49 steals and 10 caught stealing is an 83.1% success rate
Stolen base percentage is successful steals divided by total steal attempts. The MLB glossary defines it the same way: stolen bases divided by stolen base attempts. Nasim Nuñez of Washington led MLB with 49 steals in 2026 and was caught 10 times, so his rate is 49 ÷ (49 + 10) = 83.1%.
The rate alone does not tell you whether the running helped. A caught stealing erases a runner and adds an out, so it costs more than a stolen base gains. Run FanGraphs' wSB formula through the 2026 MLB run environment and one caught stealing wipes out the value of 2.06 steals, which puts the break-even rate at 67.3%. Nuñez's line added about 5.7 runs. Among the 89 MLB players with at least 15 attempts, his 83.1% ranks 27th, well above the middle of the pack, and because he ran so often his run value is still among the best in the league. The calculator above shows the rate alongside that break-even rate, SB runs, the 2026 league average and the player rank.
Why check the success rate against break-even
The MLB glossary calls stolen base percentage a key tool for evaluating base stealers and points out that the stolen base leader is often caught a lot as well. A big steal total can still cost runs when the caught stealing pile up with it. Looking at the rate and the break-even together gives a clearer answer in cases like these.
- Comparing base stealers: in 2026, Chandler Simpson (TB) stole 46 bases but was caught 14 times, a 76.7% rate worth +3.4 runs. Bobby Witt Jr. (KC) stole one fewer base, 45, yet was caught only 8 times (84.9%) and added +5.7 runs.
- Deciding when to run: managers and coaches can check whether a runner should keep going at his current rate and how many more times he can be caught before he drops below break-even.
- Amateur, rec league and youth stats: enter each player's steals and caught stealing from the scorebook to get individual rates and a team total in one place.
- Fantasy baseball and broadcast numbers: see whether an "80% success rate" is actually high for that league and season, and where it ranks among 2026 players with 15 or more attempts.
How to use it
- Pick a league: KBO, MLB or NPB. The break-even rate, league average and player rank switch to that league's 2026 regular-season values.
- In the first row of the player ledger, enter a name, stolen bases and caught stealing. If your stats only list attempts, tap "Attempts" at the top right of the ledger to turn the second column into attempts.
- To compare more players, add rows with "Add player". You can enter up to 10, and once there are two or more rows a total row appears with the combined team rate.
- Tap the number in front of a row (Σ for the total) to point the verdict and the tile strip at that row. On the success rate ruler below, the ledger's row numbers are placed on the 2026 player distribution.
- "Copy results" copies the whole ledger and the league values as text. "Copy link" copies an address that reopens exactly what you entered.
How it is calculated
- Success rate: SB ÷ (SB + CS). If you enter attempts, CS = attempts − SB. The result is shown as a percentage to one decimal place.
- SB runs: SB × 0.2 − CS × |runCS|. FanGraphs values a stolen base at 0.2 runs in every season and sets the cost of a caught stealing at runCS = −(2 × runs per out + 0.075). Runs per out is league runs ÷ league outs for that season.
- Break-even rate: |runCS| ÷ (0.2 + |runCS|). At this rate SB runs come out to exactly 0. For 2026 MLB it is 0.412 ÷ 0.612 = 67.3%.
- To break-even: above break-even, the calculator counts how many more times the runner can be caught, with steals unchanged, while SB runs stay at 0 or higher. Below it, the calculator counts how many straight steals it takes to get back to 0. Both are whole numbers.
- Next target: the number of straight steals needed to reach the next 5% step above the current rate (80%, 85%, 90% and so on).
- Total: adds up every row's steals and caught stealing, then divides. It does not average the players' rates.
| Line (2026 MLB values) | SB | CS | SB% | SB runs | To break-even |
|---|---|---|---|---|---|
| SB leader Nasim Nuñez | 49 | 10 | 83.1% | +5.7 runs | Can be caught 13 more times and stay at or above |
| Chandler Simpson | 46 | 14 | 76.7% | +3.4 runs | Can be caught 8 more times and stay at or above |
| Perfect on few attempts | 8 | 0 | 100.0% | +1.6 runs | Can be caught 3 more times and stay at or above |
| Safe half the time | 10 | 10 | 50.0% | −2.1 runs | 11 straight steals to reach break-even |
The tile strip in the calculator draws this run math as squares. In 2026 MLB one caught stealing erases 2.06 steals, so Nuñez's 10 caught stealing erase 20.6 steals' worth and leave 28.4. Those 28.4 steals × 0.2 = +5.7 runs. To lift his rate to the next target of 85%, he would need 8 straight steals.
