chance of a podium trio Formula 1 has never seen at the United States Grand PrixThe overall probability that any brand-new podium trio appears at the next race, not just the top-ranked one.
Updated after qualifying · was 36% before the grid was set
Most likely new trio · 7.0%
Across every race since 2019 the model put the real trio in its top three 35% of the time.
Most likely trios Every trio the model rates as a live chance at the next race. Highlighted ones have never shared a podium; hover or tap the rest to see when they did. grid-aware
- 13.82%ANTP1/LECP4/VERP2Happened 2 times
Last: 2026 Singapore Grand Prix
See all → - 11.68%ANTP1/VERP2/RUSP6Happened 4 times
Last: 2026 Italian Grand Prix
See all → - 8.66%ANTP1/HAMP5/VERP2Happened 2 times
Last: 2026 Bahrain Grand Prix in Malaysia
See all → - 7.03%ANTP1/HADP3/VERP2NEW
- 4.93%ANTP1/VERP2/NORP8Happened 3 times
Last: 2026 Spanish Grand Prix
See all → - 4.24%ANTP1/LECP4/RUSP6Happened 1 time
Last: 2026 Australian Grand Prix
See all → - 3.27%LECP4/VERP2/RUSP6Happened 3 times
Last: 2023 Abu Dhabi Grand Prix
See all → - 3.04%ANTP1/HAMP5/LECP4NEW
- 2.93%ANTP1/VERP2/PIAP7NEW
- 2.56%ANTP1/HAMP5/RUSP6Happened 1 time
Last: 2026 Chinese Grand Prix
See all → - 2.50%HAMP5/LECP4/VERP2Happened 5 times
Last: 2023 Mexico City Grand Prix
See all → - 2.41%ANTP1/HADP3/LECP4NEW
- 2.17%HAMP5/VERP2/RUSP6Happened 4 times
Last: 2023 Spanish Grand Prix
See all → - 1.97%ANTP1/HADP3/RUSP6NEW
- 1.82%HADP3/LECP4/VERP2NEW
- 1.75%ANTP1/LECP4/NORP8NEW
- 1.57%HADP3/VERP2/RUSP6Happened 1 time
Last: 2026 Azerbaijan Grand Prix
See all → - 1.49%ANTP1/NORP8/RUSP6Happened 1 time
Last: 2026 Dutch Grand Prix
See all → - 1.44%ANTP1/HADP3/HAMP5Happened 1 time
Last: 2026 Monaco Grand Prix
See all → - 1.37%LECP4/VERP2/NORP8Happened 5 times
Last: 2025 Mexico City Grand Prix
See all → - 1.21%HADP3/HAMP5/VERP2NEW
- 1.15%VERP2/NORP8/RUSP6Happened 3 times
Last: 2025 Singapore Grand Prix
See all → - 1.07%ANTP1/HAMP5/NORP8NEW
Show current form ▾Hide current form ▴
Each driver's modelled strength — their own rating plus their car's, learned race by race from every finishing and qualifying order since 1950.
New podiums through the years Every trio that debuted on a podium that season. Drag the slider or click a bar to explore.
Frequently asked questions
How does the prediction model work?
A Bayesian rating engine keeps a skill rating for every driver and every car, updated after each qualifying session and race since 1950 (uncertainty widens between seasons and when the rules change). For the next race it blends those ratings with each car’s reliability record (DNF risk) and the circuit’s character, then simulates the podium hundreds of times with a Plackett–Luce model to rank the never-before-seen trios.
What does the headline percentage mean?
It’s the overall probability that any brand-new podium trio appears at the next race — not just the top-ranked one, but any combination that has never happened before.
Why does the prediction update after qualifying?
Qualifying is the most informative pre-race session: it reveals current pace and fixes the starting grid, whose track-position advantage has strong historical precedent. Once the grid is set, the model feeds the qualifying order through its rating engine and adds a grid-position term scaled by how processional the circuit historically is — stronger where overtaking is hard, weaker where the grid gets shuffled. In backtests, the grid-aware update places the actual podium trio in its top 3 40% of the time.
How accurate is the model?
Backtested on seasons it never saw during tuning, the model places the actual podium trio in its top 3 predictions 35% of the time. F1 podiums are inherently high-variance, so exact-trio hits are rare by nature.
What is “current form” based on?
Each driver’s strength is their own rating plus their car’s, learned from every finishing and qualifying order — so a driver can rate higher than their recent podium count suggests when the car underneath them is quick.
What does “podigami” mean?
Podigami blends “podium” and “scorigami” — it’s the practice of tracking F1 podium trios that have never happened before. Since 1950, 755 unique trios have appeared across 1,166 races. Today’s 22-driver grid can produce 1,540 different trios per race, so most combinations simply haven’t come up yet.
Why do some 1950s podiums list four drivers?
Until 1960 a driver could hand his car to a team-mate mid-race, and both were classified at the finishing position. The 1956 Belgian Grand Prix is the clearest case: Cesare Perdisa and Stirling Moss shared the third-placed Maserati, so that race put two different trios on the podium at once. Eighteen races in history work this way, and every trio they produced counts as having happened.
What else is on this site?
Four deeper dives: Combinations lists every unique podium trio in history; Overdue ranks the trios that keep almost happening; Unlikeliest celebrates the podiums that defied the odds; and Soulmates maps which drivers keep meeting on the podium.