Correlation isn't causation — the principle behind this whole series
Some stats predict winning because they help cause it. Others predict winning because they're the scoreboard wearing a different hat. The principle behind every finding in this series — with one stat that proves the point.
Why some stats predict winning and still tell you nothing. And one stat that proves the point.
New to r-values, p-values and confidence intervals? Stats made simple for coaches explains every number in this piece in plain English — no propellor hat required.
Across this whole series we've been doing something quietly deliberate: when a stat looks like it predicts winning, we've been asking why. Some stats predict winning because they help cause it — pressure, intercepts, clean ball use. Others predict winning because they're effectively just measuring the scoreboard from a different angle — inside 50s, scores per inside 50, expected points. Both kinds correlate with winning. Only the first kind is something a coach can actually do.
Take the most extreme version: "the team that kicks more goals usually wins." It's not quite a perfect correlation — a side that's accurate enough can occasionally win on behinds — but it's about as strong a link as football offers, and it's worse than useless as coaching advice. It tells you nothing about how to kick more goals. The correlation is real. The lesson is empty.
That's the gap between correlation (this number goes up when that number goes up) and causation (doing more of this helps make that go up). The correlation-isn't-causation rule is the most-quoted line in stats. The practical version for coaches is sharper:
Favour the stats that sit upstream of the result — the ones a coach can actually move.
That single rule is the one behind which stats Powercoach lifts up as KPIs and which it deliberately leaves out.
Upstream vs downstream — the shape of the game
Most of what happens on a footy field sits on a chain — each step a little closer to the scoreboard:
Inside 50s, scores per inside 50, intercepts that turn into scores — all of them correlate with winning. But they're a half-step away from the scoreboard itself. Saying "the team with more inside 50s usually wins" is not far from saying "the team that scored more usually won." True, but coaching gold? No.
The stats Powercoach lifts up are the upstream levers: pressure, chain efficiency, possession share, turnover counts. The ones you can drill on a Tuesday and see again on Saturday.
The killer worked example: total metres gained
The cleanest demonstration of why this matters is total metres gained — a stat almost every coach instinctively thinks should be useful. It isn't. Or rather, it is — but only as a measure of who's scoring.
Think about it. In a game of footy there are really only two ways to bank a net metres-gained advantage over your opponent:
- You kick a goal. The ball goes back to the centre bounce — and depending where you scored, you've just "gained" something like 80 metres of ball movement on the reset. Score more goals than they do, gain more net metres.
- A quarter ends with the ball in your half. Small swing. Four times a game. Nothing to do with how you're playing.
For the entire rest of the game, the ball goes back and forth. You gain 40 metres up the wing; they take 40 back the other way. Net effect: zero. The only thing that systematically moves the net-metres dial is scoring (and quarter-end position). So a team with more net metres gained has, almost by definition, scored more than its opponent.
That's why you won't see total metres gained on a Powercoach dashboard. It looks like insight. It's a tautology — "team A gained more metres" decodes to "team A scored more." See it elsewhere and you'll know how much weight to give it.
The redemptive nuance — aggregates vs by-source
Here's where it gets useful, though: metres gained *by source* is a different animal entirely. Kick-in metres gained, clearance metres gained, intercept metres gained — those isolate a single phase of the game.
- Low kick-in metres gained probably means your kick-in structure is bombing it into a pack.
- Low clearance metres gained probably means your midfield is winning first hands but not breaking the line.
- High intercept metres gained probably means you're punishing turnovers cleanly.
Each one is upstream of the scoreboard. Each one points at something you can actually drill. Powercoach tracks each separately, and they're genuinely useful.
The broader lesson the metres-gained example teaches:
Aggregate often becomes a proxy for the result. Broken down by source, the same idea can be a proper process stat.
Whenever a stat correlates almost perfectly with winning, ask yourself why. If the answer is "because it's basically the scoreboard," it's a downstream proxy and you can put it down. If the answer is "because the team that does this well tends to set up the conditions to win," it's a lever you can pull.
Pulling the whole series together
That lens has been quietly running underneath every piece in this series.
- Win the clearances, lose the game? — chain efficiency (upstream: ball use) out-predicts clearance counts (just-upstream: contested wins).
- Where do scores come from? — sources matter; aggregate scoring tells you less than how you got there.
- How pressure becomes points — the full upstream chain mapped: pressure → turnovers → scores.
- The forward-half lock-in — territory (time spent in the right half) is upstream of the turnovers you'll create there.
- Footy is a turnover game — winning back the ball is the cause; the points come after.
- There's no universal formula — every team's upstream lever is different; the downstream effect (winning) is the same.
Every finding sits on the same principle: prefer the upstream stat to the downstream one; prefer the cause to the symptom; prefer the thing you can coach to the thing you can only measure. Once you have that lens, every dashboard tells a different — and far more useful — story.
A note for the curious
This piece is the methodological capstone of the series. The metres-gained debunking is a known piece of stats folklore in Aussie Rules — coaches who think hard about ball movement land on it independently. The "by source" rescue (kick-in, clearance, intercept metres gained as separate stats) reflects how Powercoach actually structures its territory metrics.
