore gained 49 wins between the endpoints while improving in both steps of the three-season window (MLB).

That combination matters. Baltimore did not make one dramatic leap and then surrender part of it. The club followed its 31-win breakthrough with another 18-win increase, reaching 101 victories in the third season.

A community discussion about Baltimore’s rise poses the natural record-book question: Was this the largest two-season improvement ever? The discussion does not verify the answer. Nor can a selected list of divisional-era turnarounds do so, because neither source represents a complete query of every eligible major-league team window.

The supported verdict is narrower: the 2021–2023 Orioles produced a 49-win endpoint gain and a sustained climb across both annual steps. That is exceptional, but it is not enough by itself to assign an all-time rank.

What counts as a two-season improvement?

The broadest definition compares a team’s wins in the first and third seasons of a three-season window:

Endpoint win improvement = W(N+2) − W(N)

For Baltimore:

101 − 52 = 49 wins

This measure answers a specific question: How many more games did the club win two seasons after the baseline year? It does not require the path between those endpoints to move steadily upward.

A stricter test asks whether the team improved in each successive season:

• W(N+1) > W(N)
• W(N+2) > W(N+1)

Those definitions create three useful trajectory labels:

• Sustained climb: The team wins more games in each successive season.
• Breakthrough followed by partial regression: The team makes a large first-year jump, then loses some ground while still finishing above its baseline.
• No positive endpoint gain: The ending season does not exceed the baseline, regardless of the intermediate result.

Consider a hypothetical sequence of 65, 100 and 85 wins. The endpoint gain is still 20 wins because 85 exceeds 65. But the team did not sustain its climb: it lost 15 wins between the second and third seasons.

Baltimore qualifies under both measures. Its endpoint change was positive, and each annual step was also positive.

One-year turnaround records must remain separate. They compare N with N+1, while a two-season endpoint ranking compares N with N+2. A club can lead one measure without leading the other, especially if it gives back much of its initial gain in the following season.

A worked comparison of four notable three-season windows

The table below contains four selected comparisons, ordered by endpoint gain. It is not an all-time leaderboard. MLB’s underlying review covers 14 divisional-era teams that gained at least 27 wins in one year, excluding comparisons before and after shortened seasons; these four windows illustrate how endpoint gains and trajectories can differ (MLB).

Team and window	Annual changes	Endpoint change	Trajectory
Orioles, 2021–2023: 52, 83, 101 wins	+31, +18	+49	Sustained climb
Braves, 1990–1992: 65, 94, 98 wins	+29, +4	+33	Sustained climb
Royals, 2023–2025: 56, 86, 82 wins	+30, −4	+26	Breakthrough followed by partial regression
Diamondbacks, 1998–2000: 65, 100, 85 wins	+35, −15	+20	Breakthrough followed by partial regression

Baltimore’s 49-win endpoint gain is 16 wins larger than Atlanta’s 33-win gain in this selected group. The difference is not merely one of scale. Baltimore and Atlanta continued moving upward after their breakthrough seasons, whereas Kansas City and Arizona retained only part of their initial gains.

“Regression” therefore should not be treated as an all-or-nothing result. Arizona’s decline from 100 wins to 85 erased 15 wins from its breakthrough, but the Diamondbacks still finished 20 wins above their 65-win baseline. Kansas City fell from 86 to 82 while remaining 26 wins above its starting total of 56.

Those teams count as positive endpoint improvements. They do not count as sustained climbs.

Why sustained improvement is different from retaining a turnaround

Within MLB’s selected 14-team sample, only the 1990–1992 Braves and 2021–2023 Orioles increased their win totals again in the following full season. Baltimore added 18 wins after its initial 31-win improvement; Atlanta added four after first gaining 29.

That makes Baltimore one of two sustained climbs in this particular thresholded sample. It does not mean only two such climbs have occurred in major-league history. The review begins with the divisional era in 1969, requires an initial increase of at least 27 wins and excludes certain shortened-season comparisons.