Break-even rate by league (2026 regular season)
The cost of a caught stealing moves with the value of an out. The more runs a league scores, the more it hurts to give up an out, and the higher the break-even rate climbs. In 2026 the KBO, the highest-scoring of the three leagues, had the highest break-even at 69.5%, while pitcher-friendly NPB had the lowest at 63.2%.
| League | Runs | Outs | Runs per out | 1 CS | Break-even | 1 CS = steals | League SB% |
|---|---|---|---|---|---|---|---|
| KBO | 7,301 | 38,282 | 0.191 | −0.456 runs | 69.5% | 2.28 | 74.5% (904 / 1,213) |
| MLB | 21,769 | 129,239 | 0.168 | −0.412 runs | 67.3% | 2.06 | 77.0% (3,289 / 4,273) |
| NPB | 6,168 | 45,828 | 0.135 | −0.344 runs | 63.2% | 1.72 | 70.4% (814 / 1,156) |
MLB's −0.412 matches the 2026 runCS that FanGraphs publishes. For the KBO and NPB, each league's runs (the sum of runs allowed by all teams) and outs (innings pitched × 3) from the official records go into the same formula. FanGraphs offers the rule of thumb that "you want to have at least twice as many SB as CS to break even" and adds that the number bounces around with the run environment. The MLB glossary says a base stealer with an SB% of 75 or higher is typically helping his team. Instead of a rule of thumb, the calculator uses the break-even rate for the league and season you pick.
2026 stolen base leaders: success rate and SB runs
Steal totals and SB runs do not rank players in the same order. The MLB player with the most SB runs was José Ramírez (CLE), who was only seventh in steals. He stole 37 bases and was caught just 4 times, a 90.2% rate worth +5.8 runs, slightly more than the +5.7 of steals leader Nasim Nuñez.
| League | Player (team) | SB | CS | SB% | SB runs |
|---|---|---|---|---|---|
| MLB | Nasim Nuñez (WSH) | 49 | 10 | 83.1% | +5.7 |
| MLB | Chandler Simpson (TB) | 46 | 14 | 76.7% | +3.4 |
| MLB | Bobby Witt Jr. (KC) | 45 | 8 | 84.9% | +5.7 |
| MLB | José Ramírez (CLE) | 37 | 4 | 90.2% | +5.8 |
| KBO | Hwang Seong Bin (Lotte) | 55 | 11 | 83.3% | +6.0 |
| KBO | Park Min Woo (NC) | 50 | 12 | 80.6% | +4.5 |
| KBO | Park Hae Min (LG) | 42 | 17 | 71.2% | +0.6 |
| NPB | Yukihiro Iwata (Yakult) | 39 | 8 | 83.0% | +5.0 |
| NPB | Shunsuke Urata (Yomiuri) | 36 | 12 | 75.0% | +3.1 |
| NPB | Ukyo Shuto (SoftBank) | 30 | 5 | 85.7% | +4.3 |
Among players with 15 or more attempts in 2026, 13 of 89 in MLB finished below break-even, compared with 1 of 21 in the KBO and 4 of 22 in NPB. Most frequent base stealers clear break-even, but how far they clear it makes a large difference in SB runs.
Team steals: getting caught less beats running more
The gaps grow at the team level. In 2026 Washington and Tampa Bay each attempted 218 steals. Washington was safe 178 times (81.7%) for +19.1 runs, and Tampa Bay was safe 164 times (75.2%) for +10.6. With the same number of attempts, a difference of 14 steals turned into a gap of more than 8 runs. Houston stole 50 bases and was caught 22 times, and its 69.4% rate sat just above the 67.3% break-even for +0.9 runs.
The KBO showed the same pattern. NC attempted 214 steals and made 154 of them, but 60 caught stealing left it at 72.0% and +3.4 runs. Doosan attempted only 146, going 117 safe and 29 caught (80.1%) for +10.2 runs. NC stole 37 more bases, yet Doosan gained nearly 7 more runs. KT (61 steals, 38 caught, 61.6%) stayed below break-even and lost 5.1 runs on the bases. In NPB, Yomiuri (99 steals, 43 caught, 69.7%) added +5.0 runs while Yakult (98 steals, 28 caught, 77.8%) added +10.0. The "10 teams" example loads the 10 teams with the most attempts in the chosen league into the ledger.
