Who has the highest WAR in consecutive seasons?

Headline answers

Which MLB player has had the highest WAR in consecutive seasons? Short answer: it depends on how you read the question and which era you allow. Walter Johnson is the answer to most versions of it, and Babe Ruth is the best hitter. A handful of 1880s pitchers beat everyone on a two-year basis. The details are below.

Data: Lahman 13.0.0 (seasons 1871-2024, via the R Lahman package) joined by playerID to Baseball-Reference WAR (daily files, downloaded 2026-10-03).

1. Best two-year stretch

• Overall, Old Hoss Radbourn 1883-84, 33.8 WAR (14.3 + 19.4). Walter Johnson 1912-13 is next with 32.0 and Jim Devlin 1876-77 with 31.7. Eleven of the top 15 are 1876-1899 pitchers.
• Best since 1900: Walter Johnson 1912-13 (32.0).
• Best hitter: Babe Ruth 1923-24 (25.5), then Barry Bonds 2001-02 (23.6) and Carl Yastrzemski 1967-68 (22.9).
• Best since 1947: Bonds 2001-02 (23.6), barely ahead of Bob Gibson 1968-69 (23.2).

2. Best 3- and 5-year stretches

• Best 3-year: Walter Johnson 1912-14 (45.1), then Radbourn 1883-85 (42.3). Ruth's 1919-21 (34.4) is #7. The best since 1947 is Gibson 1968-70 (33.2), then Mays 1963-65 (32.9) and Bonds 2001-03 (32.8).
• Best 5-year: Walter Johnson 1912-16, 69.1 WAR, far ahead of Ruth 1920-24 (56.3), Clarkson 1885-89 (55.0), Radbourn (54.9) and Cy Young 1892-96 (54.3). Mays 1962-66 (52.3) and Bonds 2000-04 (51.1) follow. Since 1947 it runs Mays, Bonds, Joe Morgan 1972-76 (47.8), and Mike Trout 2012-16 (47.1) is #3 since 1969.
• Johnson also holds the 7-year (89.4) and 10-year (120.3) records.

3. Streaks above a threshold

Only one player in history has 10 straight 8+ seasons. Four have 7 or more.

4. Both seasons elite

• With both seasons at 10+ WAR there are only 41 pairs by 24 players. Radbourn 1883-84 (33.8) and Walter Johnson 1912-13 (32.0) lead. The best by a hitter is Ruth 1923-24 (25.5).
• Ranked by the weaker season (the floor), Walter Johnson 1912-13 is #1 (15.4 / 16.6). Next are Radbourn, Tommy Bond, Devlin, then Ruth 1920-21 (11.8 / 12.6) and Bonds 2001-02 (11.9 / 11.7).
• There are just 31 pairs (19 players) where the player was #1 in MLB both years. They include Ruth 1926-27, Bonds 2001-02 and Gibson 1968-69.

TL;DR for each split

• Hitters vs. pitchers: Pitchers own the all-time lists. The best hitters are Ruth (2-year 25.5; 5-year 56.3), Mays 1962-66, Bonds 2000-04 and Yastrzemski 1967-68. Hitters hold the longest elite streaks, and Aaron's 17 is the longest. Of the two-way players, Ohtani 2020-24 has 37.2 WAR over five years and four straight 8+ seasons (2021-24). The 1880s pitcher-fielders Bob Caruthers and Jim Whitney post bigger numbers.
• Era: The leader changes with the cutoff. For two-year stretches it goes Johnson (1901+), Ruth (1920+), Bonds (1947+) and Bonds (1969+). For five-year stretches it goes Johnson, Ruth, Mays, Bonds. The all-eras ranking is a pre-1900 pitcher ranking, and from 1920 on the answer is Ruth, Mays or Bonds depending on the length.
• Position (best 5-year): C Gary Carter 1981-85 (33.9), 1B Lou Gehrig 1927-31 (47.1), 2B Rogers Hornsby 1921-25 (50.1), 3B Alex Rodriguez 2001-05 (42.5), SS Honus Wagner 1905-09 (49.2), OF Ruth (56.3), DH Frank Thomas 1991-95 (31.8), P Walter Johnson (69.1). Catchers and DHs are far behind. The positional adjustment and games played probably both play a part, but I haven't separated them.
• Decade: The record-holder for each decade is in the splits tables. The best 3-year stretch is Johnson in the 1910s (45.1) and the 1880s pitchers. After 1920 it sits in a band of about 26-34: Ruth 1926-28 (34.1), Gibson 1968-70, Bonds 2001-03, Mantle 1955-57. Trout's 2014-16 stretch is 27.6 and Ohtani's 2022-24 is 28.6.
• Age at start: The best streaks start at 24 (Johnson) or 27-28 (Devlin, Radbourn, Ruth 1923-27). Hitters can start later. Ruth's best five-year stretch beginning at 31 or older is 1926-30 (52.7).
• Franchise (one-team windows): Washington/Minnesota (Johnson), NY Yankees (Ruth), Providence (Radbourn), Cleveland Spiders (Young), San Francisco (Mays), Pittsburgh (Wagner), St. Louis (Hornsby), Anaheim/LA Angels (Trout 47.1) and Cincinnati (Morgan) lead their franchises in the 5-year table.
• League: AL's best is Johnson and NL's best is Clarkson (5-year) or Radbourn (2-year). "Other (AA/UA/PL/FL)" is the American Association and the other short-lived leagues, led by Keefe, Hecker and Caruthers.
• Peak vs. longevity (peak_vs_longevity.csv, chart 08): Johnson is balanced (41% of career WAR in his best five years). Ruth, Bonds, Mays and Aaron are longevity cases at about a third each. Sandy Koufax (80%), Hal Newhouser, Ron Santo, Jackie Robinson and Mookie Betts are peak-heavy. Trout is at 55%.

