How does a simulation compare F1 teammates “fairly” if real races aren’t equal?

It holds selected variables constant (like reliability noise or strategy bias) and re-runs outcomes many times to estimate an expected gap and its uncertainty. You’re not replacing reality—you’re isolating execution signals that points alone can’t separate.

Is a teammate simulator just predicting who is faster?

No—use it to quantify ranges and flip risk, not a single verdict. The key outputs are the expected advantage and how often that advantage disappears under plausible conditions.

What inputs matter most when modelling a teammate gap?

Qualifying conversion, race-pace consistency, tyre life, and incident/reliability assumptions usually drive the biggest swings. Track mix matters too, because some driver strengths convert to points more strongly on certain circuit types.

How should I interpret a small average gap with high variance?

It means the drivers are close on underlying performance, and weekend-to-weekend randomness can dominate the points outcome. In that case, focus on which assumptions create the tails (DNFs, penalties, track position) rather than arguing from one headline result.

Do points calculators account for the lack of fastest-lap points from 2025 onwards?

They should—without a fastest-lap bonus, the points ceiling and late-race incentives change slightly. When you run scenarios, make sure your assumptions match the current points rules so you don’t build conclusions on outdated scoring.

How does a simulation compare F1 teammates “fairly” if real races aren’t equal?

It holds selected variables constant (like reliability noise or strategy bias) and re-runs outcomes many times to estimate an expected gap and its uncertainty. You’re not replacing reality—you’re isolating execution signals that points alone can’t separate.

Is a teammate simulator just predicting who is faster?

No—use it to quantify ranges and flip risk, not a single verdict. The key outputs are the expected advantage and how often that advantage disappears under plausible conditions.

What inputs matter most when modelling a teammate gap?

Qualifying conversion, race-pace consistency, tyre life, and incident/reliability assumptions usually drive the biggest swings. Track mix matters too, because some driver strengths convert to points more strongly on certain circuit types.

How should I interpret a small average gap with high variance?

It means the drivers are close on underlying performance, and weekend-to-weekend randomness can dominate the points outcome. In that case, focus on which assumptions create the tails (DNFs, penalties, track position) rather than arguing from one headline result.

Do points calculators account for the lack of fastest-lap points from 2025 onwards?

They should—without a fastest-lap bonus, the points ceiling and late-race incentives change slightly. When you run scenarios, make sure your assumptions match the current points rules so you don’t build conclusions on outdated scoring.

Using Simulations to Compare Teammates Fairly

TL;DR

Same car doesn’t mean same year — points bundle pace, luck, strategy, and reliability into one number that can’t separate them.
An equalised simulation reruns the season under controlled assumptions (reliability, strategy, track mix) so the remaining gap is about execution, not circumstance.
Run a baseline, then an equalised version, then stress-test the calendar in the Drivers tool — read the output as a distribution, not a verdict.
Watch two numbers: the expected gap and how often it flips. A gap that holds across many runs is a signal; a gap that only appears in a few is an artifact.

Same car, same garage, same engineers. Teammates should be the easiest F1 comparison on the calendar — and somehow they’re the one everyone argues about.

Part of the problem is that the scoreboard lies. Not deliberately — points just don’t separate what you want separated. McLaren’s 2025 intra-team battle ends with Piastri ahead of Norris by some delta, but that delta is tangled up with a hydraulic failure, a red-flag timing quirk, and one of them drawing the short straw on strategy when both cars couldn’t run the same tyre. You can stare at the standings all season and still not know who drove the car better.

Simulations don’t fix this by predicting anything. They fix it by letting you rerun the season often enough that the noise starts cancelling out. That’s the whole idea.

Points aren’t a measure of ability

Points are a record of outcomes. They compress qualifying, race pace, tyre management, a well-timed Safety Car, a badly timed one, a penalty, a failure — all into one number per weekend. By the time you’re looking at a season total, you have no idea which of those threads mattered.

The points curve is also steep. A couple of seconds over a race distance can be the difference between P2 and P4, which is a six-point swing. P10 versus P12 might come down to a Virtual Safety Car landing five seconds after you pitted — not a performance gap at all. With the fastest-lap bonus gone since 2025, one distortion is off the table, but everything else is still there. Two teammates can be genuinely close on pace and finish 40 points apart by October.

If you actually want to know who was better, the season points gap is the wrong place to look.

What “equalised conditions” means

The trick isn’t pretending both drivers had the same year. It’s stripping out the stuff that wasn’t really about them and seeing what’s left.

In practice that looks like:

Same baseline car. Treat the pace potential as equal, so you’re measuring who converts it.
Reliability noise smoothed out. One DNF shouldn’t define a season-long comparison.
Strategy on a level field. Remove the “priority car” advantage and see what happens when both drivers run the same risk profile.
Traffic and clean-air separated. Some drivers are clinical in clean air and average in traffic. Worth knowing — and worth not hiding inside a points total.

