Bases Gained (BG) is a proprietary offensive contribution metric developed by The Baseball Nerd. It counts every direct action a player takes to advance himself and his teammates, applies a weighted penalty for caught stealing and a flat penalty for ground ball double plays, and adjusts for defensive position using 10 years of qualified-season data. The formula is transparent. The methodology is proprietary. The all-time leader in the dataset is Aaron Judge 2024 at 653.04 BG.
What Every Other Stat Gets Wrong
The existing metrics are not broken. They are excellent at what they were designed to measure. The problem is what they were not designed to measure.
Total Bases counts contact and stops at the moment of impact. A player who walks, steals second, and scores on a sacrifice fly contributed three advancement events to his team. Total Bases credits him with zero.
wOBA applies empirically derived run values to every plate appearance and is the most accurate rate stat publicly available. It deliberately excludes stolen bases from the batter's line. It does not penalize him for hitting into double plays. A player who steals 40 bases and a player who never runs finish with the same wOBA if their contact and walk numbers match.
wRC+ builds on wOBA, adjusts for park and league, and is the preferred tool for cross-era comparisons. It is a rate stat. It does not price caught stealing or double plays. It cannot tell you what a player cost his team in damaged rallies.
BG answers the question none of them are asking: across every direct action this player took, how many total bases of advancement did he generate for his team, against the full cost of his most damaging actions?
What Each Component Captures
The Caught Stealing Penalty
A flat -1 per caught stealing was the starting point and was rejected for two reasons. The actual cost is higher: a CS removes the runner entirely rather than simply costing a base. And a flat penalty treats a player stealing at 90% career rate identically to one stealing at 40%. They are not identical.
The tier structure applies the penalty based on career attempt volume, because single-season SB% is too noisy to reflect true skill reliably.
| Tier | Career Attempts | Formula |
|---|---|---|
| No activity | 0 | No penalty |
| Small sample | 1-4 | CS x -2 (flat) |
| Active | 5-30 | CS x -2 x (1 - Career SB%) |
| Elite | 31+ | CS x -2 x (1 - Career SB%) |
The result: Ronald Acuña Jr. with 73 stolen bases and a 10-catch season takes a penalty of just -1.6. Rougned Odor with 12 stolen bases and 12 caught stealings at a 60.5% career rate takes -9.49. BG prices the difference because the outcome for the team is not the same.
The Positional Adjustment
A catcher producing 400 raw BG is contributing something categorically different from a first baseman at the same number. The positional adjustment closes the gap between what each position historically produces among qualified hitters and what the average qualified hitter produces. It is derived from 10 years of BG distributions and does not borrow values from any other framework.
Catchers and middle infielders receive positive adjustments because they produce the least raw offense among qualified hitters. Designated hitters and corner outfielders carry negative adjustments because the position demands elite offensive output. Left field is the effective baseline position, sitting closest to zero adjustment. The full positional adjustment values are available to paid subscribers.
Left field at near-zero is effectively the baseline position. DH at a significant DH penalty reflects the highest offensive standard any position faces: no defensive obligation, justify the roster spot entirely with the bat.
The Data Behind It
BG inputs were built from three independent data sources and verified for consistency before any leaderboard was published. The stolen base and caught stealing data across FanGraphs, Baseball Prospectus, and Baseball Reference matched in 99.7% of 1,378 qualified player-seasons. CS data was identical across all three sources in every single record.
The 0.785 correlation with wRC+ confirms BG is related to the best existing public rate stat. The gap between 1.000 and 0.785 is where BG adds information. Stolen base volume, double play damage, positional context, and situational RE24 are all doing work that wRC+ does not do.
Raleigh Outscored Judge in 2025. Here Is Why.
This is the result that makes BG worth understanding. In 2025, Aaron Judge had a wRC+ of 204, one of the highest single-season marks in the dataset. Cal Raleigh had a wRC+ of 161. Judge was the better hitter by every rate measure. He won the AL MVP.
BG puts Raleigh ahead of Judge by 27.35 points. The argument is specific and the math is transparent.
Cal Raleigh C SEA
Aaron Judge RF NYY
Strip out position and penalties and look at raw production. Judge at 615.82 raw BG outproduced Raleigh at 575.36 by 40 points. That is a right fielder with a wRC+ of 204 being better than a catcher with a wRC+ of 161. BG agrees completely with that.
The positional swing of nearly 68 points reflects 10 years of historical data showing that catchers as a group produce the least offense of any qualified position, while right fielders produce the most. The GIDP gap adds another 20 points: Judge hit into 16 double plays against Raleigh's 6. wRC+ charged him nothing for that. BG charged him -20.
BG is not saying Judge was a bad player. It is saying Raleigh contributed more to his team's run production over the full 2025 season. Those are different claims. The first is obviously false. The second is what BG is designed to evaluate.
"Judge was the better hitter. Raleigh was the more complete contributor. BG measures the second thing."