Anything surprising

1. Hank Aaron is nowhere near the top of the best-window lists (his best five years is 43.8) but wins the "most straight 5+ WAR seasons" list with 17. He is the longevity champion with a modest peak.
2. Walter Johnson holds the record at every window length from 3 to 13 seasons. His 10-year 8+ streak is a record by three seasons.
3. Seven 19th-century pitchers, plus Walter Johnson, outscore Ruth's best two years (Ruth is #8), and 11 of the top 15 two-year stretches are from before 1900. The era cutoff is the single biggest decision in this project.
4. Service adjustment barely changes the top 10s. It does lift Ted Williams' best five years from 41.3 (#48) to 49.2 (#14) once 1943-45 are bridged (1941-48), and Bob Feller's from 35.9 to 39.5. The pace adjustment for 1918/19/81/94/95/2020 doesn't change who is in any top-10 list (it only reorders a few places at the 3- and 10-year lengths).
5. Ohtani is the only modern two-way player on the elite lists, and by total WAR he is still far behind Ruth's two-way production (Ruth's 1919-21 is 34.4).
6. Trout's 2012-16 (47.1) ranks #22 all-time for five years.

Charts

WAR Streak Explorer (interactive)

The extra feature I added. Search a player, drag the window length, see their best run, rank, and streak twins.

It is embedded in output/WAR_streaks_report.html, and also saved as its own offline file, output/parts/streak_explorer.html (about 260 KB, no server or internet needed). It holds all 864 players with 30+ career WAR.

• Pick a player and drag a slider for window length (1-20 seasons).
• It highlights their best run of consecutive seasons on a season-by-season "barcode", with missing years marked.
• It shows their all-time rank and what percentage of the record they hit.
• It finds their streak twins: the three players whose best window had the closest year-by-year WAR shape. You can overlay a twin's line on theirs.
• There is a toggle to stop military service from breaking a streak.
• A "crown" strip shows who holds the record at each of the 20 lengths. The record passes from Tim Keefe (1 season) to Radbourn (2) to Walter Johnson (3-13) to Ruth (14-20).

Why I picked it: a ranking table answers one question, and the real issue here is that the answer changes with the window length. Making k a slider turns that into something you can feel. The streak twins also add "who did this look like?", which a leaderboard can't. I checked it against the R results: Johnson's five-year record is 69.1, Ruth's is 56.3, and Williams is 49.2 versus 41.3 when service is bridged.

Key tables

Click a heading to expand. The complete set (47 tables, including every split) is in WAR_streaks_tables.xlsx, next to this file.

Method & caveats

Where the WAR comes from

Lahman has no WAR, so I used Baseball-Reference WAR ("bWAR") from their public daily files:

• https://www.baseball-reference.com/data/war_daily_bat.txt (position players, plus pitchers' hitting)
• https://www.baseball-reference.com/data/war_daily_pitch.txt (pitching)

The download worked, so nothing was substituted. The raw files are saved in data/raw/, so re-runs don't change unless you delete them.