MLB success rates jumped in 2023
Starting in 2023, MLB enlarged the bases from 15-inch to 18-inch squares and, together with the pitch timer, limited pitchers to two pickoff attempts or disengagements per plate appearance. If a third pickoff attempt fails, the runner automatically moves up one base. That season, stolen base attempts rose from 1.36 to 1.80 per game, and the league success rate climbed from 75.4% to 80.2%.
| MLB season | SB | CS | SB% | Attempts per game (both teams) | runCS | Break-even |
|---|---|---|---|---|---|---|
| 2019 | 2,280 | 832 | 73.3% | 1.28 | −0.435 | 68.5% |
| 2021 | 2,213 | 711 | 75.7% | 1.20 | −0.419 | 67.7% |
| 2022 | 2,486 | 811 | 75.4% | 1.36 | −0.397 | 66.5% |
| 2023 | 3,503 | 866 | 80.2% | 1.80 | −0.422 | 67.8% |
| 2024 | 3,617 | 961 | 79.0% | 1.88 | −0.405 | 66.9% |
| 2025 | 3,440 | 989 | 77.7% | 1.82 | −0.410 | 67.2% |
| 2026 | 3,289 | 984 | 77.0% | 1.76 | −0.412 | 67.3% |
2020 is left out because it was a 60-game season. The success rate has eased a little since 2023, but the break-even rate has barely moved, staying between 66% and 68%. In 2026 the league average still sits about 10 percentage points above break-even.
Common mistakes
- Counting every pickoff as a caught stealing: according to the MLB glossary, a runner who is picked off after he has already broken for the next base is charged with a caught stealing, but a runner picked off while getting back to his original base is not. The KBO records site counts pickoffs separately from caught stealing.
- Counting bases taken on a wild pitch or passed ball as steals: if the runner was not already stealing, it is not a stolen base. A runner thrown out trying to advance on a wild pitch or passed ball is not charged with a caught stealing either.
- Counting defensive indifference as a steal: when the defense makes no effort to stop the runner, as often happens late in games, the advance is not scored as a stolen base.
- Averaging players' rates: average a player with 1 steal and 0 caught (100%) and one with 30 steals and 15 caught (66.7%) and you get 83.3%, but the two players combined are 31 ÷ 46 = 67.4%. The total row always adds first and divides second.
- Treating a perfect rate on a few attempts as the best: 8 steals and 0 caught are worth +1.6 runs, while Nuñez's 49 steals and 10 caught are worth +5.7 with 2026 MLB values. The success rate measures efficiency; SB runs combine efficiency and volume.
- Using another league's break-even: the same 68% is above break-even in 2026 NPB (63.2%) and MLB (67.3%) but below it in the KBO (69.5%). Choose the league your stats come from.
Limitations
- SB runs are a context-neutral calculation that gives every attempt the same value. FanGraphs also notes that wSB does not consider the game situation. The real break-even point shifts with the base-out state, for example when a runner tries to steal third with two outs.
- The 0.2 runs and runCS are linear-weight estimates. There is no adjustment for park, opposing pitcher or catcher.
- The calculator's SB runs use 0 as the baseline. FanGraphs wSB also subtracts the runs a league-average runner would have produced with the same times on base, so its values differ slightly.
- The KBO and NPB runCS values are not published by FanGraphs. They come from plugging each league's records into FanGraphs' MLB formula.
- League totals and player lists are final regular-season stats checked on 2026-10-23. The postseason is not included, and the NPB player list was built from team-by-team player stats after the season, keeping only players with 15 or more attempts.
- What you enter is not sent to a server; everything is calculated in this browser. Values are added to the address only when you change them or press "Copy link".
As of The formulas and definitions were checked on 2026-10-23 against the MLB glossary, the MLB.com rule-change article and the FanGraphs Library. League totals and player stats were calculated the same day from 2026 regular-season figures on the KBO records site, MLB's official stats (Stats API), NPB's official records and Sports Navi team player stats.
