Brewers vs Mets Prediction, August 25, 2026 | Under 7.5
Kyle Harrison vs Zac Thornton at Citi Field
Published August 25, 2026 · MLB Prediction Data Lab
The posted total is 7.5. A naive add of season run rates (4.96 for Brewers, 4.17 for Mets) gets you 9.13 combined runs. That is not the bet. The bet is that Kyle Harrison and Zac Thornton plus Citi Field pull the night below that baseline. Partitioning each starter's own earned-run rate across expected innings and handing the remainder to each club's season staff rate returns 6.53 projected earned runs.
Full slate: MLB picks today, August 25, 2026.
The listed matchup
Milwaukee Brewers (81–50) at New York Mets (59–72), Citi Field, Flushing. Kyle Harrison, a left-hander, is listed for the visitors. Zac Thornton, a left-hander, is listed for the home club. Void the play if either listed starter is scratched.
| Side | Listed starter | Record | Park |
|---|---|---|---|
| Brewers | Kyle Harrison | 81–50 | Citi Field |
| Mets | Zac Thornton | 59–72 | Open, grass |
The two starters, measured per batter faced
Rate statistics expressed per nine innings reward pitchers who work deep. Expressing them as a share of batters faced removes that distortion. Kyle Harrison has faced 402 hitters. Zac Thornton has faced 200.
| Starter | BF | K% | BB% | K−BB | K/BB | Pitches/BF |
|---|---|---|---|---|---|---|
| Kyle Harrison | 402 | 30.3 | 6.5 | 23.9 | 4.69 | 4.22 |
| Zac Thornton | 200 | 18.0 | 7.0 | 11.0 | 2.57 | 3.69 |
| Starter | W-L | ERA | IP | WHIP | K/9 | BB/9 | HR/9 | Opp AVG |
|---|---|---|---|---|---|---|---|---|
| Kyle Harrison (L) | 10-3 | 2.84 | 98.1 | 1.11 | 11.17 | 2.38 | 1.19 | .223 |
| Zac Thornton (L) | 3-4 | 3.04 | 50.1 | 1.15 | 6.44 | 2.50 | 1.07 | .240 |
Kyle Harrison sits at a 2.84 ERA with a 1.11 WHIP and 26 walks in 98.1 innings. Zac Thornton sits at a 3.04 ERA with a 1.15 WHIP and 14 walks in 50.1 innings. Walks stabilize faster than earned runs, so the walk column is the more trustworthy August reading.
Direction inside the season
Season aggregates hide the last month. These are the prior starts, excluding today's outing so the table does not leak a result into a published prediction.
| Kyle Harrison — last starts | IP | ER | K | BB | H | HR | Pitches |
|---|---|---|---|---|---|---|---|
| 2026-07-03 | 2.2 | 3 | 3 | 1 | 5 | — | 72 |
| 2026-07-08 | 4.0 | 3 | 2 | — | 4 | 1 | 70 |
| 2026-08-05 | 5.0 | 0 | 10 | 1 | 1 | — | 69 |
| 2026-08-11 | 4.2 | 3 | 3 | 3 | 10 | 3 | 90 |
| 2026-08-18 | 5.0 | 0 | 8 | 2 | 2 | — | 79 |
| Zac Thornton — last starts | IP | ER | K | BB | H | HR | Pitches |
|---|---|---|---|---|---|---|---|
| 2026-07-27 | 6.1 | 3 | 5 | 1 | 6 | 1 | 81 |
| 2026-08-01 | 4.2 | 3 | 2 | 3 | 6 | 1 | 94 |
| 2026-08-07 | 5.0 | 0 | 3 | 2 | 4 | — | 63 |
| 2026-08-12 | 6.0 | 3 | 2 | 1 | 7 | 1 | 80 |
| 2026-08-18 | 5.0 | 3 | 7 | 1 | 7 | 2 | 99 |
The run environment, before the starters
Brewers have scored 650 runs in 131 games (4.96 per game) with a .733 OPS. Mets have scored 546 in 131 (4.17) with a .698 OPS. The pitching staffs have allowed 3.73 and 4.47 runs per game.
| Club | Record | R/G | OPS | Staff ERA | Staff WHIP | RA/G |
|---|---|---|---|---|---|---|
| Brewers | 81–50 | 4.96 | .733 | 3.42 | 1.17 | 3.73 |
| Mets | 59–72 | 4.17 | .698 | 4.07 | 1.30 | 4.47 |
Reassembling the buckets
Kyle Harrison averages 4.92 innings per start. Zac Thornton averages 5.59. Weight each starter's own earned run average across that length, then hand the remainder of the nine innings to his club's season staff rate. That is a conservative relief proxy — it mixes rotation and bullpen — and it is stated as a limitation, not hidden.
| Side | Starter IP | Starter ERA | Relief IP | Staff ERA | Projected ER |
|---|---|---|---|---|---|
| Brewers (Kyle Harrison) | 4.92 | 2.84 | 4.08 | 3.42 | 3.10 |
| Mets (Zac Thornton) | 5.59 | 3.04 | 3.41 | 4.07 | 3.43 |
| Combined | 6.53 |
A combined 6.53 projected earned runs against a 7.5 total. Unearned runs are ignored. Neither starter hits his average length every night. The opposing lineups are not in the formula; they sit in the table above.
The park
Citi Field is listed at 335 feet to left, 370 feet to left center, 408 feet to center, 380 feet to right center, 330 feet to right, on grass, open at 10 feet above sea level with a listed capacity of 42,136. Park dimensions do not decide a total by themselves. They decide whether a walk plus a ball in the air becomes two runs or a long out.
Where this reading is weakest
The season run rates do not want this number. Combined they sit at 9.13 against a 7.5 total, a gap of 1.63 runs that the starters and the park have to close. Kyle Harrison also allows 1.19 home runs per nine, and one three-run ball turns a low-event night into a push or a loss.
What would falsify it
A start from either pitcher that ends before the fifth with a pitch count above 95, or a walk total of four-plus from Zac Thornton, is the cleanest failure signal. That is how a 7.5 total gets away: free baserunners, an early bullpen, and a second time through a lineup that already scores 4.96 runs a night on the higher side.
The position
Brewers/Mets Under 7.5 at -115, 1.5 units. A price of -115 implies a 53.5% break-even. Grade it against this published number. If a listed starter is scratched, the play is void.
- Brewers/Mets Under 7.5 -115 · 1.5u
Who is on the mound?
Kyle Harrison for Brewers. Zac Thornton for Mets. Those names come from the MLB probable-pitcher feed, not from a guessed rotation.
Why use batters-faced rates instead of ERA?
Earned run average is a counting result. Strikeout and walk rates as a share of batters faced describe the process that produced it, and they stabilize on a smaller sample. Kyle Harrison at 23.9 points of K−BB separation versus Zac Thornton at 11.0 is the comparison ERA hides.
How is the pick graded?
Against the published price, on the public record. Pending is not a win. Today's full card lives at a stable URL: mlb-picks.html.
Pick, odds, and unit size come from the BetLegend daily tracker. Probable starters, season rates, and venue facts are from the MLB Stats API. Educational analysis. 18+ only. 1-800-GAMBLER.