Track Record ✦ the last 30 days

56.8% Accuracy

130–99 across 229 games

— By Simulation Type —

Simulation Correct Total Accuracy
Baseline 126 229 55.0%
Pitcher-weighted 128 229 55.9%
Platoon-weighted 115 229 50.2%
Pitcher + Platoon ★ leader 132 229 57.6%
Logistic 123 229 53.7%
Random Forest 121 229 52.8%
Consensus 130 229 56.8%
Pitcher + Platoon is leading individual model by 1.7 points. Consensus edges out 5 of 6 sims.

— Per-Team Accuracy —

★ Top 5 teams
Giants
12–2 85.7%
Athletics
8–2 80.0%
Dodgers
14–5 73.7%
Cardinals
11–4 73.3%
Astros
9–4 69.2%
✗ Bottom 5 teams
Mets
5–10 33.3%
D-backs
4–7 36.4%
Orioles
5–8 38.5%
Mariners
4–6 40.0%
Blue Jays
6–9 40.0%

52-point spread between best and worst — Mets the hardest to call.

— Confidence Calibration —

Are our confidence numbers honest? Compare what we said to what actually happened.

Confidence band Picks Avg predicted Actual rate Gap
50-55% 145 52.2% 52.4% +0.2
55-60% 63 56.9% 57.1% +0.3
60-65% 14 62.3% 78.6% +16.3
65-70% 6 66.8% 100.0% +33.2
70%+ 1 74.6% 100.0% +25.4
Calibration drifts up to 33.2 points across bands.

— Lineup Impact —

Lineup status Games Correct Accuracy
Real lineups 1724 933 54.1%
Fallback lineups 618 351 56.8%
Fallback-lineup predictions are 2.7 points more accurate. 26% of season picks rely on fallback lineups.

— Pick Distribution —

Avg confidence

54.6%

across 230 picks

Home team picks

59.1%

vs MLB home rate ~54%

Unanimous picks

13%

all sims agreed

— Recent Performance —

Last 30 days

★ Best day

2 of 2

Tuesday, October 6 · 100.0%

✗ Worst day

1 of 4

Saturday, October 3 · 25.0%

Sep 11 Sep 26 Oct 10
Date Correct Total Accuracy
Tue, Oct 6 2 2 100.0%
Mon, Oct 5 1 1 100.0%
Sun, Oct 4 1 2 50.0%
Sat, Oct 3 1 4 25.0%
Thu, Oct 1 1 1 100.0%
Wed, Sep 30 2 4 50.0%
Tue, Sep 29 2 4 50.0%
Sun, Sep 27 8 14 57.1%
Sat, Sep 26 5 11 45.5%
Fri, Sep 25 6 14 42.9%
Show all 24 days →