Wet weather splits: what a wet game actually changes

Thirteen of this season's 207 completed AFL games are flagged wet. Before asking what rain does to a player's numbers, it is worth knowing what “wet” means here and how seldom it happens.

What counts as wet

Every completed AFL game since 2019 and NRL game since 2021 carries weather, backfilled from hourly records at the venue's own coordinates. Rain is summed across the hour the game starts and the two after it, and a game is wet when that three-hour total reaches 1.0 millimetres. Wind and temperature are stored and shown nowhere, because nothing on the site has a use for them yet.

One millimetre over three hours is a low bar, and most wet games clear it by very little: of the 43 AFL games flagged wet since 2024, 17 had under 2 mm in the window, and of the 55 NRL games, 26did. “Wet” here mostly means light rain, not a downpour, and a split built on it should be read that way.

Coverage is the part worth stating plainly: the weather backfill is complete for every completed game in both codes. That matters because a game we held no weather for would fall into neitherbucket rather than quietly landing in “dry”, and there are none.

Why your twenty-game window probably has no wet split

Wet games are rare. In the AFL: 13 of 207 this season, 18 of 216 last year, 12 of 216 the year before — around 6%. In the NRL: 14 of 189, 23 of 213, 18 of 213 — around 7%. Twenty games at 6% is roughly one wet game, and the props card needs three on each side before it will draw a Wet chip or open a wet comparison.

So the counts come out where you would expect. Across the last twenty games of every player with a record in 2025 or 2026, only 85 of 818 AFL players and 71 of 622 NRL players have three or more wet games. For nine players in ten the Wet chip is absent, the wet-versus-dry door is not offered, and typing “does rain change it” into the question box returns no match. That last one is the correct answer rather than a missing feature: the site is saying it does not have the games to answer, which is a different thing from having nothing to say.

Where a wet split does exist for everybody

The wet-versus-dry island on a player page reads the whole multi-season record rather than the window on the props card, so a long-tenured player has a real wet count even when the last twenty games hold none. Joseph Tapine's page, on run metres, prints wet games at 130.1 off 16 games against dry games at 150.1 off 122, with the difference shown as −19.9. Under four wet games the island swaps its footnote for the count and the words thin evidence instead.

What the rain does, league-wide

Rather than trust an intuition about it, here is the whole of both competitions since 2024, per game, with both teams added together.

The AFL rows are the interesting ones. Disposals barely move: 721.0 in the wet against 718.5 in the dry, across 43 wet games and 596 dry. Marks fall hard, from 183.3 to 164.2. Tackles rise from 117.4 to 132.6, clangers from 111.9 to 121.1. In other words the rain does not take the ball out of an AFL game — it changes what happens to it. Fewer marks, more tackles, more mistakes, the same number of touches.

The NRL rows do show volume dropping, and that is a genuine difference between the codes rather than a rounding artefact: run metres 3,219 against 3,370, line breaks 8.51 against 10.13, tries 7.36 against 8.26, across 55 wet games and 560 dry. Every one of these figures is arithmetic we performed on the game rows — no feed publishes a wet-weather table — and every one of them rests on 43 or 55 games, which is thin for a league-wide claim and is why the denominators are printed beside them.

The roof

Docklands has a roof and the site knows it. Tonight's forecast is suppressed at a roofed ground: the conditions strip reads Roof — weather neutralinstead of a rain figure, and the projection's wet adjustment cannot fire there at all.

The history is the part that is not solved. The weather backfill stamps rain at the venue's coordinates onto every past game regardless of the roof, so 10 of the 132 Docklands games since 2024 are flagged wet on rain that may have fallen on a closed roof. The site does not record whether the roof was shut. The honest reading of a wet split for a Docklands-heavy player is that it is thinner than its count suggests. For comparison, the MCG — which is open to the sky, whatever the tile that once promised this post implied — has 15 wet games of 147 over the same period.

How to read a wet split

Count first: below four wet games the island says so itself. Then ask which stat, because the answer is not the same for all of them — marks and tackles move, disposals may not. Then check the venue mix, because a Docklands-heavy record carries the roof problem above. And if tonight's forecast is dry, the projection carries no wet row at all, which means the split is context rather than arithmetic.