The question you’re asking is a counterfactual: rerun this season a thousand times with the same driver traits and different luck — what’s the average gap? The answer is a distribution. That’s the point.

Open the Drivers tool to set up the baseline. The workflow that follows works for any pairing you want to look at.

The three-layer comparison

Three steps: baseline, equalised run, stress test. This works whether you’re looking at Verstappen and Tsunoda, Russell and Antonelli, Alonso and Stroll — any pairing.

1. Baseline — what actually happened

Start with the obvious stuff. Qualifying head-to-head. Race finishes. Points. Note the context: who took grid penalties, who retired, who got the sprint weekends that rewarded their strengths.

This isn’t the answer. It’s the shape you’re comparing against. If one driver lost 40 points to two mechanical failures, the baseline is already telling you what the next step needs to do.

Use the Drivers tool to lock in the window — season-to-date, last six races, whatever’s relevant — so you’re not moving the goalposts between runs.

2. Equalised run — where the real gap lives

Rerun the comparison under controlled assumptions. This is where conversion shows up, which is usually where the actual separation is.

Three kinds of conversion matter:

Qualifying conversion. When the tyre’s in the window and the fuel is low, does the lap happen? This is execution under pressure, and it matters a lot on tracks where overtaking is hard — Monaco, Hungary, Singapore.

Race conversion. Given similar pace, who runs cleaner stints? Fewer mistakes, better tyre management, stable lap times when the track temperature swings?

Opportunity conversion. When a race opens up — Safety Car, alternate strategy, late restart — who takes the points without giving back more than they gain?

If the headline gap is 40 points but the equalised gap is 8, you’ve learned something real: the season was noisier than the standings suggest. If it’s still 35 after equalising, the gap is closer to genuine.

Run it in the Drivers tool and look at two numbers: the expected gap and the spread. A gap that holds across many runs is a signal. A gap that only appears in a handful of runs is an artifact.

3. Stress test — change the calendar

Here’s the mistake that shows up most often in teammate analysis: assuming performance is flat across the season. It isn’t. Driver traits interact with track types in ways that can make someone look dominant on paper because the calendar happened to suit them.

A qualifying specialist looks unbeatable on a schedule heavy with track-position circuits. A tyre-whisperer looks mediocre in qualifying and then quietly outperforms in high-deg races where stint length creates strategic options. Same drivers, different calendar, different story.

Rerun with different mixes. More street circuits. More high-deg races. More rain. More Safety Car probability. You’re not forecasting the next season; you’re asking whether the advantage you saw was about the driver or about the schedule.

This is also where teammate analysis stops being isolated from the championship. If you’re asking who to back for a title push, you’re really asking how their traits will convert to points across the remaining calendar. Use the Drivers tool to work out what’s driving the gap, then push it through the Season Simulator to see what it does to the standings.

Reading the output without kidding yourself

A fair comparison doesn’t give you a winner. It gives you a range and the reasons for it. Four things are worth looking at.

Median over mode. In a noisy sport, the “most likely” single outcome can sit next to a long tail of rare but decisive swings. Use the expected gap, not the most common one.

Flip frequency. How often does Driver B beat Driver A across runs? A driver who loses on average but wins a meaningful slice of simulated seasons can be more valuable than the standings suggest — especially if their upside is timed right for a title fight.

Pace delta vs execution delta. Raw speed is one thing. Turning it into points when you add traffic, tyre wear, and decisions under pressure is another. In teammate comparisons, execution is usually what separates them.

Your own assumptions. If your run treats reliability as equal and strategy as neutral, that’s a choice, not a fact. The reason to use a tool is that you can change those choices one at a time and see whether your conclusion survives.

Two things people get wrong

“Same car means perfectly equal.” It never does. Teammates share a concept, not an experience. Upgrades land at different weekends. Setup direction pulls one way or the other. Parts allocation isn’t always symmetric — and that played out at both Red Bull and Ferrari more than once across 2024 and 2025. Equalised simulations don’t pretend this doesn’t happen; they let you remove it so you can see what’s left underneath.

“Simulation equals prediction.” A sim run is a structured what-if. The most useful output is often the one that says two drivers are separated by a small average gap with a wide spread — which means the debate isn’t about who’s better, it’s about whose profile is more robust under whatever uncertainty is still ahead.

The takeaway

Teammate comparisons get noisy because points are an output, not a diagnostic. If you want to know who actually drove the car better — and why — run the same season under equalised conditions, stress it with realistic luck, and read the result as a distribution instead of a verdict.

Start in the Drivers tool. Run an equalised scenario, then change one assumption at a time until you can point to what’s actually moving the answer. That’s how you stop arguing about teammates and start getting somewhere.