The Wet Index: F1's Actual Rain Masters, Ranked by Teammate Head-to-Head

“Rain master” is the most overused phrase in Formula 1 commentary. It is handed out on the basis of one or two memorable laps, often years old, often dependent on car and tyre and circuit rather than any specific skill the driver brought to the rain. Schumacher in 1996 at Barcelona. Senna in 1993 at Donington. Hamilton in 2008 at Silverstone. Verstappen in 2016 at Interlagos. These are the canonical reference points, and they get retrieved from memory every time the rain comes down, regardless of what the data says is happening on the actual track in front of us.

The problem is the data has been available for a long time, and almost nobody is using it.

The premise

What’s a fair test of “rain master”? Win count in the wet doesn’t work; most modern wet races are red-flagged into half-distance or won by the driver who got the lucky tyre call. Lap-time pace doesn’t work in isolation; a slow car in the rain is still a slow car. Career wet wins doesn’t work either, because drivers with long careers accumulate more wet wins simply by being on the grid longer.

The cleanest measure is lap pace versus your teammate, in the same car, on the same wet lap. That controls for everything that isn’t the driver. The Mercedes is a Mercedes whether it’s wet or dry. The Williams is a Williams. If a driver lays down a faster median wet lap than their teammate, that is a clean win on the variable we care about. Sum those wins and losses across a career and you have a stable rating that says nothing about machinery and everything about how each driver handles a wet floor underneath them.

This is exactly what we’ve built.

The method

The FastF1 archive has been recording lap-by-lap data and per-session weather since 2018. We have eight seasons (176 race sessions) of every lap each driver completed and every weather reading at the trackside Pitot tubes. We do four things:

1. Filter to truly wet laps. A “wet race” in broadcast terms can mean anything from a five-minute squall to a fully soaked Sunday. We don’t care about the broadcast label; we care about which specific laps had rain falling on them. So we tag each lap as wet or dry based on the nearest weather reading (within ± 2 minutes of the lap start). A race with two wet stints and three dry stints contributes separately to each driver’s wet and dry h2h tallies.

2. Filter to clean green-flag laps. We exclude in-laps, out-laps, safety-car laps, and laps FastF1 has flagged as inaccurate. We want pace, not pace-plus-traffic.

3. Match teammates. Within each session, for each team that fielded both its drivers, we take the median wet-lap time for each driver and compare. The faster median wins the wet h2h. The slower loses. A team where one teammate didn’t complete enough wet laps gets dropped; we require at least three wet laps per driver to call the comparison meaningful.

4. Aggregate across the career. Each driver accumulates wet wins, wet losses, dry wins, dry losses. The Wet-Pace Index is the wet-h2h win rate. The dry comparator is the dry-h2h win rate. The interesting number is the delta between them: a driver who runs +15 in the wet compared to the dry is a genuine wet specialist. A driver who runs -15 is the opposite, worse in the rain than the comparison should suggest.

Twenty drivers cleared our minimum sample threshold (at least six wet h2h samples in the eight-year window). That’s most of the current grid plus a few recently retired teammates whose history is still measurable. The full ranking is below.

The ranking

A few things jump out before any commentary.

The dashed reference line at 50% is the baseline: half your wet h2hs won. A driver above the line is, on average, faster than the teammates they’ve been paired with on wet laps. A driver below is slower. The dry bar (grey) is the comparator.

Three drivers rate above 70% wet h2h: Verstappen, Russell, Leclerc. Verstappen is the top-ranked wet performer at 81%, but he is also the top-ranked dry performer at 75%. His wet advantage is real but modest (+6.2 over dry). He is fast in everything. The bigger story sits a few rows down.

The genuine wet specialists

A “wet specialist” should mean someone who is more effective in the rain than out of it. Five drivers in the table fit that definition on a wet-minus-dry delta of +10 or higher:

Esteban Ocon (+32.1). This is the most extreme delta in the field. Ocon’s career dry h2h is a brutal 25%. He loses three out of four dry comparisons against his teammates, which is consistent with his reputation as a difficult driver to match against. In the wet, that flips to 57% wins. The shift is real. Anyone who watched the 2024 Brazilian Grand Prix from above sixth place got the live demonstration. The data confirms what the broadcast suggested: when the track gets wet, Ocon catches up.

