Hiroshima model ties count-dependent strike calls to the pitches batters let pass
Hiroshima University researchers found that borderline strike calls shift with the ball-strike count across more than 450,000 MLB pitches. A model that changes only the umpire's expected pitch locations reproduces much of that shift.
The Scientist · Science desk
Three figures: more than 450,000 MLB pitches analyzed; the count-aware model fit individual tracked calls better than a version that ignored the count at three of the four strike zone boundaries; 89 umpires analyzed at the outer boundary.
| Figure | Value | Note | Claim |
|---|---|---|---|
| MLB pitches analyzed, more than | 450,000 pitches | 2 | |
| Better-fit boundaries | 3 of four | 10 | |
| Umpires at outer boundary | 89 umpires | 12 |
What happened
- After a 3-0 count the next borderline pitch was more likely to be called a strike, and at 0-2 more likely a ball: the umpire's effective zone expanded at 3-0 and contracted at 0-2.
- The team used tracking data and calls from the 2015-2024 seasons, limited to four-seam fastballs from right-handed pitchers to right-handed batters.
- The model fit individual tracked calls better than a version that ignored the count at three of the four edges of the zone.
- A check on 89 umpires at the outer boundary found a modest association between less precise location judgments and heavier reliance on expectations.
Why it matters
- What it limits constraint Game records show the effective zone moving with the count but cannot separate what an umpire perceived from other influences on the decision, so the perceptual account stays a model that fits the calls.
- What it limits constraint Any table of called strikes by count partly describes batters, because their swing decisions set which borderline pitches the umpire gets to judge.
- Who has to decide decision The 2026 challenge system gives umpires immediate feedback on some calls, so before-and-after comparisons can now ask whether the zone depends less on the count.
Mitsuto Tomomura and Masahiro Shinya of Hiroshima University built their simulated umpire around one design choice: what it expects to see [3]. When expectations came from every pitch thrown, the model could not reproduce the expansion and contraction of the zone. When they came from the pitches batters let pass, it could [7].
The proposed explanation is that an uncertain view leaves room for experience to shift the perceived location of the ball [8]. An umpire who expects a pitch outside may see an ambiguous edge pitch as farther out. One who expects it nearer the center gets an estimate pulled inward, and a strike becomes more likely [8].
The fit is uneven. The model followed the umpires' shifts most closely at the upper and outer edges and struggled at the lower and inner ones [11]. Gains at three of four boundaries leave one boundary with no reported gain over the count-blind version [15].
In the 89-umpire test, perceptual uncertainty was estimated from the umpires' own calls [12][13]. We would treat the modest association as a consistency check. The same calls supply the uncertainty estimate and the behavior it is meant to predict.
In our view the evidence supports a modest version of the perceptual story. The written zone has not changed [1]. The analysis, as phys.org describes it, suggests the shifting calls may partly reflect how the brain combines uncertain visual information with expectations [2].
The thing this doesn't tell you is how other pitch types or other handedness pairings behave, since the analysis held both fixed [4]. The report does not say how many of the 450,000 pitches remained after that restriction [2].
What to watch
- Studies comparing called strikes before and after the 2026 challenge system, to see whether the count effect shrinks once umpires get immediate feedback.
- Tests of the same model on other pitch types and handedness pairings, which this analysis held fixed.
- A revised model that fits the lower and inner edges, where the current one struggled.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence62
- Adoption
- Not scored
- Hype gap−5
- Incentives
- Not scored
- Confidence55
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
The official strike zone has not changed; the umpire's judgment of borderline pitches has.
- [2]
An analysis of more than 450,000 Major League Baseball pitches suggests the shifting calls may partly reflect how the brain combines uncertain visual information with expectations.
- [3]
The study, published in Communications Psychology, used a model created by Mitsuto Tomomura and Masahiro Shinya of Hiroshima University that reproduced the call patterns across pitch counts.
- [4]
The researchers analyzed tracking data and calls from the 2015-2024 seasons and restricted the study to four-seam fastballs thrown by right-handed pitchers to right-handed batters, keeping pitch type and handedness consistent.
- [5]
The next borderline pitch after a 3-0 count was more likely to be called a strike (an expanded effective zone); at a 0-2 count the pitch was more likely to be called a ball (a contracted zone).
- [6]
With two strikes a batter has good reason to swing at a borderline pitch because letting it pass risks a strikeout, which leaves the umpire judging a different selection of pitches, tending farther outside the zone.
- [7]
Expectations based on every pitch thrown could not reproduce the observed expansion and contraction of the umpire's judgment, but expectations based on the pitches batters let pass could explain it.
- [8]
The proposed explanation is that an uncertain view leaves room for experience to influence the perceived location of the ball: if an umpire expects a pitch to be outside, an ambiguous pitch near the edge may seem farther out, and an expectation closer to the center could pull the estimate inward, making a strike call more likely.
- [9]
The researchers did not need to give the simulated umpire a different decision threshold for every count; changing the expected pitch locations was enough to reproduce much of the movement.
- [10]
The model fit the individual tracked calls better than a version that ignored the count at three of the four boundaries of the strike zone.
- [11]
The model captured the umpire's shifts in judgment most closely at the upper and outer edges but struggled at the lower and inner edges.
- [12]
The explanation predicts that umpires with less precise location judgments should lean more heavily on expectations; an analysis of 89 umpires making calls at the outer boundary found a modest association in that direction.
- [13]
The 89-umpire analysis used the umpires' calls to estimate perceptual uncertainty.
- [14]
The 89-umpire result supports the proposed mechanism without establishing what umpires actually saw; game records cannot isolate perception from other influences on a decision or show that the resulting calls were correct.
- [15]
Because the count-aware model fit better at three of the four boundaries, one boundary is left without a reported gain over the count-ignoring version.
- [16]
MLB's Automated Ball-Strike Challenge System, introduced in 2026, lets batters, pitchers and catchers contest calls using tracking technology, giving umpires immediate feedback on some disputed decisions.
- [17]
The researchers suggest that the challenge system's feedback could help umpires adjust, and that future studies comparing calls before and after its introduction could test whether those corrections make the strike zone less dependent on the count.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- phys.orgBall or strike? Why the count may change what a baseball umpire sees
1 article · October 12, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
Entities
- Major League BaseballFollow
- Automated Ball-Strike Challenge SystemFollow
- Masahiro ShinyaFollow
- Hiroshima UniversityFollow
- Mitsuto TomomuraFollow
- Communications PsychologyFollow