How often the big favourite wins, and how often they cover, across the NRL, AFL, NFL and NBA.
Loading…
How often does a heavy favourite win?
Bands are percentiles of that sport’s own spreads, so “top 5%” means a very different
number of points in the AFL than in the NFL. The actual range is shown.
Band
Games
Spread range
Favourite wins
Favourite covers
Over
Avg total
Which teams cover as big favourites?
Team
Games
Avg spread
Wins
Covers
vs 52.4%
Lines and results from aussportsbetting (NRL, AFL) and SportsDatabase (NFL, NBA). A push —
margin exactly equal to the spread — is a returned bet and is excluded from cover rates rather than counted
as a loss. Informational only, not betting advice · 18+ · Gamble responsibly.
Backing a heavy favourite: what the record actually shows
The intuition is obvious and correct: a team favoured by a huge margin usually wins. Across the biggest 5% of
spreads, favourites win between 88% and 96% of the time depending on the sport, against roughly 67%
across all games with a line. What follows from that is the part that catches people out.
Why the win rate rises and the cover rate does not
A spread exists to make both sides of a bet equally attractive. As a mismatch grows, the line grows with it, so
the favourite has to win by more to return anything. The record bears this out almost exactly: cover rates sit
between roughly 46% and 57% in every band of every sport here, with no trend as the spread gets bigger. Win rate
climbs steeply; cover rate is flat. The market has already done the work.
What counts as heavy depends on the sport
A 10-point NFL line is a big favourite. A 10-point AFL line is close to a coin flip. That is why the bands on
this page are percentiles of each sport’s own spreads rather than fixed point counts: the top 5% means
17.5+ in the NRL, 49.5+ in the AFL, 12.5+ in the NFL and 14.5+ in the NBA.
Reading the per-team table honestly
Splitting one question across 25 or 32 teams turns a single test into dozens, on a couple of dozen games each.
At that sample the standard error is close to ten percentage points, so the top of the table will always look
remarkable. The figure printed above the table is the one that matters: how many teams would clear break-even on
luck alone, measured against the pooled rate rather than against 50%. When the observed count is at or below the
expected one, the spread between best and worst is noise.
Is there a bet here?
Of sixteen bands tested across four sports, two cleared the 52.4% break-even and chance alone predicts
about three. The strongest single cell was the NBA’s biggest 5% of spreads; even that sits barely over one
standard deviation above break-even, and across sixteen cells there is a meaningful chance one looks that good by
itself. Nothing here is a system. It is the record, with the arithmetic that stops it being mistaken for one.
How is a push handled?
A push is a margin exactly equal to the spread. The bet is
returned, so it is removed from the cover-rate denominator rather than counted as a loss, which would understate
every figure on the page.
Why is the per-team table built on the top 20% and not the top 5%?
Because the top 5%
of a sport’s games leaves most clubs two or three appearances, which is not a record. The top 20% is the first
band where most teams reach a double-digit sample, and teams below the minimum are excluded rather than shown with
a misleading percentage.