Home Underdogs Trends

Methodology

We filtered to regular-season games where the home team was the underdog (home_underdog = TRUE, schedule_playoff = FALSE) and measured cover rate (home_spread_outcome) and outright win rate (home_win). We then re-ran the same query restricted to 2001–2025 to test whether the pattern survives in the modern market.

import pandas as pd
df = pd.read_csv("spreadspoke_enhanced.csv", low_memory=False)
hu = df[(df.home_underdog==True) & (df.schedule_playoff==False)]
c, n = (hu.home_spread_outcome=='Cover').sum(), (hu.home_spread_outcome=='No Cover').sum()
print(f"All-time home dog ATS: {100*c/(c+n):.1f}% (n={c+n})")
m = hu[hu.schedule_season>=2001]
c2, n2 = (m.home_spread_outcome=='Cover').sum(), (m.home_spread_outcome=='No Cover').sum()
print(f"2001-2025 home dog ATS: {100*c2/(c2+n2):.1f}%")

The Analysis

The disrespected home team is one of the oldest trends in the book. When a home team is getting points, the public tends to pile onto the shiny road favorite, and historically the line tilted a hair too far: regular-season home dogs covered 51.8% of the time across the full dataset. For decades that was a quiet, durable lean you could lean on.

Here's the honest part — and the reason we always run a modern-era check. Restrict the same query to the last 25 seasons and the trend basically evaporates: home dogs cover right at 50%. As betting markets got sharper and information got cheaper, oddsmakers (and bettors) arbitraged the old discount away. It's a textbook example of a pattern that was real, got famous, and then got priced in.

And throughout, the framing never changed: home underdogs win outright only about 35% of the time, so the value was always in the points, not the upset. The lesson for readers is as much about method as money: a trend that looks great over 60 years can be dead in the last 20, which is exactly why you check both.