A field guide to Trevor Story's stolen-base stats

Put a denominator under the steal total

The first calculation is small: successful steals divided by successes plus caught stealings. Using that completed 2025 season, 31 divided by 32 is 96.875 percent, or 96.9 percent rounded to one decimal place. That is our arithmetic on the recorded line, not another statistic supplied by the source. The denominator matters. Thirty-one successes out of thirty-two recorded attempts says something different from thirty-one out of forty.

There was a delightful run of tidiness before that final season line. MLB’s September 18, 2025 report records Story beginning the season with 31 consecutive successful stolen-base attempts, an American League record. The streak is a historical achievement. It does not make a later attempt automatic, and its charm survives perfectly well without pretending perfection lasted forever.

When reading a fresh table, match both the season and game type before pairing its SB and CS columns, because a regular-season steals total joined to a postseason caught-stealing count would describe a sample that nobody actually played. Check the labels first. A neat decimal cannot rescue a mismatched denominator.

Sprint speed has a different measuring tape

Story’s Baseball Savant profile gives a useful comparison: 29.6 feet per second in 2018 and 28.6 in 2025. Yet the recorded steal totals for those years were 27 and 31 respectively. The later season had more steals despite the lower sprint-speed measurement. That observation does not establish the cause of the change. It does show why the two columns should not be treated as synonyms.

Savant's sprint-speed methodology measures the fastest one-second portion of selected runs, averaging approximately the best two-thirds for a seasonal figure. Qualifying plays include certain runs of two or more bases and home-to-first runs on weakly hit or topped balls. This is not a stopwatch timing of every steal from first to second.

A recorded speed can therefore be valuable without answering whether a particular attempt was wise. The selection of runs, the unit and the season all belong beside it. I would use those dated numbers to describe a measured tool, then use the attempt ledger to describe what happened when Story tried to steal. Neither one needs to win an argument against the other.

An out costs more than a pretty total admits

FanGraphs’ Weighted Stolen Base Runs explanation makes the next distinction explicit. Its wSB model gives value to steals, assigns a cost to caught stealings and compares the result with league-average activity given opportunities. Zero is average on that scale. This is an estimate in runs, not a percentage and not a count of how often a player looked fast.

That is why two runners with the same number of steals can have different value. The outs matter, and so do the opportunities. A success percentage helps explain efficiency; a run-value model asks how much the activity helped compared with its baseline. Those are related answers, but replacing one with the other would hide useful information.

Leave room for the bases nobody stole

Savant's Extra Bases Taken Run Value measures advancement against fielders on batted balls and explicitly excludes stolen bases. It models an opportunity using such inputs as runner position, speed, the outfielder's arm and the outfielder's distance from the ball and bases. A trip from first to third on a hit belongs to a different assessment from a steal.

Here is the useful distinction when a profile displays both measures: check what each includes before adding anything together. An extra-base-running panel is not automatically a complete baserunning total. A steals panel does not record every helpful advance either. Their labels are the instructions for reading them, and a shared word such as running does not make their definitions interchangeable.

We can now give Story's 2025 line a fair description: unusually efficient recorded stealing, alongside a speed measurement that was below his earlier peak. The evidence does not tell us how much of that efficiency came from any particular tactical choice. For the broader player profile, the major in this package connects his running with his hitting and infield work.

Keep SB and CS together, keep the season attached to the speed number, and read the scope of any run-value panel. The next time a baseball quietly disappears from first and turns up at second, the box score will have somewhere sensible to put both the pleasure and the risk.