Those boundaries omit several possible categories:

• Three-season windows before 1969
• Windows built from two substantial increases that were each below 27 wins
• Raw-win comparisons affected by shortened schedules
• Other positive endpoint gains that did not begin with a 27-win leap

The sample nevertheless reveals something useful about Baltimore’s trajectory. A dramatic first-year improvement does not guarantee another increase. Arizona and Kansas City remained well above their baselines but won fewer games in the third season. Baltimore and Atlanta both moved higher.

The records describe outcomes, not causes. Win totals alone cannot prove that a particular player, roster decision, health pattern or close-game performance caused a turnaround. Those explanations require evidence beyond the standings.

Raw wins and winning percentage answer different questions

Raw wins are the clearest measure when the seasons being compared had comparable schedules. Baltimore’s rise from 52 to 101 is readily interpretable because the totals come from completed full-length seasons.

Schedules have varied across baseball history, however. A team playing a 154-game schedule had fewer opportunities to accumulate wins than one playing 162. Shortened seasons create an even larger distortion. For cross-schedule comparisons, winning-percentage change provides a necessary secondary measure:

Endpoint winning-percentage change = PCT(N+2) − PCT(N)

The calculation should use the club’s completed record and actual games played. A projected final record cannot be substituted for history.

Schedule situation	Primary treatment	Secondary treatment
Comparable full schedules	Compare raw-win change	Also report percentage change
Unequal historical schedules	Emphasize percentage change	Show wins with schedule context
Shortened seasons	Exclude or flag in raw-win rankings	Use actual percentage and label the season

A 162-game equivalent can help translate a winning percentage into familiar terms, but it is only a normalization. It does not represent games actually played and should never be inserted into a historical leaderboard as a real win total.

Final standings also should not be replaced by other measures. Expected records estimate results from runs scored and allowed. Relative Power Index incorporates schedule-related factors. In-season standings remain incomplete. Each may answer a useful question, but none measures how many games a club actually won between two completed seasons.

What a reproducible all-time ranking would require

SABR describes the freely available Lahman Baseball Database, whose January 2026 release contains historical team standings through the completed 2025 season. That provides a suitable foundation for an all-time team-window analysis, although the database page does not itself name a record holder.

A reproducible protocol should establish its rules before calculating the rankings:

• Scope: Include completed team seasons in leagues classified as major leagues within the selected database release.
• Window construction: Match each franchise’s seasons N, N+1 and N+2, following the database’s franchise continuity across relocations and name changes.
• Raw-win schedule groups: Rank 162-game, 154-game and other standard schedule formats separately. Exclude officially shortened seasons from those full-schedule tables.
• Winning percentage: Use actual completed records. Where ties exist, calculate percentage as (W + 0.5 × T) ÷ (W + L + T) and state that convention.
• Shortened seasons: Include them only in a separately flagged percentage comparison, not in an unqualified raw-win ranking.
• Minimum games: Require a completed team season in the chosen dataset; disclose its games played rather than converting it into a fictional full schedule.
• Ties in rankings: Preserve shared places when the underlying values are equal. Compare exact values rather than rounded display percentages.

Each eligible window should report:

• Baseline, intermediate and ending records
• Actual games played in all three seasons
• First annual win change
• Second annual win change
• Endpoint win change
• Endpoint winning-percentage change
• Schedule classification and any exclusion flag
• Trajectory label

The results should then be divided into two rankings. One should include every positive endpoint improvement, including teams that declined in the third season but remained above their baselines. The other should include only sustained climbs in which wins increased in both annual steps.

For raw wins, separate leaderboards by standard schedule length would avoid treating 154- and 162-game opportunities as identical. A broader winning-percentage table could then compare across schedule formats, with shortened seasons plainly identified rather than silently normalized.

Incomplete 2026 standings do not belong in such a ranking because they are not final and the cited database release ends with 2025. On the documented evidence through that release, Baltimore’s 49-win gain is the strongest of the selected comparisons and one of only two sustained climbs in MLB’s thresholded 14-team sample. Determining whether it stands first across the full major-league record requires the defined all-team query rather than an extrapolation from that sample.