How F1 changed · Chapter 1 of 3

When was F1 actually a contest?

If a regulation reset really opens the sport up, the wins should spread out. This page tracks how concentrated race victories were, season by season, and asks which rule changes actually changed who got to the front.

Data: constructor race wins from Jolpica-F1, 1958–present.

How concentrated were the wins?

The first view uses the Herfindahl-Hirschman Index: 1.0 if one constructor wins everything, smaller numbers if the wins are shared more evenly. You do not need to love the acronym; just read lower as more open.

Higher values mean one constructor is taking most of the wins; lower means the grid is sharing them. The Mercedes hybrid-era stretch (2014–2020) is easy to spot, as is the early-2000s Ferrari run.

How many teams actually won a race?

The other angle on the same question: rather than measuring concentration, just count distinct race-winning constructors per season. A harder lower bound than HHI — a season with six winners is always competitive, even if two of them took twelve wins apiece.

Peaks in 1982 (eleven winning constructors) and 2012 (six) are the most open years in the data; the troughs through the mid-60s and mid-2010s tell the opposite story.

What the series says

The hybrid era (2014–2021) was not an era of single-team dominance in the Ferrari-early-2000s sense — it was dominance by two Mercedes drivers against each other. HHI stays elevated, but unique_winners dipped to three or four for years. That's a different kind of closed championship than 2002-2004.
Regulation resets sometimes do nothing. The 1983 flat-bottom ban and 2017 wider-car formula barely register in either series. Changing the rules and changing the pecking order are distinct outcomes.
Sparse seasons are noisy. Early-1950s fields were tiny and entries were inconsistent; we cut the chart at 1958 (start of the constructors' championship) for this reason. HHI values are stable from the mid-60s onward.