ABS Is Rewiring Umpire Behavior, Just Not the Way Anyone Expected

When MLB introduced ABS challenges in 2026, the obvious hypothesis was that catchers would steal fewer strikes. If a batter can challenge a borderline called strike and get it overturned, framing becomes less valuable. Six weeks of data says that isn't happening. What is happening is something more interesting.

The Numbers Through May 7

All data below is regular season only, comparing 2025 and 2026 through the same date — May 7.

The headline: umpires are dramatically more accurate at calling strikes on pitches in the zone. They are not meaningfully more accurate at calling balls on pitches out of the zone. Those two things are moving independently, and it's the in-zone accuracy driving everything.

Stolen Strikes Didn't Go Down

The intuitive assumption going in was that challenges would reduce stolen strikes — if a batter can dispute a borderline called strike and get it overturned, catchers who used to steal those calls would lose that edge. But stolen strikes per game in 2026 (8.51) are actually higher than 2025 (7.46). Out-of-zone call accuracy is essentially flat year-over-year.

Lost strikes are a completely different story. In-zone pitches called balls dropped from 4.11 per game in 2025 to 2.15 in 2026. That 48% decline is big enough to be real, and it's coming from umpires, not challenges.

The Effect Is Concentrated in Two-Strike Counts

Breaking accuracy down by count reveals where the change is actually coming from. Umpires are calling strikes at a much higher rate on in-zone pitches specifically in two-strike counts — the exact situations where they historically gave batters the most benefit of the doubt.

The intuition behind this makes sense. When a pitcher challenges a ball call on a pitch that was clearly in the zone, the umpire looks bad in real time. On two-strike counts, the stakes of that miss are highest — it's the difference between striking a batter out and extending the at-bat. Umpires appear to be responding by calling more strikes when the pitch is in the zone, rather than giving the batter the benefit of the doubt.

Three-Ball Counts Didn't Move

The other place where count bias historically showed up was three-ball counts — umpires were reluctant to issue walk calls on close pitches when the count was already 3-0 or 3-1. If ABS challenges were correcting that bias too, we'd expect out-of-zone accuracy to improve on those counts. It didn't.

The three-ball count numbers are essentially flat. On 3-0 counts, out-of-zone accuracy actually dropped slightly (81.4% to 79.6%). Umpires are not more willing to call ball four on borderline pitches in 2026. The bias toward pitchers in full-count situations doesn't appear to be correcting, at least not yet.

The improvement on non-two-strike counts (the right chart above) is also real but modest — 2 to 4 percentage points on most counts. The two-strike effect is in a different tier.

So Why Didn't Called Strikeouts Go Up?

Better in-zone accuracy on two-strike counts should produce more called strikeouts. It mostly didn't.

Called strikeouts per game in 2026 (3.99) are virtually identical to 2025 (3.97). Swinging strikeouts are up a little (12.48 to 12.75). The share of strikeouts that are called actually fell from 24.2% to 23.8%.

The most likely explanation is batter adaptation. Hitters appear to have already adjusted to the tighter two-strike zone — swinging more on borderline pitches they might have taken a year ago, knowing the umpire is more likely to ring them up. That converts potential called strikeouts into swinging strikeouts, or contact, keeping the called K total flat even as the underlying accuracy improves.

The Big Picture

Six weeks into 2026, ABS challenges are doing something, just not what most people assumed. Stolen strikes haven't declined. Catchers are still framing effectively. What has changed is umpire behavior on in-zone pitches, specifically in two-strike counts where the historical bias was the strongest.

Lost strikes — pitches clearly in the zone that get called balls — are down almost 50%. That's a meaningful correction to one of the most consistent biases in umpiring. Whether it's umpires responding to the threat of challenges, or some broader recalibration that comes with having a parallel automated system in the room, the in-zone accuracy improvement is too large to be noise.

Three-ball count bias hasn't budged yet. Umpires are still reluctant to walk batters on borderline pitches. If that's going to change, it likely needs more time — or more pitchers willing to challenge those calls.

We'll revisit this analysis at the end of the season when we have a full picture of how it all played out.

© 2026 Malter Analytics · MalterAnalytics.com · Data sourced from MalterAnalytics mlb_pbp_prepped, through May 7, 2026.