• Total WAR = bat file WAR + pitch file WAR, summed per player-season. This is how B-Ref totals a player. It matters for two-way players: Ruth comes to 182.6 WAR (162.3 as a hitter plus 20.4 as a pitcher). The often-quoted 162.1 is his position-player total only.
• Joined on People$bbrefID = B-Ref player_ID. Only 4 B-Ref MLB players lacked a Lahman match (0.7 WAR in total), listed in output/parts/tables/diag_unmatched_war_players.csv.
• Version drift: B-Ref does not publish WAR version numbers and revises the methodology now and then. These results are a snapshot of the files as served on 2026-10-03. I compared 16 icons' career totals with the B-Ref figures I know. 15 matched within 3 WAR. Ruth was 20 off, which is the hitting-only versus total difference above. See sanity_career_totals.csv. Treat those reference numbers as approximate.
• Range: the files also contain 2025-26. I stopped at 2024 because Lahman 13.0.0 does. Ohtani's and Judge's 2025 seasons, for example, are not counted. Change MAX_YEAR in 00_config.R once Lahman catches up.
• What counts as MLB: NL, AL, AA, UA, PL and FL. I excluded the 1871-75 National Association (so the data starts in 1876) and the Negro Leagues. The WAR files include the Negro Leagues, but Lahman does not, and their shortened schedules make season-to-season streaks hard to compare.
• 1880s pitching WAR is huge (Radbourn threw 678 IP in 1884). Those seasons are legitimate under B-Ref's method, but they dominate every all-time ranking, which is why the era splits matter so much.

Definitions

• Consecutive means calendar-consecutive seasons with an MLB player-season in the data. A missed year breaks the streak. Traded players' stints are summed into one season.
• Window of k seasons is any k back-to-back seasons. "Best" is the highest total WAR, and each player appears once, at their best window. Tables are a1_* (k=2) and a2_* (k=3, 5, plus bonus lengths 7 and 10).
• Threshold streak is a maximal run of consecutive seasons at or above a WAR threshold (tables a3_*, thresholds 4, 5, 6, 7, 8, 10).
• Both-elite pair (a4_*) uses six definitions:
  • both seasons at or above 6, 8 or 10 WAR;
  • both seasons top-5 in MLB, or both #1 in MLB;
  • the highest "floor" (the weaker of the two seasons).
• Hitter / Pitcher / Two-way. Each season is Two-way if the player threw 50+ IP and played 30+ games at non-pitcher positions (DH included) that year. It is Pitcher if they threw 50+ IP otherwise, and Hitter otherwise. A streak is labelled "Hitter" or "Pitcher" only if every season in it is. Anything else, including mixed streaks like Ruth's 1919-21, is "Two-way / mixed". Total WAR always counts both halves, so nobody loses credit. Ruth's best hitters-only stretches are in the "Hitter" rows.
• Eras. A streak belongs to an era only if all of its seasons lie on one side of the cutoff. Streaks that straddle it are left out of the pre/post comparison. Results are shown for four cutoffs, 1901, 1920, 1947, 1969, in every *_splits.csv and in era_cutoff_sensitivity.csv.
  • Primary cutoff: 1947. Integration is when MLB first drew on its full talent pool, and it also sits after the dead-ball era and the 1893 pitching-distance change.
  • 1901 is the AL's start and the 20th century. 1920 is the live-ball era. 1969 is the mound-lowering and expansion.
• Edge cases:
  • Shortened seasons (1918, 1919, 1981, 1994, 1995, 2020) are flagged. The main tables use raw WAR. A pace-adjusted sensitivity scales those seasons to full length (2020 x2.7).
  • Military service: a curated list of 13 documented cases (data/reference/military_service_gaps.csv) lets a "service-adjusted" mode bridge those exact gap years. Partial seasons (Williams 1952-53, Mays 1952) are in the data as small seasons, so they count as played years. Every other gap breaks the streak.
  • Gaps: sanity_best_gap_pairs_not_counted.csv shows the best elite pairs that were not counted because a year is missing.

Sanity checks

output/parts/sanity_checks_report.txt lists all 22 checks, and all pass. They cover:

• Structure: one row per player-season, WAR equals bat plus pitch everywhere, no missing WAR, leagues limited to MLB.
• Career totals within 8 WAR of B-Ref for Ruth, Bonds, Mays, Mantle, Gibson, Clemens, Maddux, Trout, Williams, Cy Young, Walter Johnson, Aaron, Cobb, Musial, Wagner and Randy Johnson.
• Ranks: Ruth is the #1 hitter for 2 years and #2 overall for 5 years. Bonds is a top-3 post-1946 3-year stretch and #10 overall for 5 years.
• Edge cases: trades and stints, two-way labels (Ruth 1918-19, Ohtani 2021-23), shortened-season flags, the Williams and Mays service gaps, and the exclusion of gapped pairs.

Three of my first-draft checks failed and I changed them. I had expected Ruth to be a top-3 two-year player overall and Bonds a top-10 three-year player overall, which ignored the 1880s pitchers. I also had Ruth's career total at 162.1 when that is his hitting-only figure. I rewrote those checks to what the data should show, and explained that in the script comments.