Atlanta Braves @ Houston Astros
ABAwayAtlanta Braves69%
@HAHomeHouston Astros31%
Model PickAtlanta Braves 69%Market edge — disclosed after closing odds are collectedEvidence 12 structured
AI Summary grounded in real data · zero hallucination
Our model projects an Atlanta Braves victory with a 69% win probability. This forecast is primarily driven by the Braves' advantages in 'Team ERA' (-7.5 deltaPct) and 'Last 10 Games Form' (-3.7 deltaPct). A key swing risk is the Houston Astros' 'Home Advantage' (2.5 deltaPct), which could partially mitigate the Braves' edge. The market edge is currently uncalculated.
Logistic-regression statistical model — predicts Atlanta Braves vs Houston Astros by accumulating the log-odds contributions of 12 proven structured metrics. Only structured metrics enter the probability; qualitative and folklore factors get a label only. Predictions are sealed with SHA-256 before the game.
Model factor contribution How the 31% home win was derived
Home Advantage+2.5
Pythagorean Expectation-4.7
Season Win%-1.6
Last 10 Games Form-3.7
Home/Away Split-0.4
Team OBP+0.3
Team SLG-0.1
Team ERA-7.5
Team WHIP-2.2
Starting Pitcher Matchup-1.2
Baseline 50% → final home win31.3%
Each factor's log-odds contribution accumulated · only structured metrics reflected (qualitative/folklore excluded) · each %p converts log-odds to probability, so it's non-linear — a simple sum may not exactly equal (final − baseline)
Structured metric analysis 12 metrics · fully disclosed
| Metric | Analysis (home vs away) | Source |
|---|---|---|
| Home Advantage | Houston Astros at home — MLB historical home win rate ~.535 (ballpark familiarity, no travel) | MLB StatsAPI |
| Pythagorean Expectation | Home 46.4% (686 RS/743 RA) vs Away 58.1% (718/600) — strength from runs scored/allowed | MLB StatsAPI |
| Season Win% | Home 77-78 (.497) vs Away 91-64 (.587) | MLB StatsAPI |
| Last 10 Games Form | Home 3-7 vs Away 6-4 · Home L2 | MLB StatsAPI |
| Home/Away Split | Home team home 39-41 · away team away 40-37 | MLB StatsAPI |
| Team OBP | Home .313 vs Away .311 — foundation of scoring | MLB StatsAPI |
| Team SLG | Home .410 vs Away .412 (OPS .723/.723) | MLB StatsAPI |
| Team ERA | Home 4.53 vs Away 3.62 — lower is better | MLB StatsAPI |
| Team WHIP | Home 1.35 vs Away 1.25 — baserunners allowed per inning (lower is better) | MLB StatsAPI |
| Starting Pitcher Matchup | Home Hunter Brown (ERA 3.43) vs Away Martín Pérez (ERA 3.07) | MLB StatsAPI |
| Hitting Detail (BABIP, K, BB) | Home BABIP .277 · K 1271 · BB 500 / Away BABIP .288 · K 1252 · BB 446 (BABIP is luck-driven → not factored into probability) | MLB StatsAPI |
| Pitching Detail (K/BB, HR9) | Home K/BB 2.16 · HR9 1.31 / Away K/BB 2.52 · HR9 1.12 | MLB StatsAPI |
Qualitative & folklore factors None
This game has no extracted qualitative factors — we don't invent factors without a basis.
12 source headlines (auto-collected · basis disclosed)
- Atlanta Braves at Houston Astros Game Story, Scores/Highlights - 09/18/2026
- Atlanta Braves at Houston Astros - Where to Watch, Stream Info, TV Channel, and Live Updates (September 19, 2026)
- Braves 6-2 Astros (Sep 18, 2026) Game Recap
- How to watch Braves vs. Astros: TV channel and streaming options for September 18
- Astros can't get the timely hits and fall to Braves and into AL West tie
- Atlanta Braves at Houston Astros - Where to Watch, Stream Info, TV Channel, and Live Updates (September 18, 2026)
- How to watch Braves vs. Astros: TV channel and streaming options for September 19
- Atlanta Braves at Houston Astros Game Story, Scores/Highlights - 09/19/2026
Frequently asked questions This game's prediction summary
Atlanta Braves vs Houston Astros — who's more likely to win?
SealedSlate's logistic-regression statistical model gives Atlanta Braves a 69% chance to win (Home 31% · Away 69%). MLB · 2026-09-21 03:10 KST.
How was the 31.3% home-win probability calculated?
Starting from a 50% baseline, we additively accumulate the log-odds contributions of 12 proven structured statistical metrics to reach the final 31.3%. Each factor's %p contribution is fully disclosed on the page; qualitative and folklore factors are not reflected in the probability.
Can this prediction be altered later?
No. Every prediction is sealed with a SHA-256 hash before the game starts and permanently recorded in the track record, making after-the-fact changes impossible. SealedSlate runs a transparent ledger that never inflates the hit rate and leaves missed predictions exactly as they were.
We don't compete on hit rate — every factor is fully disclosed and labeled. Only 📊 structured stats feed the probability (each factor's contribution shown in %p); ⚔️🔮 are context to weigh. Predictions are locked and sealed before the game — verify the hits directly on the track record.