Sources
- 01MLB Glossary — Stolen-base Percentage (SB%)Stolen bases ÷ total attempts; a caught stealing hurts more than a steal helps; rule of thumb of 75% or higher
- 02MLB Glossary — Stolen Base (SB)Not a steal: advancing on a wild pitch or passed ball when not already stealing, advancing on an error, another runner thrown out on a double steal, defensive indifference
- 03MLB Glossary — Caught Stealing (CS)Picked off after breaking for the next base counts; picked off returning to the base or thrown out on a wild pitch does not
- 04MLB Glossary — Caught Stealing Percentage (CS%)Catcher's caught stealing ÷ steal attempts against that catcher
- 05FanGraphs Library — wSBMatt Klaassen (2013-02-10): runSB 0.2, runCS = −(2 × runs per out + 0.075), twice as many SB as CS to break even, context neutral
- 06FanGraphs — Guts! wOBA and FIP ConstantsrunCS by season: 2026 −.412, 2025 −.410, 2024 −.405, 2023 −.422, 2022 −.397, 2021 −.419, 2019 −.435
- 07MLB.com — Pitch timer, shift restrictions among announced rule changes for '23Anthony Castrovince: bases from 15 to 18 inches, two pickoffs or disengagements per plate appearance
- 08MLB Glossary — Pitch TimerThe runner advances automatically when a third pickoff attempt fails
- 09KBO official records — team baserunning2026 regular season, 10 teams: 904 SB and 309 CS (1,213 attempts)
- 10KBO official records — player baserunning21 players with 15 or more attempts in the 2026 regular season
- 11KBO official records — team pitching2026 regular season: 7,301 runs allowed in 12,760⅔ innings (38,282 outs)
- 12MLB Stats API — 2026 team hitting30 teams: 3,289 SB and 984 CS; team pitching (group=pitching): 21,769 runs and 129,239 outs; 2019 to 2025 from the season parameter of the same address
- 13NPB — 2026 Central League team pitching (2026年度 セントラル・リーグ チーム投手成績)With the Pacific League page (tmp_p.html), final figures after all games: 12 teams, 6,168 runs allowed and 45,828 outs
- 14Sports Navi (スポーツナビ) — team batting statsPlayer steals and caught stealing summed over 12 teams: 814 SB and 342 CS; 22 players with 15 or more attempts
- 15Pro Yakyu Data Freak (プロ野球データFreak) — team battingFinal team steal totals, checked against the Sports Navi sums for all 12 teams
Frequently asked questions
How do you calculate stolen base percentage?
Divide successful stolen bases by total steal attempts (stolen bases + caught stealing). The MLB glossary uses the same definition. A runner with 49 steals and 10 caught stealing is at 49 ÷ 59 = 83.1%. If your stats only list attempts, caught stealing = attempts − stolen bases.
What is a good stolen base percentage in MLB?
Start with the break-even rate, the point where steals stop adding runs. With the 2026 MLB run environment it is 67.3%, and anything above it adds runs. The 2026 MLB league average was 77.0%, and the MLB glossary's rule of thumb is that a base stealer at 75% or higher is typically helping his team.
Who has the best stolen base percentage?
Among 2026 MLB players with 15 or more attempts, Bryson Stott (PHI) had the best rate: 27 steals without being caught, 100%. Measured by run value, José Ramírez (CLE) led MLB with +5.8 runs on 37 steals and 4 caught stealing (90.2%), ahead of Stott's +5.4. The calculator shows where any line you enter would rank among them.
Why is the break-even rate different in each league?
A caught stealing gives up an out, so it costs more in leagues that score more runs. FanGraphs sets the cost of a caught stealing at −(2 × runs per out + 0.075) and the value of a steal at 0.2 runs. In 2026 the KBO scored 0.191 runs per out for a 69.5% break-even, MLB 0.168 for 67.3%, and NPB 0.135 for 63.2%.
Does a pickoff count as caught stealing?
If the runner had already broken for the next base when he was picked off, it is scored as a caught stealing. If he was picked off while getting back to his original base, it is not. That is how the MLB glossary explains it. The KBO records site counts pickoffs separately from caught stealing, so enter only the caught stealing figure instead of adding the two.
How do you calculate a team's stolen base percentage?
Add up every player's stolen bases and caught stealing, then divide. Averaging the players' rates gives too much weight to players with only a few attempts. With two or more rows in the ledger, the total row does this for you, and you can also enter a team's totals in a single row.
Is defensive indifference or a wild pitch a stolen base?
No. When the defense makes no attempt to stop the runner, as often happens late in games, the advance is not scored as a steal, and a base taken on a wild pitch or passed ball is not a steal unless the runner was already stealing. A base taken on an error is never a steal. A runner thrown out trying to advance on a wild pitch is not charged with a caught stealing either.
Can I calculate a catcher's caught stealing percentage?
A catcher's CS% is the runners he threw out divided by the steal attempts against him. Enter the steals he allowed and the runners he caught in the ledger, and 100% minus the success rate is his caught stealing percentage. Runners who advanced on wild pitches or passed balls are not included in a catcher's CS%.