F1 does not publish a clean overtaking count all the way back to
1950, but it does publish where cars started and where they finished.
That is enough to build a long-run proxy for how much the race order
actually changed on Sunday.
Data: classified finishers, 1950–present. DRS arrived in 2011; ground-effect rules returned in 2022.
The first proxy is the absolute mean of |grid - finish| across
classified finishers. Higher numbers mean more movement through the race;
lower numbers mean Sunday ended close to Saturday's order.
The late-70s aero revolution, the grooved-tyre era, and the 2022
ground-effect return each show step changes. DRS (2011) also bumps the
metric as expected — though notably, the hybrid era with its three-year
Mercedes dominance suppresses it again even *with* DRS.
Another angle: Pearson correlation between grid position and finish
position. Closer to 1.0 means the race order is determined by Saturday;
closer to 0 means the race materially reshuffles the order.
A complementary view: high correlation = races that start
pre-determined. The rising trend into the late-2010s is a mirror of the
churn chart above, and the 2022 return to ground effect visibly breaks
the pattern.
Taken together, the three charts retell the last seventy-five years
from the inside. Regulation rewrote the rulebook every decade or
so but rarely changed who won. Reliability did what the rulebook
couldn't — quietly turning a two-in-three DNF sport into a
one-in-twenty one. And racecraft paid the bill: when cars stopped
breaking and aero made dirty air a barrier, the mean grid-to-finish
delta quietly fell by half. Every modern rule package since 2011
(DRS, 2017 wider cars, 2022 ground effect) has been an attempt to
pay some of that bill back — with mixed results you can read
off the same chart.