Oscar Piastri (+43.8). This is the largest delta in the table, but it’s also the smallest sample (six wet h2hs). Piastri has only had wet matchups against Norris, and across that small window he runs 50% wet h2h against a 6% dry h2h. The dry number is genuinely bad; Norris has been substantially faster in the dry across their pairing. The wet number is genuinely not bad. The question is whether the +43.8 is real signal or six samples of noise. We won’t know for another two seasons of wet races. The current best estimate is that Piastri is a much closer match for Norris in the rain than in the dry. That’s enough to flag.

Zhou Guanyu (+17.8), Charles Leclerc (+15.9), and George Russell (+15.0) round out the list of consistent wet outperformers. Russell’s 75% wet h2h sits behind only Verstappen and Leclerc; his dry h2h is 60%, which is excellent but not exceptional. The 15-point delta is what makes him a real wet specialist rather than a fast driver in everything. The reputation got built in 2020 at Imola, when Russell led the Williams from sixth to second; the data ratifies that the underlying skill has held.

Leclerc’s number is similar, at 71% wet against a 56% dry h2h, and consistent with his Monaco 2024 drive and the 2022 Spielberg wet qualifying lap that won him pole by half a second. He is a genuine wet asset. The reputation is correct.

The overrated

This is the part of the piece where we have to be specific.

Lewis Hamilton sits seventeenth. Out of twenty rated drivers, Hamilton is fourth from the bottom in wet h2h. He runs 41% in the wet against a 59% dry h2h, meaning he is worse than his teammate in the wet substantially more often than in the dry. Delta: -17.8.

That number requires explanation, because it sits squarely against thirty years of accumulated “Hamilton is a rain master” commentary. Some of it is sample composition. Hamilton’s recent wet samples are dominated by 2018-2024, when his teammates were Bottas (consistently good in the wet himself, +0 delta), Russell (a top-tier wet driver), and Leclerc (also a top-tier wet driver). Hamilton lined up against the strongest set of wet teammates of any driver in our window. So a 41% wet h2h is partly a function of who he was racing.

But not entirely. Hamilton’s dry h2h against the same teammates is 59%; he beats them in the dry six times in ten. He is unambiguously slower in the wet relative to where he sits in the dry. The legendary wet drives of his career (Silverstone 2008, Hockenheim 2016, Istanbul 2020) happened in a window when his teammates were Heikki Kovalainen, Nico Rosberg, and Valtteri Bottas (year one). None of those drivers were the wet equals Bottas-era-late, Russell, or Leclerc became. The reputation was earned against weaker comparators and has not transferred to the stronger ones.

This is, we think, the central reframe of this piece. Hamilton is not a bad wet driver. He is a top-tier driver who is, by the data, a tier weaker in the wet than he is in the dry. The Hamilton-the-rain-master narrative is overstated. He is Hamilton-the-everything-master with a wet-rate slightly below his own dry rate.

Pierre Gasly (-28.1). Gasly’s reputation has always sat more with technical circuits than weather, but the data is unambiguous: dry h2h 58.8%, wet 30.8%. He loses to his teammate in the wet seven times in ten while beating them in the dry six in ten. That is a 28-point delta in the wrong direction. Among recent grid members, Gasly is the most clearly worse in the wet than the dry.

Lando Norris (-30.6). This one will surprise readers. Norris’s reputation is for consistency and increasingly for outright pace; he has been a podium-magnet across all conditions since 2024. But the wet h2h says something more specific: against Piastri specifically, Norris loses more often in the rain than he wins. Dry h2h: 80.6%, top of the field. Wet h2h: 50%, middle of the pack. The 30-point gap is the largest negative delta in the table.

Read alongside Piastri’s +43.8, this is the same data point viewed from each side. In the dry, Norris is dramatically faster than Piastri. In the wet, the gap collapses entirely. Whether the wet evens them out because Piastri has a specific wet skill, or because Norris has a specific wet deficit, the table can’t tell us. We can say with confidence that the Norris-Piastri matchup is much closer in the rain than it is in the dry.

Two patterns the data exposes

The “everything-master” pattern. Verstappen and Russell both rank in the top five for both wet and dry h2h. They are not “wet specialists” or “dry specialists.” They are drivers whose pace transfers across conditions almost perfectly. Verstappen’s wet 81 / dry 75 split says he is slightly better in the wet, which is consistent with his reputation, but the gap is small compared to drivers like Ocon and Russell where the wet shift is much sharper.