The Baseball Nerd, 2026What a Good BG Score Looks Like
BG tiers recalibrate each season based on the percentile distribution of all qualified hitters. The cutoffs below are 2025 full-season finals.
| Tier | 2025 BG | Description | 2025 Examples |
|---|---|---|---|
| Elite | 491.9+ | Top 5%. MVP conversation. | Raleigh, Judge, Schwarber |
| Great | 403.8 - 491.9 | Top 25%. All-Star caliber. | Tatis, Ramirez, Carroll |
| Above Average | 352.1 - 403.8 | Above median. Productive starter. | Guerrero Jr., Nimmo |
| Average | 293.5 - 352.1 | League average contribution. | Albies, Kwan, Frelick |
| Below Average | 249.2 - 293.5 | Lineup liability. | Semien, Steer |
| Struggling | Below 249.2 | Bottom 10%. Roster concern. | Castellanos, Sheets |
The All-Time Leaderboard
| # | Player | Season | Pos | BG | wRC+ |
|---|---|---|---|---|---|
| 1 | Aaron Judge | 2024 | RF | 653.04 | 205 |
| 2 | Shohei Ohtani | 2024 | DH | 650.75 | 192 |
| 3 | Bobby Witt Jr. | 2024 | SS | 619.53 | 169 |
| 4 | Ronald Acuña Jr. | 2023 | RF | 618.35 | 170 |
| 5 | Aaron Judge | 2022 | RF | 616.17 | 211 |
| 6 | Cal Raleigh | 2025 | C | 613.45 | 161 |
| 7 | Mike Trout | 2016 | CF | 603.08 | 174 |
| 8 | Cody Bellinger | 2019 | CF | 587.59 | 167 |
| 9 | Aaron Judge | 2025 | RF | 586.10 | 204 |
| 10 | Matt Olson | 2023 | 1B | 579.61 | 150 |
Bobby Witt Jr. at third all-time with a 2024 wRC+ of 169 is the most instructive result on this leaderboard. His 37 stolen bases, 73 non-HR RBI, and shortstop positional bonus (a shortstop positional bonus) combine to produce a score that no single public metric surfaces in one place. BG found it because it was designed to find exactly that combination.
What BG Does Not Measure
BG is a direct-action offensive framework. It measures what a player personally did to advance himself and his teammates at the plate and on the basepaths. Several real contributions fall outside its boundary by design.
It is not park-adjusted. Coors Field inflates everything. BG records what actually happened. A park-adjusted version is a documented future extension.
It is not predictive. A player who hits into 31 double plays might be unlucky on contact. BG records the result and leaves the cause to the evaluator.
It does not measure defense. The positional adjustment accounts for the offensive demands of each position at the population level. Individual defensive performance is not part of the formula.
Non-scoring sacrifice bunts and productive outs that do not produce an RBI are excluded. These require play-by-play intent data that public sources do not reliably provide. The exclusion applies uniformly and does not distort comparisons across qualified hitters.
How to Use It
BG is most useful alongside wRC+ and FIP rather than instead of them. The 0.785 correlation with wRC+ tells you they substantially agree on who the best hitters are. The gap tells you where the most useful information lives.
A player ranking materially higher in BG than wRC+ is producing baserunning and situational value that rate stats cannot see. A player ranking materially lower in BG than wRC+ is carrying penalty weight that rate stats do not charge. Junior Caminero in 2025 had a wRC+ of 112 and a GIDP total of 31, a -62 penalty that his rate stats never acknowledged. BG did.
The live leaderboard is available at tbnbasesgained.netlify.app, updated weekly through the 2026 season. Click any player to see the full component breakdown.
Frequently Asked Questions
What is Bases Gained in baseball?
Bases Gained (BG) is a proprietary offensive contribution metric developed by The Baseball Nerd. It counts every direct action a player takes to advance himself and his teammates, applies a weighted penalty for caught stealing based on career stolen base rate, applies a flat penalty for ground ball double plays, and adjusts for defensive position using 10 years of qualified-season data. The full methodology and formula are proprietary to The Baseball Nerd.
Why did Cal Raleigh outscore Aaron Judge in 2025?
Raleigh scored 613.45 BG against Judge's 586.10. In raw production before adjustments, Judge leads Raleigh by 40 points. The positional adjustment swings nearly 68 points: catchers receive a significant catcher bonus because they produce the least offense of any qualified position, right fielders receive a right field penalty because they produce the most. Judge also hit into 16 double plays against Raleigh's 6, a -20 point gap in the GIDP penalty that no public metric charges. BG is not saying Judge was a worse hitter. It is saying Raleigh contributed more to his team's run production over the full season.
Is Bases Gained park-adjusted?
No. BG records what actually happened in actual games. A hitter at Coors Field benefits from that environment and BG reflects it. This is a deliberate design choice. A park-adjusted version of BG is a documented future extension of the framework.
How was Bases Gained verified?
BG inputs were verified across FanGraphs, Baseball Prospectus, and Baseball Reference. Caught stealing data matched 100% across all three sources across 1,378 qualified player-seasons. Stolen base data matched in 99.7% of cases. The baserunning component correlates at 0.740 with Baseball Prospectus's independently derived Downgraded Runs on Basepaths and 0.751 with their Stolen Base Runs Above Average, confirming that two separate frameworks built by different organizations identify the same players as elite and poor baserunners.
What is the highest single-season Bases Gained score in the dataset?
Aaron Judge 2024 leads all seasons at 653.04 BG. Judge combined 411 total bases, 79 walks, 10 stolen bases, a +48.3 RE24, and only 6 double plays. His right field penalty is the steepest in the formula and he still leads all qualified seasons in the 2015-2025 dataset.
Where can I find the Bases Gained leaderboard?
The live leaderboard is at tbnbasesgained.netlify.app. It covers the 2026 current season updated weekly and full 2024 and 2025 final seasons. Every player row links to a full component breakdown showing exactly how the BG score was assembled.
The Baseball Nerd maintains a white paper and a changelog for Bases Gained. The white paper covers the full methodology, formula derivation, component rationale, positional adjustment construction, and cross-validation findings. The changelog documents any updates to the framework over time. Both documents can be reviewed by request. Contact information is at thebaseballnerd.com/about.