There's no universal formula
Platform-wide averages can't tell you what wins for your side. Capture your own games and find the stat that lights up your wins.
Platform-wide averages tell you what tends to matter. They don't tell you what makes your team win. For one side in our data, the "most overrated" stat in football is their single best predictor.
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.
Everything we publish about what predicts winning comes with an asterisk: it's an average across the games captured on the Powercoach platform. Averages are useful — they cut through folklore — but they hide the thing a coach most wants to know. Your team isn't the average. You have a way of playing, a list, a set of strengths. What's your winning lever?
When we split the data team by team, the averages stop being the whole story.
The platform says one thing; individual teams say another
Take clearances. Across the whole platform, winning the clearance count is the weakest predictor of winning a quarter we measure — barely better than a coin flip. So "win the clearances" is, on average, weak advice.
But look at one particular team in our data:
For this side, the clearance count is their single strongest predictor of winning a quarter (r = 0.74) — not their weakest. They're built to win on the contested ball and convert that dominance. The platform-wide "clearances are overrated" verdict is exactly backwards for them. Telling this team to stop scoreboarding clearances would be telling them to ignore the thing they're best at.
Different teams, different engines
Look across the teams with enough captured football to read a pattern and the top levers genuinely differ:
| Team | Their distinctive winning lever | r |
|---|---|---|
| A | Contested ball — clearances (a coin-flip stat across the platform) | 0.74 |
| B | Possession — controlling the ball and grinding sides down | 0.68 |
| C | Transition — turning intercepts straight into scores | 0.71 |
| D | Territory — driving the ball into the forward 50 | 0.66 |
One thread does run through almost everyone: chain efficiency — the share of your possession chains that end in a score — shows up near the top for most teams. It sits closer to the scoreboard than the truly upstream stats (pressure, intercepts, possession) because scoring is baked into the ratio — but the way you push that ratio up is coachable: clean hands, hitting targets, taking the corridor option. That's the universal piece. The distinctive lever — the one in the table above — is where teams actually diverge: some win on territory, some on turnover, some on the contest.
That's the case for capturing your own football rather than coaching to platform averages. The benchmark that matters isn't ours — it's the pattern in your own games, the stat that lights up in the quarters you win and goes dark in the ones you lose.
What this means for your team
- Don't import another team's formula. "Win the clearances" or "it's all transition" might be gospel for the side you read about and irrelevant — or backwards — for you.
- Find the stat that moves with your results. After a handful of games, look at which numbers separate your wins from your losses. That's your lever. Defend it, build around it, and know when an opponent is taking it away.
- Use the platform average as a starting hypothesis, not a verdict. It tells you where to look first. Your own data tells you whether it's true for you.
- Almost everyone needs conversion. Chain efficiency turns up for nearly every team — making the most of what you create is close to universal. The differentiator is how you create it.
In short
- The averages hide your team. Across the platform, clearances are the weakest predictor of winning — yet for one side in our data they're the single strongest.
- Teams win on different levers — contest, possession, transition, conversion — and your distinctive one is yours alone to find.
- Chain efficiency is the near-universal exception — almost everyone needs to convert — but the second lever is what defines your style.
The capstone of this series — Correlation is not causation — names the broader principle: the most useful stat for a team is the upstream one they can actually move. For most teams that's a process lever; for some it's how they convert. Either way it's theirs.
A note on the data
This splits the per-quarter analysis by team and, for each side with enough captured football to read a signal, ranks which stat's advantage best tracks winning their quarters. These are indicative, not definitive — per-team samples are far smaller than the platform-wide pool, so treat a single team's ranking as a hypothesis to confirm with more football, not a settled fact. The point is the variety between teams, which is robust even where any one team's exact order isn't.