The “wet-conversion” pattern. Several drivers with weak overall dry h2hs find a meaningful equaliser in the wet. Ocon (dry 25%, wet 57%) and Piastri (dry 6%, wet 50%) are the clearest cases. Zhou (dry 27%, wet 44%) is a milder version. These drivers all get closer to their teammates in the rain than they manage in the dry. Whether the equaliser is technical (better wet feel) or strategic (better tyre management in changing grip) the data can’t isolate, but the fact of the equalisation is real.

There is also a quiet, opposite pattern: drivers like Albon and Norris, whose strong dry pace evaporates in the rain. Their wet h2h drops sharply. They look elite in normal conditions and merely competent in wet ones.

What we couldn’t measure

A piece like this has to be honest about its boundaries.

Sample size. Twenty drivers, six to seventeen wet h2hs each. Most ratings are reliable enough to put in writing; some (Piastri, Hülkenberg) are based on small samples and the rank is provisional. We require six wet h2hs as a minimum, but the difference between six and seventeen is meaningful. A driver with seventeen samples (Hamilton, Verstappen, Bottas) has a tighter confidence interval than a driver with six.

Pre-2018 history. FastF1 doesn’t cover seasons before 2018. So we can’t extend this analysis to Senna, Schumacher, Prost, or even mid-career Hamilton. The “historical overlay” of all-time wet greats would need a different data source, likely Ergast race results plus a manual labelling of which races were wet, and would produce a much coarser measure (wet win rate, wet podium rate) instead of the per-lap pace measure we’re using here. That’s a separate project. For now, this is a rating of every driver who has raced in the FastF1 era against the teammates they were actually paired with.

Track and condition heterogeneity. Wet at Spa is not the same as wet at Singapore. Heavy rain is not the same as a damp track drying out. We’re treating all wet laps equivalently, which is the right call for a single composite number but loses the texture between “racing in a downpour” and “racing on a drying line.” A more granular version of this rating would split each driver’s record by rain intensity, by track type, by year. We have the data; we don’t have the article-length budget. That’s a v2.

Teammate strength asymmetry. Hamilton’s wet record looks worse partly because he raced strong wet teammates. Piastri’s wet record looks better because he only has Norris on his card. The adjustment for teammate strength is what an Elo model would give you, and the engine actually has Elo-rated drivers from a related codepath. A future version of the Wet Index would feed wet h2hs into a wet-only Elo so that beating Verstappen in the rain counts more than beating Tsunoda. That would shuffle the rankings (Hamilton would rise a few places, Ocon might drop one), but we don’t think the broad reframes change. Norris would still be roughly even with Piastri in the wet; Hamilton would still be below his own dry rate; Verstappen would still be at the top. The deltas would tighten but not invert.

What this changes

Three calls the data argues against:

1. “Hamilton is the best wet driver of his generation.” Not in the FastF1-era data. He’s a top-ranked everything driver and a middle-ranked wet driver. The wet rate is below his own dry rate. The reputation was earned against an earlier era of teammates and has not transferred to the modern era. If “best wet driver of his generation” means anything in the post-2018 window, it means Verstappen, Russell, or Leclerc.

2. “Norris will dominate Piastri in a wet race.” Not according to the data so far. Dry, Norris wins six in ten convincingly. Wet, the matchup is level. If Round 5 in Montreal turns wet (forecast permitting) it’s a much harder call than the dry-condition odds suggest.

3. “Ocon is a steady, unremarkable driver.” False on the wet axis. Ocon is the most-improved-by-rain driver on the current grid. A wet weekend changes what he can credibly fight for. Strategy teams should be planning for that.

A few calls the data confirms:

1. Verstappen is the most all-condition driver in F1. True, with a small wet preference.
2. Russell and Leclerc are wet assets. True, with both sitting comfortably above 70% wet h2h.
3. Sergio Pérez was not a wet driver. True. Dry h2h 38%, wet h2h 33%. Already weak in the dry, no better in the wet.

What this piece is for

Like the Chaos Index, this is a reference piece. The numbers update after every wet race weekend. The next time a commentator says “X is a rain master” you can check whether the data agrees, whether the data disagrees, and how big the sample is.

The reputation economy in F1 is built on a handful of unforgettable laps from a handful of unforgettable races. The data economy is built on every lap of every race that’s been run for the last eight years. Both are useful. But when they disagree, the data wins. And right now, the data is saying that “rain master” is a less-distributed honour than the broadcast booth treats it as.