How much is pole worth? 709 races say half, and that the circuit hardly matters
F1 Legend Sim · Staff · 1 September 2026 · 8 min read
There is a line you hear in every broadcast and that nobody has ever checked: that Monaco is won on Saturday and that at Monza the grid is only a suggestion. It sounds obvious. We counted the 709 Grands Prix run between 1988 and 2026, one by one, looking at where the winner started. The line does not hold.
The overall number is clean, and higher than most people assume: since 1988, pole converts into a win 50% of the time. One in two. Starting first guarantees nothing, but it is comfortably the best single place to be.
The interesting part starts when you split those races by circuit, because that is where everyone has an instinct and where the instinct turns out to be wrong.
Monza converts more poles than the Hungaroring
Monza is the temple of the slipstream, the place where track position supposedly counts for little because somebody is always tucked in behind. The Hungaroring is its opposite: narrow, twisty, a circuit where overtaking is treated as nearly impossible. If the instinct were right, the gap between the two would be enormous.
Across 38 Italian Grands Prix at Monza, pole won 47% of the time. Across 39 in Hungary, 44%. Monza converts more.
Nor is it a one-off. Silverstone, all long straights and heavy braking zones, sits at 35%. Magny-Cours, one of the tighter tracks, is near the bottom of the list at 28%. Monaco, which ought to be the absolute ceiling, manages 55%: high, certainly, but a long way from the "the race is already over" that gets repeated every May.
| Circuit | Races | Wins from pole | Polesitter avg. finish |
|---|---|---|---|
| Silverstone Circuit | 40 | 35% | P6.2 |
| Hungaroring | 39 | 44% | P3.9 |
| Circuit de Monaco | 38 | 55% | P3.8 |
| Autodromo Nazionale di Monza | 38 | 47% | P5.4 |
| Circuit de Spa-Francorchamps | 37 | 43% | P4 |
| Circuit de Barcelona-Catalunya | 36 | 69% | P3.4 |
| Circuit Gilles Villeneuve | 36 | 50% | P5.4 |
| Suzuka Circuit | 35 | 54% | P4 |
| Autódromo José Carlos Pace | 35 | 49% | P4.4 |
| Albert Park Grand Prix Circuit | 29 | 45% | P5.7 |
| Hockenheimring | 26 | 62% | P4.8 |
| Autodromo Enzo e Dino Ferrari | 24 | 46% | P6 |
The twelve circuits with the most Grands Prix run between 1988 and 2026. Tracks with eight or nine races are deliberately left out of the table: at that sample size, one race either way moves the figure by ten points.
A circuit's reputation predicts nothing
This can be measured rather than argued. Our simulator stores, for every track, a number describing how easy overtaking is there, from 0.15 at Monaco to 0.75 at Monza. If the instinct were right, that number should predict how many poles convert: the harder it is to pass, the more wins from pole.
Cross that parameter with the real races at 25 circuits and the correlation comes out at -0.145. With twenty-five circuits, that is noise. It points the expected way, but it is far too weak to predict anything.
Put another way: tell me how easy overtaking is at a circuit and I still cannot guess how often pole turns into victory there.
Why, most likely
The explanation that best fits the numbers is that we were measuring the wrong thing. Whether a pole converts depends far less on whether passing is possible than on who took the pole. In a season of one-car dominance, whoever starts first usually has the quickest car too, and then wins anywhere. In a tight season, pole can fall to a car that cannot hold the pace on Sunday, and it loses anywhere.
That would explain why the spread is so flat: what decides this is the season, not the track. We measured that other half separately in what is the most dominant car in F1 history.
And this is where our own simulator gets it wrong
This part is uncomfortable to publish, which is why it belongs in the article. We ran the same experiment inside our own engine: same grid, same starting order forced at every circuit, 120 races at each one. The only thing that changed was the track.
| Circuit | Reality | Our simulator |
|---|---|---|
| Circuit de Barcelona-Catalunya | 69% | 25% |
| Marina Bay Street Circuit | 69% | 25% |
| Circuit de Monaco | 55% | 25% |
| Suzuka Circuit | 54% | 18% |
| Autódromo José Carlos Pace | 49% | 17% |
| Autodromo Nazionale di Monza | 47% | 17% |
| Autodromo Enzo e Dino Ferrari | 46% | 26% |
| Hungaroring | 44% | 26% |
| Circuit de Spa-Francorchamps | 43% | 16% |
| Silverstone Circuit | 35% | 17% |
| Baku City Circuit | 33% | 19% |
The engine converts pole between 16% and 26% depending on the circuit. Reality is 50%. We fall short by roughly half.
There is a second, deeper error. In the engine, the advantage of starting ahead depends on the circuit: the harder it is to overtake, the more pole is worth. That is precisely the instinct the real data has just contradicted. So it is not only that our number is too small: the formula rests on an idea that does not hold up.
We checked it with a test real Formula 1 cannot run: reverse the grid and put the quick cars at the back. If the circuit mattered, the top six should claw back many more places at Monza than at Monaco. They recover 15.2 and 15. Effectively the same.
What we are going to change
Two things, in this order. First, raise the weight given to starting position until conversion looks like the real 50% rather than today's 20%. Second, and more interesting: try making that weight stop depending on the circuit at all, because the data says the track does not explain it.
We have not touched it yet, and not out of laziness. That number feeds all eleven modes of the simulator and the calibration of every driver, so moving it means re-running the whole bench. We would rather publish the measured error than patch it by eye.
In the meantime, if you want to play with this, in What If? you can force the circuit and the starting grid and see how long a slow car holds up a quick one.