Home Cooking Is Real (Sort Of): 12,449 Games of Evidence

It's the oldest accusation in baseball: the umpire favors the home team. The crowd noise gets in his head. He squeezes the visiting pitcher. He gives the home team's ace an extra inch on the outside corner. Every road fan has felt it. Every home fan has denied it.

We tested it with 12,449 games of Statcast-verified pitch data across five seasons (2021-2025). Here's what the data says.

How We Measured It

For every game in our database, we calculate net favor — the cumulative run value that shifted from one team to the other due to missed calls. Positive net favor means the home team benefited; negative means the away team benefited.

We also calculate net WPA favor — the same concept, but weighted by game context. A missed call in the 9th inning of a tie game counts more than one in the 3rd inning of a blowout.

If umpires systematically favor the home team, we'd expect the league-wide average of net favor to be consistently positive across seasons. If it's random, the average should hover near zero.

The 2025 Team-Level Picture

Here's every team's net ump run favor for 2025, from most favored to most squeezed:

| Team | Run Favor | WPA Favor | Games | |------|-----------|-----------|-------| | TOR | +47.3 | +3.26 | 178 | | SF | +24.7 | +1.94 | 162 | | NYY | +16.2 | -0.97 | 169 | | CLE | +15.4 | +2.24 | 163 | | COL | +12.6 | -0.72 | 161 | | AZ | +12.4 | -0.65 | 161 | | DET | +7.7 | -0.13 | 170 | | CHC | +6.5 | -0.33 | 168 | | STL | +5.9 | +2.57 | 162 | | TEX | +4.9 | -0.56 | 162 | | SEA | +3.2 | -0.42 | 174 | | WSH | +1.9 | +0.62 | 160 | | KC | -0.2 | -0.64 | 161 | | PHI | -0.7 | -0.90 | 165 | | BOS | -1.0 | -1.07 | 165 | | HOU | -1.1 | -0.20 | 160 | | MIL | -1.5 | +0.59 | 170 | | SD | -1.7 | -0.13 | 164 | | NYM | -2.1 | +0.43 | 161 | | OAK | -3.1 | -1.60 | 162 | | CIN | -5.9 | +0.35 | 163 | | ATL | -6.3 | -0.89 | 162 | | LAD | -12.3 | -0.37 | 177 | | PIT | -14.5 | +0.63 | 161 | | MIN | -16.8 | -0.13 | 162 | | MIA | -17.0 | -0.47 | 162 | | TB | -17.1 | -0.27 | 161 | | CWS | -17.8 | +0.02 | 162 | | LAA | -18.9 | -0.25 | 160 | | BAL | -20.7 | -1.69 | 162 |

The spread is enormous. Toronto received +47.3 runs of ump favor in 2025 — a full starting pitcher's worth of value delivered by incorrect calls. Baltimore was on the other end at -20.7 runs. That's a 68-run gap between the most and least favored teams.

But Is It "Home Cooking"?

Here's the critical distinction: these numbers measure net favor for each team across all their games — both home and away. A team with positive net favor got more favorable calls overall, regardless of whether they were at home or on the road.

True "home cooking" would mean umpires systematically favor whichever team is at home. To test that, you'd look at the league-wide aggregate: across all 12,449 games, does the home team get a positive net favor on average?

The answer is complicated by the fact that net favor is calculated from the home team's perspective by definition. What matters is whether the distribution is centered above zero. Our data shows the effect, if it exists, is small — far smaller than the team-to-team variance within a single season.

What Explains the Team-Level Differences?

If it's not home cooking, why does Toronto get +47.3 and Baltimore get -20.7?

Several factors likely contribute:

Pitcher repertoire and location. Pitchers who consistently hit the edges of the zone — painting corners with elite command — will naturally accumulate more borderline calls that go their way. The Blue Jays' rotation in 2025 featured Kevin Gausman (+17.3 runs of gift favor), José Berríos (+14.9), and Chris Bassitt (+14.1). Three pitchers who live on the edge, and all three ranked in the top 10 for most gifted calls league-wide.

Catcher framing. Though we don't directly measure framing in our model, a catcher who presents borderline pitches well may influence which side of the line an umpire's call lands on. The effect shows up in our data as differential gift rates for pitchers throwing to different catchers.

Random variation. Over 162 games, even perfectly random umpiring will produce team-level variance. Some teams will be on the lucky end; some won't. The magnitude of the variance we see — a 68-run spread — is large enough to raise questions, but season-to-season consistency (or lack thereof) can distinguish signal from noise.

Cross-Year Consistency

If team-level ump favor were mostly random, you'd expect teams to bounce around the rankings year to year. If it reflects something structural — pitcher command, catcher framing, park effects — you'd see consistency.

The data shows a mix. Some teams are consistently favored or squeezed:

• LAD was the most squeezed team in 2024 (-27.0 runs) and heavily squeezed in 2025 (-12.3). Two straight bottom-5 finishes.
• BAL was bottom-3 both years: -14.4 in 2024, -20.7 in 2025.
• CLE was top-5 both years: +30.7 in 2024, +15.4 in 2025.
• NYY was top-5 both years: +7.8 in 2024, +16.2 in 2025.

But others flip dramatically:

• TOR went from +2.1 (near neutral) in 2024 to +47.3 (runaway #1) in 2025.
• DET went from -4.7 (squeezed) in 2024 to +7.7 (favored) in 2025.

The teams with consistent multi-year trends likely reflect real roster effects — the same pitching staff locating to the same spots. The teams that flip are probably seeing rotation turnover, catcher changes, or plain variance.

The WPA Wrinkle

Run favor and WPA favor don't always tell the same story. The Yankees illustrate this perfectly: they received +16.2 runs of ump favor in 2025, but their WPA favor was -0.97. The calls went their way in terms of runs, but not in high-leverage moments. Their favorable calls came in blowouts and low-stakes situations.

Pittsburgh shows the opposite: -14.5 runs of favor (squeezed) but +0.63 WPA favor (the bad calls came in low-leverage spots). Getting squeezed in garbage time doesn't hurt your win-loss record.

This divergence is why we report both metrics. Run favor tells you the total magnitude of umpire impact. WPA favor tells you whether it mattered.

The Verdict

Is home cooking real? The team-level data tells a more nuanced story than the simple accusation suggests.

What's real: Some teams consistently receive more favorable calls than others, and the magnitude is large enough to matter — 47 runs for Toronto is not trivial. The effect correlates more with pitching staff characteristics than with home/away status.

What's not proven: A systematic league-wide bias toward the home team that operates independently of roster composition. The team-to-team variance dwarfs any potential home-field umpire effect.

What ABS changes: The challenge system gives both teams — home and away — equal access to corrections. If there is a home-field umpire effect, challenges should neutralize it. We'll be watching the 2026 data to find out.

---

See every team's ump favor breakdown on our teams pages, or explore the full umpire leaderboard to find which umpires produce the most total favor.