MLB Umpire Stats

MLB Umpire Stats: Runs & Over Rates by Home-Plate Umpire

Q: Where does this data come from?

From our own game database: finished MLB games since 2017 with the home-plate umpire assignment recorded per game. It refreshes automatically as new games settle, so the table stays current through the season.

Every home-plate umpire calls a slightly different strike zone — a tighter zone means more walks, deeper counts and more runs; a generous zone speeds hitters into outs. Over a large sample that shows up as a measurably different run environment per umpire, which is why totals bettors track these numbers. The honest caveat: even a full season behind the plate is a small sample, and most of the spread you see below is noise around the league average. We use umpire tendencies as one small input to our MLB totals model — never as a betting rule on their own.

How our MLB model works →Public track record →

#	Umpire	Games	Avg total runs	Over 8.5 %
1	Vic Carapazza	148	9.00	48.6%
2	Edwin Moscoso	145	9.12	54.5%
3	Bill Miller	144	8.63	47.2%
4	Adam Hamari	142	8.62	49.3%
5	Dan Bellino	141	9.70	58.9%
6	Ryan Wills	141	8.76	46.1%

Method: built from the finished MLB games in our own model database (2017 onwards), using the recorded home-plate assignment for each game. "Avg total runs" is the average combined final score in that umpire's games; "Over 8.5 %" is the share of his games with 9+ total runs. Umpires with fewer than 30 plate games are excluded as too thin to report.

Frequently asked questions

Do MLB umpires really affect totals?

At the margins, yes. The strike zone is called by a human, and zone size correlates with walks, strikeouts and ultimately runs. But the effect is small compared to pitchers, lineups, weather and park — an umpire is a minor adjustment, not a handicapping shortcut.

Should I bet the over when a high-runs umpire is behind the plate?

Not on its own. Sportsbooks see the same public data, and a single season of plate games is a noisy sample — an umpire at 55% overs may just have caught high-scoring matchups. We feed umpire tendencies into our model as one small input alongside dozens of stronger signals.

Where does this data come from?